Chapter 1 draft

cocoddy

CMPT-DESSERTATIONPAPERAcceptance_of_Artificially_Int.pdf

Home >Business & Finance homework help >Management homework help >Chapter 1 draft

Acceptance of Artificially Intelligent Autonomous Self-Governing Technology

(AIASGT): A Qualitative Case Study

Dissertation Manuscript

Submitted to Northcentral University

Graduate Faculty of the School of Business and Technology Management

in Partial Fulfillment of the

Requirements for the Degree of

DOCTOR OF PHILOSOPHY

Dr. ROBERT E. CARMACK Jr.

Prescott Valley, AZ

November 2016

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

Microform Edition © ProQuest LLC.

ProQuest LLC. 789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, MI 48106 - 1346

ProQuest

Published by ProQuest LLC ( ). Copyright of the Dissertation is held by the Author.

ProQuest Number:

10244826

2016

Approval Page

Acceptance of Artificially Intelligent Autonomous Self-Governing Technology (AIASGT): A

Qualitative Case Study

Approved by:

Chair: MC. Clowes, Ph.D.

Certified by:

Dean of School: Dr. Peter Bemski

Robert Carmack

November 28, 2016

Date

Abstract

The purpose of this qualitative study was to examine the acceptance and use of

Artificially Intelligent Autonomous Self-Governing Technology (AIASGT). AIASGT

will give machines full autonomy, allowing them to make independent decisions without

any human intervention. No empirical research has been found regarding the impact of

what has been described as technology creep associated with the assimilation of

sophisticated Artificial Intelligence (AI) into autonomous systems. The focus of this

study was on the assimilation of AI into self-governing systems that are utilized by

members of the American Legion (AL). The AL sample pool was of those who have

used self-governing technology. Data collection began with a general search for a

purposeful sample of 10 retired or separated military personnel from the AL. The

expectation was that the study would show that the rules and regulations currently in

place for technology assimilation and development do not adequately address AIASGT.

The analysis of the findings of the study revealed the effect of AIASGT may have on the

Technology Acceptance Model (TAM). The recommendation is to specifically consider

AIASGT in all technology acceptance models. Another recommendation is for the

creation of a regulatory commission that will monitor and control AIASGT. The

consequence of not having the right constraints in place may become increasingly

difficult to correct.

iii

Table of Contents

Chapter 1: Introduction ........................................................................................................1

Background ..................................................................................................................... 3

Statement of the Problem ................................................................................................ 4

Purpose of the Study ....................................................................................................... 5

Theoretical Framework ................................................................................................... 6

Research Questions ....................................................................................................... 12

Nature of the Study ....................................................................................................... 13

Significance of the Study .............................................................................................. 14

Definition of Key Terms ............................................................................................... 16

Summary ....................................................................................................................... 18

Chapter 2: Literature Review .............................................................................................20

Documentation .............................................................................................................. 21

Technology Acceptance Model (TAM2) ...................................................................... 23

Technology Acceptance Model (TAM3) ...................................................................... 30

Unified Theory of Acceptance and Use of Technology (UTAUT) / TAM4 ................ 43

Self-Governance and Autonomous Technology ........................................................... 46

Advanced Autonomous Technologies .......................................................................... 49

AIASGT and Autonomous Technologies ..................................................................... 53

Ethical Considerations and Autonomous Technology .................................................. 54

AIASGT Moral, Ethical Adaptation, and Data Collection Integrity............................. 60

Various Developed Branches of Artificial Intelligence ................................................ 61

Applications of Artificially Intelligent Technology ...................................................... 64

AIASGT Software Development .................................................................................. 67

AIASGT Systems of Systems Common Operating Environment (SOSCOE) ............. 70

AIASGT Test and Evaluation ....................................................................................... 71

AIASGT Software Integration and Test (SWIT) .......................................................... 72

AIASGT Software Defect Documentation, Correction and Reintegration ................... 75

AIASGT and Cyber Security ........................................................................................ 78

Summary ....................................................................................................................... 81

Chapter 3: Research Method ..............................................................................................83

Research Method and Design ........................................................................................ 84

Population...................................................................................................................... 85

Sample ........................................................................................................................... 86

Materials/Instruments .................................................................................................... 88

Data Collection, Processing, and Analysis.................................................................... 90

Assumptions .................................................................................................................. 92

Limitations .................................................................................................................... 93

Delimitations ................................................................................................................. 95

Ethical Assurances ........................................................................................................ 96

Summary ....................................................................................................................... 97

Chapter 4: Findings ..........................................................................................................100

Results ......................................................................................................................... 103

Evaluation of Findings ................................................................................................ 109

Summary ..................................................................................................................... 113

Chapter 5: Implications, Recommendations, and Conclusions .......................................115

Implications ................................................................................................................. 118

Recommendations ....................................................................................................... 127

Conclusion ................................................................................................................... 135

References ........................................................................................................................138

Appendixes ......................................................................................................................143

Appendix A: Informed Consent Form .............................................................................144

Appendix B: Screening Demographic Questionnaire ......................................................145

Appendix C: Survey Questionnaire .................................................................................146

Appendix D: AIASGT Acceptance and Use Survey .......................................................147

Appendix E: Survey Participation Advertisement ...........................................................148

List of Figures

Figure 1. High Level Schematic of the Ethical Hybrid-Deliberative Reactive

Architecture....................................................................................................................... 60

Figure 2. Software Development Life Cycle ................................................................. 125

Chapter 1: Introduction

Many fields of robotics have been developed for U.S. military defense

applications as well as for industry, medical, and research fields (Singer, 2013). With the

employment of Artificially Intelligent Autonomous Self-Governing Technology

(AIASGT) utilized by the U. S. military, multiple permutations have been associated with

this use along with risks that have not been quantified with the rapid assimilation of

Artificially Intelligent (AI) powered machines. The U.S. military has utilized AI for

conducting military related purposes on land, sea, air, and in space (Singer, 2013). The

U.S. Air Force (USAF) categorized the use of AI for Aerospace, Command, Control,

Intelligence, Surveillance, and Reconnaissance (AC2ISR) that includes direct

engagement of hostile forces, and the sophistication and capabilities of AI has increased

exponentially (Sleeman, 2008).

The intent of the military is to assimilate autonomous technology into its war

making capabilities in an auspicious attempt to mitigate friendly force casualties (Arkin,

2011). Organizations like the U.S. Army have developed a continuous acquisition life

cycle at specific locations, such as Aberdeen Proving Ground (APG) Maryland for the

sole purpose of replacing and upgrading existing technologies to achieve full autonomy

(Arkin, 2011). A symbiotic relationship between the advances of autonomous and

robotic platforms means that as the technology maturates, robotic platforms will possibly

become increasingly autonomous (Singer, 2013). The incorporation of AI in systems,

such as the Predator Drone, whereby AI powered robotic platforms have made major

advancements, has increased the capabilities of the technology (Arkin, 2011). A purpose

of autonomous systems, such as the Predator Drone, is to eliminate the need for human

involvement or the physical requirement of deliberate execution (Arkin, 2011) and the

utilization of autonomous systems has given the U.S. military exponential flexibility in

pursuing the enemy along with a substantial mitigation of friendly force causalities

(Mackey, 2013). Mackey (2013) and Arkin (2011) noted a key goal for military

commanders is to mitigate friendly-force casualties by utilizing autonomous technology

that is fully independent of human interaction.

The ethical discussions associated with the use of autonomous technology are as

revolutionary as the technology itself (Arkin, 2011). Neutrality is considered nonexistent

in AIASGT and ethics are not a current consideration in the development of AIASGT.

The distinctive attribute of autonomy in robotics utilizes is revolutionary, producing

ethical dilemmas associated with its control, employment, and acceptance (Arkin, 2011).

Concerns that develop because of the acceptance and use of autonomous technology in

war will possibly generate unknown ethical problems (Sparrow, 2012). The technology

has unpredictable permutations and is different compared to legacy triggered systems that

followed a set state of rules, such as artillery (Sparrow, 2012). The U.S. military

currently uses autonomous robotic technology for conducting military related mission

purposes on land, sea, in air, and space. The USAF categorizes the use of these systems

for Aerospace Command and Control, Intelligence, Surveillance, and Reconnaissance

(AC2ISR) activities, which includes the direct engagement of hostile forces, however all

U.S. military branches utilize AIASGT. The sophistication and capabilities of

autonomous technology are increasing exponentially. The main impetus behind the

military’s current and future acquisitions strategy are based on the development of war

fighting machines that can operate with full autonomy (Singer, 2013).

Background

The rapid assimilation and sophistication of fully autonomous systems like

Unmanned Aerial Vehicles (UAVs) have unprecedented utilization and expansion in both

military and civilian applications (Donley, 2010). The systems that are currently

employed by the military for the specific purpose of taking human lives are incorporating

AI to ensure their full autonomy (Arkin, 2011). The omission of commands once the

system is deployed is the impetus of incorporating the increasingly sophisticated

technology. Researchers, such as Mackey (2013) and Donley (2010), noted the lack of

specificity in considering the permutations of autonomous technology and associated

dynamics with embedded AI on autonomous platforms and recommended case study

exploration. The emphasis on considering the acceptance of fully autonomous artificially

intelligent systems has not been fully theorized (Arkin, 2011).

The organizational theological perspective for studying the influences and

ramifications that fully autonomous self-governing systems will have on humankind will

be closely tied to existing and future military operations. It is the military that is

currently developing and exploiting the technology specifically developed to take human

lives. The utilization of autonomous platforms for the USAF is a top priority as the

organizational structure of the military is rapidly incorporating fully autonomous self-

governing systems into their daily operations (Donley, 2010). Artificially intelligent

platforms were a reality in today’s battle space (Sharkey, 2012). The organizational

theories associated with Sharkey’s research indicated the recalcitrant nature of military

commanders when questioned about their employment of autonomous systems against

human targets. Sharkey noted that there is an increased trend for regulation and

quantification of autonomous systems the more these platforms are utilized and theorized

that this new way of persecuting war will be different, especially when evaluating their

use in real world combat-related scenarios. The organizational theological perspective is

relevant to the theory behind the military’s use of fully autonomous systems and its

increases assimilation into the military’s organizational structure.

The economical theological perspective for studying the influences and

ramifications are viewed as being associated to the budgets. The single greatest theme to

emerge from technology horizons is believed to be based on the need, opportunity, and

potential to dramatically advance technologies that can allow the Air Force to gain the

increased capability, manpower efficiencies, and cost reductions available through far

greater use of autonomous systems in essentially all aspects of Air Force operations

(Donley, 2010). The increased utilization of fully autonomous self-governing systems

are posited to increase because of the noted economic factors and advantages (Donley,

2010).

Statement of the Problem

The problem is the opinions of ex-military personnel who have experience with

the operation and development of AIASGT have not been consulted concerning its actual

employment. The rapid proliferation of autonomous, self-governing platforms and the

secret nature of their employment by the U.S. military have increased pressure to fully

quantify and delineate their capabilities (Singer, 2013). Arkin (2011) and Geis (2011)

both noted that new technologies used by the military have historically made killing more

efficient than the longbow, artillery, armored vehicles, aircraft carriers, or nuclear

weapons. Autonomous self-governance has not been identified as a technology that

requires comprehensive policies and procedures to mandate compliance within known

and acceptable parameters (Arkin, 2011; Geis, 2011; Singer 2013).

The determination of the military’s use of self-governing systems is frequently

based upon observation of their employment on the battlefield and the belief that the uses

of these systems are based on specific methodologies, and much of the current research in

autonomous systems is often viewed as a tool capable of some autonomy that a remote

human operator commands (Arkin, 2011). Systems, such as Pathfinder, an unmanned

combat vehicle that can accompany or precede a manned platform into adversarial

territory, has created the specific problem related to the ethical concerns of an

autonomous system that is currently terminating human lives (Arkin, 2011). Pathfinder

and similar systems are capable of launching and controlling fleets of small, unmanned

vehicles without human intervention, and capable of defense while attacking a range of

targets without human intervention or control (Geis, 2011; Singer, 2013). The autonomy

demonstrated by systems, such as Pathfinder, result in the perceived problem of omitting

an operator who is capable of controlling the actions of the system (Geis, 2011) and the

autonomous nature has raised ethical concerns related to autonomous technology that has

the omission of the human element (Arkin, 2011; Geis, 2011; Singer, 2013).

Purpose of the Study

The purpose of this qualitative case study was to explore the perceptions of AL

members who have developed or used some form of artificially intelligent self-governing

technology designed to take human lives. Only AL members who were knowledgeable

of AIASGT were asked to participate. The study objective was to explore the

perceptions of AL members for two study constructs: the acceptance of artificially-

intelligent technology (Arkin, 2011; Clarke, 2011; Cook, 2013; Davis, 1989; Geis, 2011;

Kroeker, 2011; Mackey, 2013; Singer, 2013, Sparrow, 2012), and autonomous self-

governance (Arkin, 2011; Brock, 2010; Clarke, 2011; Cook, 2013; Erlebach 2013; Geis

2011; Ghazizadeh et al., 2011; Phalp, 2008; Singer, 2013; Sleeman, 2008, Sparrow,

2012; Valavanis, 2007). A qualitative study was appropriate in this type of research as

the qualitative instruments needed to measure the impact of autonomous systems do not

yet exist (Yin, 2013). Davis’ (1989) TAM and its five factors (external factors, perceived

usefulness, perceived ease of use, intention to use, and attitude to using) will be used as

the theoretical lens of this study. The target population was 600 ex-military personnel

associated with the AL Post 346 who may have experience operating AIASGT and

purposeful sampling was used as it was best suited to construct multi-perspectival,

emancipatory, participatory, and deconstructive interpretations for qualitative data

collection (Harsh, 2011). Purposeful screening of the ex-military personnel associated

with the AL Post 346 was used to determine if they have experience with AIASGT. The

screened personnel were the participant pool for the study. The utilization of a

purposeful sample of 10 ex-military personnel associated with the AL Post 346 who have

used self-governing technology on the battlefield were accepted as participants. Data

collection was in the form of web-based questionnaires to conduct a systematic content

analysis for emergent themes (Harsh, 2011; Yin, 2013).

Theoretical Framework

The theoretical framework for this study was the technology acceptance model

(TAM; Davis, 1989), which is rooted in the Theory of Reasoned Action (TRA)

developed by Ajzen and Fishbein (1973). TRA is used for the prediction of behavioral

intention, spanning predictions of attitude and predictions of behavior, and the

subsequent separation of behavioral intention from behavior allows for explanation of

limiting factors on attitudinal influence (Ajzen & Fishbein, 1973). This theory serves as

an appropriate framework for this study to view AIASGT and the effects on behavioral

intention within a military setting. The military’s use of technology has been centric to

the termination of human lives (Singer, 2013), and theory is appropriate for the

consideration of theoretical positions to understand the risk factors in technology

adoption and ethical challenges in the deployment of autonomous technology for military

purposes (Singer, 2013).

TRA is used for the prediction of behavioral intention (Ajzen & Fishbein, 1973).

TRA theorists explain the intention to perform a behavior is a function of attitudes toward

engaging in the behavior and perceived normative pressure to perform the behavior

(Ajzen & Fishbein, 1973). This means that behavior is a function of the attitude towards

a specific action and the subjective norms regarding that action. Performing a behavior is

often a function of intention to perform the behavior and attitudes do not directly predict

behavior, but more so to predict intention. Norms do not directly predict behavior, they

are used to predict intention and intention predicts behavior (Ajzen & Fishbein, 1973).

The concept of TRA is based on what individuals think regarding the consequences and

implications of their actions occurs before they decide whether or not to take action.

Furthermore, intention was highly correlated with behavior (Ajzen & Fishbein, 1973).

The acceptance of AIASGT and its influence on adoption, acceptance, assimilation, and

use from the lens of TRA was a component of the theoretical framework for this study.

The opinions of ex-military personnel who have the operations experience

employing AIASGT were advertisement to the chosen theoretical framework. The

intention of those who operate AIASGT is significant in that it allows for assessment of

unadulterated opinions. Ex-military personnel are especially significant because they are

civilians with real-world operational experience. The predictions of attitude and behavior

is different for ex-military because they are no longer actively conducting military

operations. They have had time to reflect on the impact of AIASGT and have opinions

that can shape the way national policy is made. The behavior of ex-military is different

than that of an active duty participant of military operations. Many ex-military personnel

may have a different opinion because of their chance to reflect upon the current and

continued use of AIASGT. The subsequent separation of the behavioral intention of ex-

military personnel from the behavior of current operators allows for explanation of

limiting factors on attitudinal influence.

Behavioral intention based on TRA subsequently leads to a connection with

TAM. Davis (1989) postulated on what he identified as the Technology Acceptance

Model (TAM) theory, and identified five elements for any technology to be accepted

overall. These five elements are (a) external factors, (b) perceived usefulness, (c)

perceived ease of use, (d) intention to use, and (e) attitude to using (Davis, 1989). An

expansion of the TAM theory created a theoretical postulation that would consider the

element of autonomous, artificially intelligent self-governance.

TAM is a theory application that is employed to accurately describe the general

acceptance of technology (Davis, 1989), but has not been expanded upon regarding the

acceptance or exercise of the dynamics associated with the increasingly sophisticated

technology of AI. The utilization of artificially intelligent systems that make independent

human-free decisions into the existing TAM has not been fully considered or theorized

(Arkin, 2011). The example of AIASGT for military applications that will ultimately

take human lives is an aspect of the acceptance of technology that would expand the

existing TAM model.

Numerous empirical studies have indicated that TAM could be applied to

consistently explain a substantial proportion of the variance (typically about 40%) in

usage intentions and behavior of those who use technology and that TAM compared

favorably with alternative models, such as the TRA (Venkatesh & Davis, 2000). The use

and public acceptance of any technology (including AIASGT) are dependent on the five

factors delineated in TAM (Davis, 1989). The military’s use of AIASGT would further

test theory when introduced as a consideration as continued research using the TAM

framework may benefit from examinations of AIASGT as a determining factor for the

overall acceptance of such technology for military applications. Davis’ (1989) five

external factors would need consideration of a sixth factor. This factor would be

AIASGT and the unpredictability of this type of technology. The construct of

autonomous technology would enhance the TAM external factors that are associated with

the overall acceptance of autonomous technology, as the military’s use of autonomous

technology has been conducted in secret, thereby limiting the postulation or theoretical

acceptance of AI for military purposes (Singer, 2013). Concerns regarding the secretive

nature of the development of AI may allow the development of questionable technologies

(Singer, 2013).

Davis (1989) noted technology adoption and use in the workplace has remained a

central concern of information systems research and practice. This is important for this

study because the postulation of AIASGT will affect the adaptation and use of

technology in the workplace. The use of AIASGT and its effects on the workplace would

be to test theory in a new field within a military setting. Despite impressive advances in

hardware and software capabilities, the problem of underutilized or undesired systems

has continued (Venkatesh & Davis, 2000). To understand how TAM could be used to

predict conditions under which technology can be embraced by human organizations has

remained a high-priority for all aspects of technology adoption (Venkatesh & Davis,

2000), which would include the military’s use of AIASGT. Technology acceptance, and

the conditions under which technology will be embraced by human organizations, can

and will affect the acceptance of the military use of AIASGT once it is understood that

machines are designed to autonomously kill a human being. The acceptance of AIASGT

and its effect on adoption, acceptance, assimilation, and use from the lens of TAM is a

component of the theoretical framework for this study.

Venkatesh and Davis (2000) suggested that another important avenue for future

research concerned the temporal dynamics of the determinants of user acceptance. A

theoretical extension of TAM known as TAM2 was developed and tested from pre-

implementation to post-implementation. The perceived usefulness and intent to use were

quite stable in the expanded TAM2 (Venkatesh & Davis, 2000). However, perceived

ease of use, in contrast, appeared to be less stable over time, which attributed to the

known role of direct hands-on experience in forming this belief (Venkatesh & Davis,

1996, 2000). Consideration of autonomous self-governance would likely test a theory in

a new field. The methodologies of theory development by Venkatesh and Davis (1996,

2000) would benefit with the consideration of what autonomous technology would have

on the theory. TAM is not specific to the development of autonomous self-governing

technology; however, Venkatesh and Davis (2000) noted that TAM2 focused on a

number of related criteria that determine technology acceptance. The consideration of

the changes that AIASGT will have on the overall acceptance of technology would

expand TAM. The self-governance and ability of technology to make independent,

artificially intelligent decisions could affect the way technology is accepted. The

parameters identified in TAM and TAM2 for the acceptance of technology within a

military setting were used to postulate the military’s use of AIASGT and serve as a

theoretical lens to explore the problem of AIASGT used by the USAF (Davis, 1989;

Venkatesh & Davis, 1996, 2000). As a combined theoretical framework, TRA (Ajzen &

Fishbein, 1973), Davis’ (1989) TAM, and Venkatesh and Davis’ (2000) TAM2 will

contribute to the TAM literature within a military context.

When inserting the element of AI into TAM, the permutations that AI causes will

possibly alter all of the known factors identified by applying TAM. In particular, AI will

probably alter the intent to use as well as the attitude of use in unpredictable ways. The

intent to use will possibly be permutated by the unknown variables introduced by the

self-governing properties of AI. This independence is desirable with systems that the

military develops and classifies as “fire and forget.” Therefore, it makes sense that any

research regarding AI in relation to technology acceptance would likely be aligned, at

least in part, with TAM.

The formulation of a theological educational perspective is the possible influence

on the associated technological advances and benefits of the technology. Theological

studies, like the Phalp (2008) study on the increased sophistication of platforms that

utilize the particle swarm guidance system, further indicate the need for a new look at

TAM, especially when it is applied to fully autonomous, artificially intelligent platforms

designed to take human lives. The benefits for the formulation of a theological

educational perspective will center on the permutations associated with the advancement

of multiple technologies associated with the maturation of autonomous technology.

Research Questions

The background of the questions posted are directly associated with autonomous

systems that have the ability to make independent decisions that will or may reasonably

be expected to terminate a human life. The independent functionality and the ability to

make reasonable decisions in relation to terminating a human is the impetus for the

questions. The reasoning that these systems are being integrated into the military’s

defense methodology and the acceptability of systems that can make independent

decisions along with the uncertainty of their continued acceptance is a theory needing

expansion and testing. The study was inductively built on the reasoning behind the use of

AIASGT integrated into the U.S. military’s defense methods from the user perspective.

Q1. How do ex-military personnel perceive the usefulness of AIASGT?

Q2. What are the perceptions of ex-military personnel regarding the attitude to

use AIASGT?

Nature of the Study

The impetus of this qualitative case study is to explore the perceptions of military

personnel who administrate artificially intelligent self-governing technology designed to

take human lives. The study objective is to explore the perceptions of the military

personnel for two study constructs: the acceptance of artificially-intelligent technology

(Arkin, 2011; Clarke, 2011; Cook, 2013; Davis, 1989; Geis, 2011; Kroeker, 2011;

Mackey, 2013; Singer, 2013, Sparrow, 2012), and autonomous self-governance (Arkin,

2011; Brock, 2010; Clarke, 2011; Cook, 2013; Erlebach 2013; Geis 2011; Ghazizadeh et

al., 2011; Phalp, 2008; Singer, 2013; Sleeman, 2008, Sparrow, 2012; Valavanis, 2007).

A case study design was used in order to develop an in-depth depiction of the

participants’ perspectives (Yin, 2013) regarding the acceptance and use of the AIASGT

on the battlefield.

A qualitative study was appropriate for this research as the quantitative

instruments needed to measure the impact of autonomous systems do not yet exist (Yin,

2013). Davis’ (1989) TAM and its five factors (external factors, perceived usefulness,

perceived ease of use, intention to use, and attitude to using) was used as the theoretical

lens of this study. The study target population is 600 ex-military or ex-military

associated with AL Post 346 who has operated AIASGT, and the target population served

as the study-sampling frame. Purposeful sampling was used as it is best suited to

construct multi-perspectival, emancipatory, participatory, and deconstructive

interpretations for qualitative data collection (Harsh, 2011), and gather a purposeful

sample of a minimum of 10 ex-military personnel from the AL active duty and ex-

military personnel who have used AIASGT on the battlefield. Data collection was in the

form of web-based surveys upon email receipt of informed consent (see Appendix A).

Participants were identified by their responding to an AL Post 346 newsletter and

webpage article that delineates procedures on how to access the Google Docs

questionnaire for data collection. A letter of collaboration (see Appendix B) was received

from the AL Post 346, which is the participating organization.

Data collection involved an open-ended survey questionnaire regarding their

acceptance and perceptions concerning the use of AISAGT (see Appendix C). Online

data collection was conducted through the utilization of Google Docs to deliver the

informed consent form and questionnaire to the sampling frame. Data analysis was

manually processed to conduct a systematic content analysis for emergent themes (Harsh,

2011; Yin, 2013). The results allowed the downloading and storage of the data as an Excel

table once the respondents populated the questionnaire and were reviewed for missing data,

coded and analyzed.

Significance of the Study

The U.S. armed services have been affected by the increased use of fully

autonomous robotic systems in both civilian and military applications (Sharkey, 2012),

and perceptions and acceptance of those that have used these systems to take human lives

is something that the military should be concerned about because the nation’s acceptance

in the global village could be affected if this technology is not employed correctly (Arkin,

2011; Geis, 2011; Sparrow, 2012; Singer, 2013). The civilian use of autonomous systems

is on the rise, especially with law enforcement and for agricultural applications (Singer,

2013), and the proposed study may contribute to the worldview of the use of AIASGT by

relating the use of AIASGT from the user and or operator’s perspective to better

understand user perceptions and experiences in the acceptance of AIASGT. Singer’s

(2013) findings were important to consider for this study as they illustrated the

proliferation and associated challenges with military developed AI used for civilian

applications as the study may be contributive to examples of how AI affects many aspects

of autonomous technologies (Singer, 2013). For example, maritime applications have

increased so much that sailors are becoming extinct in many maritime occupations, and

the use of autonomous systems is fast becoming the preferred choice of military

commanders on the battlefield (Mackey, 2013). The significant contribution that

Mackey’s (2013) findings support the need for further study of AI as it identified the

inevitable exponential proliferation of AI technology due to its efficiency, utility, and

mitigation of friendly force casualties. Mackey also noted that the proliferation of

AIASGT is devoid of any regard of acceptance of the technology and based solely on the

needs of military commanders and the efficiencies the technology creates.

There has been an increased trend for regulation and quantification of

autonomous systems the more these platforms are utilized and theorized that this new

way of persecuting war is different, especially when evaluating the use autonomous

systems in real world combat related scenarios (Sharkey, 2012). Since constructivism, as

it applies to how the population is affected by the military’s employment of fully

autonomous robotic self-governing platforms, is defined as active construction of new

knowledge based on prior experience, the worldview of military personnel and others

may benefit from the military being as transparent as possible (Singer, 2013). Singer’s

(2013) findings further support the study of AIASGT by identifying the lack of data

concerning the acceptance or the perception by those that use and manufacture AIASGT.

Singer (2013) also reported this will be mainly because of the secretive nature of

AIASGT’s development and deployment, especially in hostile environments, and called

for transparency by the manufactures and the military’s commanders that use AIASGT.

This transparency may allow all people, who have any knowledge or firsthand experience

in the military’s use of fully autonomous robotic self-governing systems that are used to

take human lives, the opportunity to strengthen their knowledge to make rational and

informed decisions on the practicality and utility of these systems (Arkin, 2011; Geis,

2011; Singer, 2013; Sparrow, 2012). Since constructivism is defined as active

construction of new knowledge and the subject mediates input from the outside world to

determine what the person will learn, there is the potential for a dichotomy of opinions

concerning the utility and ethicality of the military’s use of these systems to take human

lives (Arkin, 2011; Geis, 2011; Sparrow, 2012; Singer, 2013). These varying views and

opinions support the need for the study to explore the worldview and associated influence

on society from the user’s perspective (Arkin, 2011; Davis, 1989).

Definition of Key Terms

Artificial intelligence (AI). The term AI is described as the computer science

that creates the intelligence of machines. Traditional AI concepts, such as pattern

recognition, numerical optimization, and data mining, are simple types of AI algorithms

(Kroeker, 2011).

Autonomous. Acting independently and having the ability to execute the

independence is referred to as autonomous. A crucially important aspect for mission-

critical robotic operations is ensuring as best as possible that an autonomous system be

able to complete its task (Erlebach, 2013).

Permutation. Permutation is the act of permuting or permutating; alteration; or

transformation. A transformation or alteration is referred to as a permutation (Ching,

2010).

Self-governance. Self-governance is act of exercising control or rule over

oneself or itself. Self-governance means self-organizing and independent of human

intervention. (Arkin, 2011).

Semi-autonomous. The term semi-autonomous refers to partial self-government

acting semi-independently. Semi-autonomy involves a team composed of robots and

humans that comprises the overall system (Franceschini, 2010).

System. A system is a group of interacting, interrelated, or interdependent

elements forming a complex whole. This term is synonymous with the term sensor-based

autonomous system, which can be an automatically or autonomously self-governing

system, including both hardware and software components (Geis, 2011).

Subsystem. A subsystem is a system that is part of some larger system.

Subsystem is a term used to describe either hardware or software components that form a

system (Geis, 2011).

System’s context. The operating environment the system will perform in and

will be designed for is referenced as the system’s context. This includes all

environmental as well as technical interrelationships and dependencies (Geis, 2011).

Surroundings. Surroundings are the external objects, conditions, and

circumstances that affect existence and development; the environment. These elements

are space, land, sea, and air (Arkin, 2011).

Summary

The key point of the study was to postulate the acceptance and the use of

AIASGT by the personnel who have utilized and developed these technologies. The

study purpose surrounds the acceptance of AIASGT. The research plan involved use of a

purposeful sample of 10 personnel that have used AIASGT. Data collection were in the

form of web-based surveys, and data analysis was processed by the software application

to conduct a systematic content analysis for emergent themes. The background statement

delineated the rapid assimilation of AIASGT in the military. The emphasis on

considering the acceptance of fully autonomous artificially intelligent systems has not

been fully theorized (Arkin, 2011). The statement of the problem noted the rapid

proliferation of autonomous, self-governing platforms and the secret nature of their

employment by the U.S. military. The problem is that autonomous self-governance has

not been identified as a technology that requires comprehensive policies and procedures

to mandate compliance within known and acceptable parameters (Arkin, 2011; Geis,

2011; Singer 2013). The purpose of the study is to explore the perceptions of the military

personnel that utilize AIASGT. The acceptance of the technology and the willingness to

use the technology and its associated impact has not been measured (Arkin, 2011; Clarke,

2011; Cook, 2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013,

Sparrow, 2012).

The theoretical framework for this study is the Technology Acceptance Model

(TAM; Davis, 1989), which is rooted in the theory of reasoned action (TRA) developed

by Ajzen and Fishbein (1973). TRA theorists explain the intention to perform a behavior

is a function of attitudes toward engaging in the behavior and perceived normative

pressure to perform the behavior (Ajzen & Fishbein, 1973). Davis (1989) postulated on

what he identified as the Technology Acceptance Model (TAM), and identified five

elements for any technology to be accepted overall. These five elements are (a) external

factors, (b) perceived usefulness, (c) perceived ease of use, (d) intention to use, and (e)

attitude to using (Davis, 1989). The research questions are directly associated with

autonomous systems (AIASGT) that have the ability to make independent decisions that

will or may reasonably be expected to terminate a human life. The nature of this study is

to explore the perceptions of the military personnel that operate AIASGT. A qualitative

study is appropriate for this research as the quantitative instruments needed to measure

the impact of autonomous systems do not yet exist (Yin, 2013).

The significance of the study may Advertisement to the worldview by correlating

the advancements in AIASGT to the users or operators perspective. The rapid

assimilation of AIASGT and the impact of the technology on the users, operators, or

targeted population have not been quantified (Singer, 2013). The key terms delineated

the significant terms associated with AIASGT. AI, as defined by Kroeker (2011), and

self-governance (Arkin, 2011) are the foundation for discussing AIASGT.

Chapter 2: Literature Review

Information technology (IT) and the development of AIASGT provide ease of use

and are rapidly introduced to the consumer, however, there are no limits currently set by

the international community on how autonomous technology is accepted or developed. If

the technology has a degree of perceived safety, society is quick to adapt the technology

simply because human nature is the path of least resistance and ease. Increase in the

innovations associated with AIASGT have awesome advantages and have brought

changes to human life and work endeavors. As people, organizations, and governments

move toward the use of AIASGT, interaction with this technology will cause great debate

(Singer, 2013), but the debate issues are not currently considered in any of the current

technology acceptance models.

Most consumers of technology assume that computers run on electricity and that

they use binary logic to carry out programmed instructions (Popkin, 2015). Actually,

computers shuttle information using materials known as semiconductors and its brains

are built on integrated circuit chips packed with tiny switches known as transistors

(Popkin, 2015). In the nearly 70 years since the first modern digital computer that were

built, the above specs have become all but synonymous with computing (Popkin, 2015).

A computer is defined not by a particular set of hardware, but by being able to take

information as input and to change or “process” the information in some controllable way

in order to deliver new information as output (Popkin, 2015). This information and the

hardware that processes it can take an almost endless variety of physical forms (Popkin,

2015).

Over nearly two centuries, scientists and engineers have experimented with

technology driven computerized designs that use mechanical gears, chemical reactions,

fluid flows, light, DNA, living cells, and synthetic cells (Popkin, 2015). Many of the new

technologies, such as chemical, wetware, fluidic, and ternary computing, will further

revolutionize AI and the reasonable expectation is that AI will become much more

capable in the near future (Singer, 2013). However, interaction between humans and

AIASGT are affected by quite a number of human factors that are not singled out or even

considered in the current technology models.

Following the documentation section, the chapter begins with four sections

devoted to the various models of the TAM. This is followed by a discussion of self-

governance and autonomous technology, advanced autonomous technologies, AIASGT

and autonomous technologies, ethical considerations and autonomous technologies, and a

summary.

Documentation

Peer-reviewed journals were the primary means of literary review. A high

percentage of articles were obtained from academic databases, such as Emerald Insight,

ProQuest, e-books, and EBSCOHost’s Business Source Premier. Several military

publications in the form of books or journals were also reviewed. TAM was developed

based on the theorization that an individual’s behavioral intention to use a system is

determined by two beliefs: perceived usefulness, defined as the extent to which a person

believes that using the system will enhance his or her job performance, and perceived

ease of use, defined as the extent to which a person believes that using the system will be

free of effort (Davis, 1989). The effects of external variables (e.g., system

characteristics, development process, training) on intention to use are believed to be

mediated by perceived usefulness and perceived ease of use (Davis, 1989). Based on

TAM, perceived usefulness is also influenced by perceived ease of use because, other

things being equal, the easier the system is to use, the more useful it can be (Venkatesh &

Davis, 2000).

Since perceived usefulness is such a fundamental driver of usage intentions, it is

important to understand the determinants of this construct and how their influence

changes over time with increasing experience using the system (Venkatesh & Davis,

2000). Perceived ease of use, TAM’s other direct determinant of intention, has shown a

less consistent effect on intention across studies (Venkatesh & Davis, 2000). A better

understanding of the determinants of perceived usefulness would enable the designing of

organizational interventions that would increase user acceptance and usage of new

systems. Therefore, the goal of the present research is to extend TAM to include

Advertisementitional key determinants of TAM’s perceived usefulness and usage

intention constructs and to understand how the effects of these determinants change with

increasing user experience overtime with the target system (Venkatesh & Davis, 2000).

The reality is that neither of the current TAM models includes consideration of

the impact of AIASGT used for military purposes. The integration of AI systems that

make independent human-free decisions into the existing TAM has not been fully

considered or theorized (Arkin, 2011). The assimilation of AIASGT for military

applications that will ultimately take human lives as an aspect of the acceptance of

technology would expand the existing TAM models.

Technology Acceptance Model (TAM2)

The first of the variations is called the TAM2 (Taiwo & Downe, 2013). As with

the original Davis (1989) TAM, TAM2 does not include the consideration of the

acceptance or use of AIASGT for any purpose, including use by the military. Using the

original TAM as a starting point, TAM2 incorporates theoretical constructs spanning

social influence processes (subjective norm, voluntariness, and image) and cognitive

instrumental processes (job relevance, output quality, result demonstrability, and

perceived ease of use). Below, each of these constructs is defined along with the

theoretical rationale for the causal relationships of the TAM2 (Venkatesh & Davis, 2000).

TAM2 reflects the impact of three interrelated social forces impinging on an

individual facing the opportunity to adopt or reject a new system: (a) subjective norm, (b)

voluntariness, and (c) image (Venkatesh & Davis, 2000). The extended model, referred

to as TAM2, was tested using longitudinal data collected regarding four different systems

at four organizations: two involving voluntary usage and two involving mandatory usage

(Venkatesh & Davis, 2000). Model constructs were measured at three points in time at

each organization: (a) pre-implementation, (b) 1 month post implementation, and (c) 3

months post implementation (Venkatesh & Davis, 2000). The extended model was

strongly supported for all four organizations at all three points of measurement,

accounting for 40–60% of the variance in usefulness perceptions and 34–52% of the

variance in usage intentions (Venkatesh & Davis, 2000).

Both social influence processes (subjective norm, voluntariness, and image) and

cognitive instrumental processes (job relevance, output quality, result demonstrability,

and perceived ease of use) significantly influenced user acceptance (Venkatesh & Davis,

2000). These findings were applied to advance theory and contribute to the foundation

for future research aimed at improving understanding of user adoption behavior

(Venkatesh & Davis, 2000).

The direct effect of subjective norm on intentions for mandatory usage contexts

were theorized to be strong prior to implementation and during early usage, but weaken

over time as increasing direct experience with a system provides a growing basis for

intentions toward ongoing use (Venkatesh & Davis, 2000). The authors expectations

were that the effect of subjective norm on perceived usefulness (internalization) to

weaken over time, since greater direct experience will furnish concrete sensory

information, supplanting reliance on social cues as a basis for usefulness perceptions

(Venkatesh & Davis, 2000). In contrast, the influence of image on perceived usefulness

(identification) to weaken over time since status gains from system use will continue as

long as group norms continue to favor usage of the target system is not expected

(Venkatesh & Davis, 2000).

One key component is a potential user’s judgment of job relevance, which is

defined as an individual’s perception regarding the degree to which the target system is

applicable to his or her job. In other words, job relevance is a function of the importance

within one’s job of the set of tasks the system is capable of supporting (Venkatesh &

Davis, 2000). TAM2 is posited to be over and above considerations of what tasks a

system is capable of performing and the degree to which those tasks match their job goals

(job relevance), which people will take into consideration as to how well the system

performs those tasks (i.e., perceptions of output quality; Venkatesh & Davis, 2000).

Empirically, the relationship between perceived output quality and perceived usefulness

has been shown before (Venkatesh & Davis, 2000). The expectation is for output quality

to be empirically distinct from and to explain significant unique variance in, perceived

usefulness over and above job relevance because a different underlying judgmental

process is involved (Venkatesh & Davis, 2000). Even effective systems can fail to garner

user acceptance if people have difficulty attributing gains in their job performance

specifically to their use of the system (Venkatesh & Davis, 2012).

Individuals can be expected to form more positive perceptions of the usefulness of

a system if the covariation between usage and positive results is readily discernable

(Venkatesh & Davis, 2000). Conversely, if a system produces effective job relevant

results desired by a user, but does so in an obscure fashion, users of the system are

unlikely to understand how useful such a system really is (Venkatesh & Davis, 2000).

TAM2 retains perceived ease of use from the original TAM as a direct determinant of

perceived usefulness (Davis, 1989) since, all else being equal, the less effortful a system

is to use, the more using it can increase job performance (Venkatesh & Davis, 2000). The

authors suggested that there is extensive empirical evidence accumulated over a decade

that perceived ease of use is significantly linked to intention, both directly and indirectly,

via its impact on perceived usefulness (Venkatesh & Davis, 2000).

Although beyond the scope of the present extension of TAM, other research has

begun to be modeled after the antecedents of perceived ease of use (Venkatesh & Davis,

2000). For example, Venkatesh and Davis’ (1996) model perceived ease of use will be

anchored to the aspect of one’s general computer self-efficacy and adjusted to account for

a system’s objective usability via direct behavioral experience using the target system

(Venkatesh & Davis, 2000). The mechanisms by which ease of use perceptions are

theorized to form and evolve are conceptually distinct from, and complementary with, the

social influence and cognitive instrumental processes within TAM2 (Venkatesh & Davis,

2000). Consistent with the theories of mental representation discussed above, the

expectation is that even over time, people will continue to rely on the match between

their job goals and the consequences of system usage (job relevance) as a basis for their

ongoing usefulness perceptions (Venkatesh & Davis, 2000). Just as the role of what a

system does remains influential, how well a system does what it does (output quality)

will remain a significant determinant of perceived usefulness over time (Venkatesh &

Davis, 2000). Similarly, there is no basis to expect the effect of result demonstrability on

perceived usefulness to become either stronger or weaker over time (Venkatesh & Davis,

2000).

Although there is no theoretical account of such temporal shifts, the effect of

perceived ease of use on perceived usefulness has sometimes been found to increase over

time, whereas the direct effect of perceived ease of use on usage intention has been

observed in some research to decrease over time (Davis, 1989), and in other research to

increase over time (Venkatesh & Davis, 2000). Presently, there is a lack of sufficient

theoretical rationale to hypothesize specific temporal shifts in the strength of any of the

effects of cognitive instrumental determinants (job relevance, output quality, result

demonstrability, or ease of use; Venkatesh & Davis, 2000). TAM2 will be strongly

supported across four organizations and three points of measurement (pre-

implementation, one month post-implementation, and three months post-implementation;

Venkatesh & Davis, 2000).

Encompassing both social influence processes (subjective norm, voluntariness,

and image) and cognitive instrumental processes (job relevance, output quality, result

demonstrability, and perceived ease of use; Venkatesh & Bala, 2008), TAM2 provides a

detailed account of the key forces underlying judgments of perceived usefulness,

explaining up to 60% of the variance in this important driver of usage intentions

(Venkatesh & Davis, 2012). The TAM2 extends TAM by showing that subjective norm

exerts a significant direct effect on usage intentions over and above perceived usefulness

and perceived ease of use for mandatory (but not voluntary) systems (Venkatesh &

Davis, 2012). The effects of social influence processes were consistent with TAM2

(Venkatesh & Davis, 2000).

Subjective norm significantly influenced perceived usefulness via both

internalization, in which people incorporate social influences into their own usefulness

perceptions, and identification, in which people use a system to gain status and influence

within the work group and thereby improve their job performance (Venkatesh & Davis,

2000). Beyond these two indirect effects via perceived usefulness, subjective norm had a

direct effect on intentions for mandatory, but not voluntary usage contexts (Venkatesh &

Davis, 2000). The authors suggested that this may explain previous research that

indicated a non-significant role for social and that as individuals gained direct experience

with a system over time, they relied less on social information in forming perceived

usefulness and intention, but continued to judge a system’s usefulness on the basis of

potential status benefits resulting from use (Venkatesh & Davis, 2000).

The effects of cognitive instrumental processes were also consistent with TAM2

(Venkatesh & Davis, 2000). An important and interesting finding that emerged will be

the interactive effect between job relevance and output quality in determining perceived

usefulness. This implies that judgments about a system’s usefulness are affected by an

individual’s cognitive matching of their job goals with the consequences of system use

(job relevance), and that output quality takes on greater importance in proportion to a

system’s job relevance (Venkatesh & Davis, 2000). There are several practical

implications of the findings (Venkatesh & Davis, 2000). Mandatory, compliance-based

approaches to introducing new systems appear to be less effective over time than the use

of social influence to target positive changes in perceived usefulness (Venkatesh &

Davis, 2000). Practical alternatives to usage mandates based on social information

should be developed and tested, such as increasing the source credibility of social

information to increase internalization or designing communication campaigns that raise

the prestige associated with system use to increase identification (Venkatesh & Davis,

2000).

On the instrumental side, in designing systems to better match job relevant needs,

improving the quality of their output, or making them easier to use, research shows that

practical interventions for increasing result demonstrability, such as empirically

demonstrating to users the comparative effectiveness of a new system, may provide

important leverage for increasing user acceptance (Venkatesh & Davis, 2000). The

current research represents an important contribution to theory by extending TAM to

Advertisementress causal antecedents of one of its two belief constructs, perceived

usefulness (Venkatesh & Davis, 2000). Venkatesh & Davis (2000) noted that they have

begun to model the determinants of TAM’s other major belief, perceived ease of use

(Davis, 1996). Further research on TAM, in refining the models of the determinants of

perceived usefulness and perceived ease of use, should include the role of other direct

determinants of usage intentions and behavior and continue to map the major contingency

factors moderating the effects of perceived usefulness, perceived ease of use, subjective

norm, and other constructs on intention (Venkatesh & Davis, 2012). Another important

avenue for future research concerns the temporal dynamics of the determinants of user

acceptance.

Researchers found that TAM2 held up well at three points of measurement

spanning from pre-implementation to 3 months post-implementation (Venkatesh &

Davis, 2000). Perceived ease of use, in contrast, will be less stable over time, which is

attributed to the known role of direct hands-on experience in forming this belief

(Venkatesh & Davis, 2000). Future researchers might profitably seek to establish how

early in a system development process, for example, even before a working prototype is

built, one can measure key user reactions, such as perceived usefulness and intention, and

still rely on them as indicators of post-implementation success of the system concept

(Venkatesh & Davis, 2000). Research is needed to elucidate the processes involved in

cognitively matching important job goals to the consequences of system use (Venkatesh

& Davis, 2000). Understanding the process is a vital research direction since it explicitly

links the functional design characteristics of a system to perceived usefulness and

ultimately user acceptance (Venkatesh & Davis, 2012).

Future researchers should also seek to further extend models of technology

acceptance to encompass other important theoretical constructs, such as the choice sets of

available alternative technologies, learning and training, misperceptions of usefulness or

ease of use and how to correct them, changes in work content or job goals, and changing

social environments (Venkatesh & Davis, 2000). Future research should involve

comparison of past findings regarding the role of subjective norm with other more recent

findings (Venkatesh & Davis, 2000). Of particular interest would be a comparison of

TAM2 with recent approaches from the media choice paradigm that attempt to integrate

normative and utilitarian determinants. The continuing trend in organizations away from

hierarchical, command-and-control structures toward networks of empowered,

autonomous teams underscores the finding regarding the limits of organizational mandate

as a lever for increasing usage (Venkatesh & Davis, 2000). To adapt user acceptance

theory to this trend, the conceptualization of perceived usefulness will need to be

expanded from its current focus on expected individual-level performance gains to

encompass team-based structures and incentives (Venkatesh & Davis, 2000).

As the adoption decision becomes more of a team rather than an individual-level

decision, the nature and role of social influence processes (both within teams and across

teams) will need to be elaborated beyond TAM2 (Venkatesh & Davis, 2012). User

acceptance of IT in the workplace remains a complex, elusive, yet extremely important

phenomenon and the development and test of TAM2 advances theory and research on

this important issue (Venkatesh & Davis, 2000).

Technology Acceptance Model (TAM3)

The second of these variations is called the TAM3 (Venkatesh & Bala, 2008). As

with the original TAM and TAM2, TAM3 does not consider AIASGT as a determining

factor for perceived usefulness or acceptance of technology. Combining TAM2

(Venkatesh & Davis, 2000) and the model of the determinants of perceived ease of use

(Venkatesh, 2000), led to the development of an integrated model of technology

acceptance known as TAM3 (Venkatesh & Bala, 2008). TAM3 presents a complete

nomological network of the determinants of individuals’ IT adoption and use (Venkatesh

& Bala, 2008). Three theoretical extensions beyond TAM2 and the model of the

determinants of perceived ease of use were posited.

Prior research has provided valuable insights into how and why employees make

a decision about the adoption and use of (IT) in the workplace (Venkatesh & Bala, 2008).

From an organizational point of view, Venkatesh and Bala (2008) noted that the more

important issue is how managers make informed decisions about interventions that can

lead to greater acceptance and effective utilization of IT. The other suggestion is that

there is limited research in the IT implementation literature that deals with the role of

interventions to aid such managerial decision making (Venkatesh & Bala, 2008).

Particularly, there is a need to understand how various interventions can influence the

known determinants of IT adoption and use (Venkatesh & Bala, 2008). To

Advertisementress this gap in the literature, the vast body of research on TAM,

particularly the work on the determinants of perceived usefulness and perceived ease of

use, needs to be drawn upon in order to develop a comprehensive network (integrated

model) of the determinants of individual level (IT) adoption and use, empirically test the

integrated model, and present a research agenda focused on potential pre- and post-

implementation interventions that can enhance employees’ adoption and use of IT

(Venkatesh & Bala, 2008). The authors suggested that their findings and research agenda

have important implications for managerial decision making on IT implementation in

organizations.

The expectation is that the general pattern of relationships suggested by

Venkatesh and Davis (2000) and Venkatesh (2000) holds in TAM3 (Venkatesh & Bala,

2008). The determinants of perceived usefulness are believed to not influence perceived

ease of use and the determinants of perceived ease of use will not influence perceived

usefulness (Venkatesh & Bala, 2008). Furthermore, TAM3 does not indicate the

presence of any cross-over effects (Venkatesh & Bala, 2008). As noted earlier, two

theoretical processes explain the relationships between perceived usefulness and its

determinants: social influence and cognitive instrumental processes (Venkatesh & Bala,

2008). The effects of the various factors—that is, subjective norm, image, job relevance,

output quality, and result demonstrability—on perceived usefulness are tied to these two

processes (Venkatesh & Bala, 2008). However, there is no theoretical and empirical

basis to expect that these processes will play any role in forming judgments about

perceived ease of use (Venkatesh & Bala, 2008). Perceived ease of use has been

theorized to be closely associated with individuals’ self-efficacy beliefs and procedural

knowledge, which requires hands-on experience and execution of skills (Venkatesh &

Bala, 2008).

Venkatesh (2000) suggested that individuals form perceived ease of use about a

specific system by anchoring their perceptions to the different general computer beliefs

and later adjusting their perceptions of ease of use based on hands-on experience with the

specific system. Social influence processes (i.e., compliance, identification, and

internalization) in the context of IT adoption and use represent how important referents

believe about the instrumental benefits of using a system (Venkatesh & Davis, 2000).

Even if an individual gets information from important referents about how easy a system

is to use, it is unlikely that the individual will form stable perceptions of ease of use based

on the beliefs of referent others over and above his or her own general computer beliefs

and hands-on experience with the system (Venkatesh & Bala, 2008). The determinants

of perceived ease of use represent several traits and emotions, such as computer self-

efficacy, computer playfulness, and computer anxiety (Venkatesh & Bala, 2008). There

are no theoretical and empirical reasons to believe that these stable computer-related

traits and emotions will be affected by social influence or cognitive influence processes

(Venkatesh & Bala, 2008). The determinants of perceived ease of use are stated to not

believe to influence perceived usefulness (Venkatesh & Bala, 2008).

The determinants of perceived ease of use suggested by Venkatesh (2000) are

primarily individual differences variables and general beliefs about computers and

computer use (Venkatesh & Bala, 2008). These variables are grouped into three

categories: (a) control beliefs, (b) intrinsic motivation, and (c) emotion. Perceived

usefulness is an instrumental belief that is conceptually similar to extrinsic motivation

and is cognition (as opposed to emotion) regarding the benefits of using a system

(Venkatesh & Bala, 2008). The perceptions of control (over a system), enjoyment or

playfulness related to a system, and anxiety regarding the ability to use a system do not

provide a basis for forming perceptions of instrumental benefits of using a system

(Venkatesh & Bala, 2008). Therefore, control over using a system does not guarantee

that the system will enhance one’s job performance (Venkatesh & Bala, 2008).

Similarly, higher levels of computer playfulness or enjoyment from using a system do not

mean that the system will help an individual to become more effective (Venkatesh &

Bala, 2008). However, the expectation is that the determinants of perceived ease of use

will not influence perceived usefulness (Venkatesh & Bala, 2008).

Venkatesh and Bala (2008) posited that TAM3 has three relationships that were

not empirically tested by Venkatesh (2000) and Venkatesh and Davis (2000). The

recommendation is that experience will moderate the relationships between perceived

ease of use and perceived usefulness, computer anxiety and perceived ease of use, and

perceived ease of use and behavioral intention (Venkatesh & Bala, 2008). Perceived ease

of use to perceived usefulness is thought to be or can be moderated by experience

(Venkatesh & Bala, 2008). The concept is that with increasing hands-on experience with

a system, a user will have more information on how easy or difficult the system is to use

(Venkatesh & Bala, 2008). While perceived ease of use may not be as important in

forming behavioral intention in a later period of system use (Venkatesh et al., 2003),

users will still value perceived ease of use in forming perceptions about usefulness

(Venkatesh & Bala, 2008).

The premise for the afore mentioned belief is action identification theory

(Vallacher & Kaufman, 1996); wherein it is hypothesized that there is a clear distinction

between high-level and low-level action identities (Venkatesh & Bala, 2008). High-level

identities are related to individuals’ goals and plans, whereas low-level identities refer to

the means to achieve these goals and plans (Venkatesh & Bala, 2008). For example

regarding the context of a word processing software use, a high-level identity can be

writing a high quality report and a low-level identity can be striking keys or use of a

specific feature of the software (Venkatesh & Bala, 2008). Venkatesh and Bala (2008)

proffered that perceived usefulness and perceived ease of use are considered high-level

and low-level identities respectively. With increasing experience, the influence of

perceived ease of use (a low-level identity) on perceived usefulness (a high-level identity)

will be stronger as users will be able to form an assessment of their likelihood of attaining

high-level goals (i.e., perceived usefulness) based on information gained from experience

of the low-level actions (i.e., perceived ease of use; Venkatesh & Bala, 2008).

Computer anxiety to perceived ease of use would be moderated by experience

(Venkatesh & Bala, 2008). Experience will moderate the effect of computer anxiety on

perceived ease of use, such that with increasing experience, the effect of computer

anxiety on perceived ease of use will diminish (Venkatesh & Bala, 2008). The

expectation is that with increasing experience, system specific beliefs, rather than general

computer beliefs, will be stronger determinants of perceived ease of use of a system

(Venkatesh & Bala, 2008). Venkatesh (2000) argued that system-specific objective

usability and perceived enjoyment will be stronger determinants over time and the effects

of general computer beliefs (e.g., computer anxiety) will diminish because with

increasing experience as users will develop accurate perceptions of effort required to

complete specific tasks (i.e., objective usability) and discover aspects of a system that

leads to enjoyment (or lack thereof; Venkatesh & Bala, 2008). However, computer

anxiety (Venkatesh, 2000) is theorized by Venkatesh and Bala (2008) as an anchoring

belief that inhibits forming a positive perception of ease of use of a system. The effect of

computer anxiety on perceived ease of use will be argued by Venkatesh and Bala to

decline with increasing experience as individuals will have more accurate perceptions of

the effort needed to use a system and perceived ease of use to behavioral intention will be

moderated by experience.

Venkatesh and Bala (2008) posed the expectation that experience will moderate

the effect of perceived ease of use on behavioral intention, such that the effect will be

weaker with increasing experience. Perceived ease of use—that is, how easy or difficult

a system is to use—is an initial hurdle for individuals while using a system (Venkatesh,

2000). Once individuals become accustomed to the system and gain hands-on experience

with the system, the effect of perceived ease of use on behavioral intention will be

observed to recede into the background as individuals possessed more procedural

knowledge about how to use the system (Venkatesh & Bala, 2008). Consequently,

individuals place less importance on perceived ease of use while forming their behavioral

intentions to use the system (Venkatesh & Bala, 2008).

The findings of Venkatesh and Bala (2008) regarding perceived usefulness were

generally consistent with the legacy TAM2 Venkatesh and Davis (2000) results. In

particular, perceived ease of use, subjective norm, image, and result demonstrability were

significant predictors of perceived usefulness at all time periods (see Table 5). Also

consistent with Venkatesh and Davis (2000), job relevance and output quality had an

interactive effect on perceived usefulness; such that with increasing output quality, the

effect of job relevance on perceived usefulness will be stronger (Venkatesh & Bala,

2008). Experience will be found to moderate the effects of subjective norm on perceived

usefulness, such that the effect will be weaker with increasing experience (Venkatesh &

Bala, 2008). Also, the effect of image on subjective norm will be significant at all points

of measurements (Venkatesh & Bala, 2008). Therefore, it will be posited that with

TAM3, the effect of perceived ease of use on perceived usefulness will be moderated by

experience and the determinants of perceived ease of use (i.e., computer self-efficacy,

perceptions of external control, computer anxiety, computer playfulness, perceived

enjoyment, and objective usability) will not have any significant effects on perceived

usefulness over and above the determinants of perceived usefulness (Venkatesh & Bala,

2008). In short, experience moderated the effect of perceived ease of use on perceived

usefulness, such that with increasing experience the effect became stronger and the

determinants of perceived ease of use had significant effects on perceived usefulness at

any point in time (Venkatesh & Bala, 2008).

Overall, TAM3 could be used to explain between 52% and 67% of the variance in

perceived usefulness across different time periods and models (Venkatesh & Bala, 2008).

Consistent with Venkatesh (2000), the anchors, that is, computer self-efficacy,

perceptions of external control, computer anxiety, and computer playfulness, were

significant predictors of perceived ease of use at all points of measurement (Venkatesh &

Bala, 2008). Experience moderated the effect of computer anxiety on perceived ease of

use such that the effect became weaker with increasing experience and none of the

determinants of perceived usefulness had a significant effect on perceived ease of use.

Overall, TAM3 could be used to explain between 43% and 52% of variance in perceived

ease of use across different points of measurements and models (Venkatesh & Bala,

2008).

The theoretical contributions of TAM3 were to present a complete network of the

determinants of IT adoption and use. The key strength of TAM3 is its

comprehensiveness and potential for actionable guidance (Venkatesh & Bala, 2008).

While TAM presented a parsimonious model, the follow-up research on the general

determinants of perceived usefulness and perceived ease of use indicated pointers to

constructs that could be levers (Venkatesh & Bala, 2008). TAM3 is believed to

Advertisement richness and insight to the understanding of user reactions to new IT in the

workplace. The comprehensiveness ensures whether all relevant factors are included in a

theory and parsimony dictates whether some factors should be deleted because they

Advertisement little value to our understanding of a phenomenon. Therefore, the

comprehensiveness of TAM3 is important as the move is more toward a research agenda

related to various interventions (Venkatesh & Bala, 2008).

The emphasis of the TAM3 is on the unique role and processes related to

perceived usefulness and perceived ease of use and is based on the theory that the

determinants of perceived usefulness will not influence perceived ease of use and vice

versa (Venkatesh & Bala, 2008). This is an important theoretical contribution by itself

because there have been many inconclusive findings regarding the relationships among

some of these determinants, specifically perceived usefulness and perceived ease of use

(Venkatesh & Bala, 2008). Venkatesh (2000) found that perceived ease of use fully

mediates the effect of computer self-efficacy on behavioral intention (Venkatesh & Bala,

2008). The theoretical justification and empirical support of why the determinants of

perceived ease of use (e.g., computer self-efficacy) will not have significant effects on

perceived usefulness over and above the known determinants of perceived usefulness that

are driven by the social influence and cognitive instrumental processes were provided by

Venkatesh and Bala (2008). An example of this is while self-efficacy may have a weak

influence on perceived usefulness; this influence will become nonsignificant in the

presence of other important social and cognitive constructs (Venkatesh & Bala, 2008).

TAM3 is posited to promote new theoretical relationships, such as the moderating effects

of experience on key relationships, ergo it is important to understand the role of

experience in IT adoption and use contexts (Venkatesh & Bala, 2008).

With increasing experience, the effect of perceived ease of use on behavioral

intention will diminish while the effect of perceived ease of use on perceived usefulness

will increase (Venkatesh & Bala, 2008). This clearly indicates that perceived ease of use

is still an important user reaction toward IT even if users have substantial hands-on

experience with IT (Venkatesh & Bala, 2008). This important theoretical relationship has

significant practical utility as there has been increasing concerns about the ease of use of

various IT, particularly enterprise systems that are inherently complex to understand and

use (Venkatesh & Bala, 2008). There have been numerous cases of enterprise system

failures because of user resistance (Venkatesh & Bala, 2008). In many cases, the users

stopped using an enterprise system as they saw no benefits of using the new system; it is

possible that a lack of perceived ease of use contributed to unfavorable perceptions of

perceived usefulness in the context of those systems (Venkatesh & Bala, 2008). The

most important theoretical contribution is the delineation of relationships among the

suggested interventions and the determinants of perceived usefulness and perceived ease

of use (Venkatesh & Bala, 2008). As a note, though, these perceptions did not include

consideration of fully autonomous technology or the use of it for military purposes.

While prior research has shown important relationships between interventions and

key IT adoption determinants, extension of this research can be attained by providing a

comprehensive list of interventions, suggesting potential relationships of these

interventions with the determinants of perceived usefulness and perceived ease of use,

and offering important future research directions (Venkatesh & Bala, 2008). Unless

organizations can develop effective interventions to enhance IT adoption and use, there is

no practical utility of the rich understanding of IT adoption (Venkatesh & Bala, 2008).

The importance of interventions in enhancing IT adoption will be underscored by

Venkatesh (1999) who provided an example of how different modes of training can be

used to manipulate system-specific enjoyment, which enhances the salience of perceived

ease of use of a system as a determinant of behavioral intention (Venkatesh & Bala,

2008). At this juncture, however, there is little or no scientific research aimed at

identifying and linking interventions with specific determinants of IT adoption. The

theoretical arguments about the relationships between the interventions and the

determinants of IT adoption are thus an important contribution that could direct future

research (Venkatesh & Bala, 2008).

The implications for decision making indicate that the findings and research

agenda focusing on interventions have direct implications for two types of decision

making in organizations: employees’ IT adoption decisions and managerial decisions

about managing the IT implementation process (Venkatesh & Bala, 2008). Given that IT

is becoming increasingly complex and pertinent to employees’ decision making and work

processes, this research has implications for broad IT-enabled organizational decision

making (e.g., collaborative forecasting, inventory management, replenishment, service

delivery; Venkatesh & Bala, 2008). The authors of TAM3 suggested that interventions

primarily focus on the complex IT to understand how pre- and post-implementation

interventions can help employees make better adoption decisions about these complex

systems and managers make effective implementation decisions (Venkatesh & Bala,

2008). This is consistent with Venkatesh (2006) who argued that in order to be relevant

to organizational decision-making processes, individual-level IT adoption research

should focus on phenomena that are pertinent to decision making (knowledge sharing,

business process outsourcing) and IT that is critical for organizational decision making

(enterprise resource planning, supply chain management, collaborative forecasting,

inventory management systems; see also Venkatesh & Bala, 2008).

The interventions and future research agenda discussed here have implications for

these types of phenomena and systems (Venkatesh & Bala, 2008). Due to the complexity

of IT, it is increasingly difficult for employees to make effective decisions about

adoption, utilization, and coping with new IT (Venkatesh & Bala, 2008). As discussed

earlier, implementation of complex IT and associated changes in business processes have

a profound impact on employees’ job and cause changes in their job characteristics,

relationships with others in the workplace, and other aspects of their job (Venkatesh &

Bala, 2008). Consequently, employees’ job outcomes, such as job satisfaction and

performance, can be affected (Venkatesh & Bala, 2008).

Other types of reactions, such as avoidance, sabotage, workarounds, and shortcuts

are also prevalent. Interventions that can help employees make appropriate decisions

about adopting and utilizing a new IT are presented below (Venkatesh & Bala, 2008). In

the context of enterprise systems, certain design characteristics, such as extent of

customization or complexity of the system, can reduce changes in employees’ jobs as

these characteristics can potentially enhance the fit between a system and employees’

jobs (Venkatesh & Bala, 2008). Beaudry and Pinsonneault (2005) suggested that some

other interventions, such as user participation and training, can help employees decide

how to cope with or adapt to new IT (Venkatesh & Bala, 2008). Venkatesh (2006) called

for work on employees’ reactions to business process changes and process standards

implementation (see also Venkatesh & Bala, 2008). Interventions can help organizations

generate favorable individuals’ reactions toward business process changes and process

standards implementation (Venkatesh & Bala, 2008).

The findings and discussion of interventions can support managerial decision

making in two ways (Venkatesh & Bala, 2008). First, managers will now have a

framework to decide what interventions to apply during pre- and post-implementation

stages and for what types of system, such as for a complex system, perhaps interventions

that will create favorable ease of use perceptions will be relevant design characteristics,

user participation, training, and peer support; for a voluntary system, interventions that

will influence the determinants of perceived usefulness will be important to implement

design characteristics, user participation, incentive alignment, training, organizational and

peer support; and for interorganizational systems that affect organizational business

processes or a customer relationship management system that is critical to service

delivery (Froehle, 2006). Interventions, such as user participation, peer support, and

management support, will be particularly relevant (Venkatesh & Bala, 2008). The

question of ethics and culpability for autonomous platforms is a major topic of debate and

research (Arkin, 2011; Clarke, 2011; Cook, 2013; Davis, 1989; Geis, 2011; Kroeker,

2011; Mackey, 2013; Singer, 2013, Sparrow, 2012). Managers can decide on resource

allocation for interventions based on the impact of interventions on different determinants

of IT adoption and type of systems (Venkatesh & Bala, 2008). For example, if design

characteristics cannot be changed in a system, managers can allocate more resources to

training and user participation to make employees familiar with the systems (Venkatesh

& Bala, 2008).

The implementation of interventions is, of course, not a “silver bullet” for greater

IT adoption and effective utilization as implementation of interventions can increase

system development costs substantially (Venkatesh & Bala, 2008). Managers have to be

mindful in their decisions about implementing interventions and identify specific

interventions that can serve as levers for managers. Based on a comprehensive network

of IT adoption and use, a set of pre- and post-implementation interventions were

presented that tare recommended as the object of future scientific inquiry (Venkatesh &

Bala, 2008).

Unified Theory of Acceptance and Use of Technology (UTAUT) / TAM4

The most recent update of the TAM theory, which can be considered as TAM4,

will be classified by the authors as the unified theory of acceptance and use of technology

(UTAUT; Venkatesh, Morris, Davis, & Davis, 2003). UTAUT is applied to propose that

performance expectancy, effort expectancy, and social influence predict behavioral

intention towards the acceptance of IT (Venkatesh et al., 2003). The theory is further

based upon the concept that facilitating conditions and behavioral intention predicts use

behavior in the acceptance of IT (Taiwo & Downe, 2013). Ever since its inception, the

theory has been assessed using different applications and has become a dè factor model

of measuring user acceptance (Taiwo & Downe, 2013). Like all other models that

preceded it, UTAUT does not include consideration of the impact that AI has on

technology as far as its acceptance and perceived use. In terms of statistical significant

magnitude and direction, reports on the UTAUT model are diverse (Taiwo & Downe,

2013). UTAUT is based on a meta-analysis of 37 selected empirical studies in an effort

to harmonize the empirical evidence (Taiwo & Downe, 2013).

The outcome of the study indicated that only the relationship between

performance expectancy and behavioral intention is strong, while the relationships

between effort expectation, social influence, and behavioral intention are weak as well as

the relationship between facilitating condition, behavior, intention, and use behavior will

be also weak (Taiwo & Downe, 2013). The significance of the relationship between

facilitating condition and use behavior did not pass the fail-safe test, while the

significance of the relationship between behavioral intention and use behavior also did

not pass the fail-safe test satisfactorily (Taiwo & Downe, 2013).

UTAUT is a unified model that will be developed based on social cognitive

theory with a combination of eight prominent information acceptance research models

(Taiwo & Downe, 2013). The predictive validity of eight models in determining the

behavioral intention and usage to allow fair comparison of the models will be examined.

The eight models are (a) TRA, (b) the theory of planned behavior (TPB), (c) TAM, (d)

the motivational model, (e) a model combining TAM and TPB (C-TAM-TPB), (f) the

model of PC utilization, (g) innovation diffusion theory, and (h) socio cognitive theory.

The unified model outperformed the eight individual models. The UTAUT model

consists of four core determinants of usage and intention (performance expectancy, effort

expectancy, social influence, and facilitating conditions) alongside four moderators

(gender, age, experience and voluntariness of use) of key relationships (Taiwo & Downe,

2013). Through the meta-analysis, the researchers succeeded in combining and

investigating existing empirical literatures on UTAUT. The relationship between

UTAUT will be examined by Taiwo and Downe (2013) using a large sample size of over

11,000 and on the basis of the meta analyses reported, generally the findings confirm

Venkatesh et al.’s (2003) findings.

Initial findings among the five constructs of UTAUT indicated only some

relationships were strong, while others were slightly weak, but significant (Taiwo &

Downe, 2013). In conclusion, the majority of researchers cited UTAUT to support an

argument rather than using it, while others only partially used UTAUT and only a few

noted actual use of the theory (Taiwo & Downe, 2013). The findings of the study of

UTAUT contributes to the area of IT adoption and diffusion research by showing the

inadequacy and inconsistency in the use and output of a theory (Taiwo & Downe, 2013).

Similarly, there is no current permutation of TAM wherein the use of AIASGT is

considered.

This literature review contains a presentation of the dominant themes found in the

current research of the study problem. Researchers have tested micro level impacts of AI

and the field of AI is continually expanding, therefore there is a need to consider the

overall impact of AI (Arkin, 2011; Epitropakis, 2012; Geis, 2011; Mackey, 2013; Singer,

2013). The use of AI by the military is expanding because of increased productivity,

decreased operating costs, and the mitigation of friendly force casualties (Arkin, 2011;

Epitropakis, 2012; Geis, 2011; Mackey, 2013; Singer 2013). Although inquiries have

been made to test and obtain specifics from the military on their involvement with AI

(Arkin, 2011; Epitropakis, 2012; Geis, 2011; Mackey, 2013; Singer 2013), disclosures of

AI and how it is used are made reluctantly (Arkin, 2011; Geis, 2011; Mackey, 2013).

Self-Governance and Autonomous Technology

The military is at the forefront of AI development (Geis, 2011). Major

development and testing of AIASGT has been primarily by the military and the

perceptions of the users of AIASGT will possibly significantly Advertisement to the

understanding of how AIASGT will enhance or impede the acceptance of technology.

Researchers have suggested that the use of fully autonomous artificially intelligent

systems by the military to conduct combat operations will give commanders a resource to

utilize as well as mitigate friendly-force casualties (Arkin, 2011; Epitropakis, 2012; Geis,

2011; Mackey, 2013; Singer 2013). There are continual and escalating ethical, legal, and

societal challenges associated with the continued and expanding utilization of AI (Singer,

2013). Autonomy in machines employed to fight future wars have to have tested and

proven limitations programmed in them to limit their actions or the ramifications may

cause atrocities (Arkin, 2011; Epitropakis, 2012). Autonomous terminator-style

machines can be capable of human destruction even if proper precautions are met (Arkin,

2011; Epitropakis, 2012). Sleeman’s (2008) research indicated the technology that is

developed by the military will eventually transcend into peaceful commercial

applications.

Arkin’s (2011) research findings showed that the major development of AI has

been primarily by the military, and that ethical controls of autonomous robotic platforms

that were utilized on the battlefield need to be fully tested and theorized. The perceptions

of the use and acceptance of technology that includes the use of AIASGT will

significantly assist in the understanding of how AIASGT will enhance or impede the

acceptance of technology. The opinion is that effective artificially-intelligent control

must be developed to place limitations on AIASGT deployed by the military (Arkin,

2011).

Singer (2013) researched the need for ethical controls of AI used by the military

and elaborated that these machines with AI capabilities have to have limitations that will

allow the regulation and constraint of lethal actions. Similarly, Arkin (2011) noted the

need to bring self-governing machines under regulation and have limitations delineated in

the military rules of engagement that govern conduct on the battlefield. Both Arkin and

Singer suggested the need to ensure that machines deployed on the battlefield do not

conduct illegal and unethical actions. Arkin (2011) opined that an unbiased assessment

that delineates the current use of AI is needed.

An unbiased assessment would possibly give military planners insight into the

dynamics associated with the utilization of AI that is capable of making and changing

decisions independently that can cause the loss of human lives (Arkin, 2011; Singer,

2013). Epitropakis’ (2012) research findings paralleled Arkin’s (2011) research.

Epitropakis (2012) noted there will be a challenge with the collection of research

data from the secretive military environment; however, collection of research information

has become possible due to the declassification of many forms of AI (Cook, 2013). This

declassification has led to researchers being capable of gathering unclassified input from

military personnel employed by the organizations that will deploy the systems on the

battlefield (Arkin, 2011; Epitropakis, 2012). One example of the availability of

information is knowledge of the current use of AI expansion into many new, advanced

AI, such as ambient intelligence (Cook, 2013).

AmI has been identified as an emerging methodology that blends sensor

networks, pervasive computing, and AI (Cook, 2013). Advancements in technology,

such as AmI, is posited to transcend into commercial applications, and Singer (2013)

noted that fully autonomous aircraft and ground mobile platforms could conceivably

bring the commercial industry significant cost savings and efficiency. Cook (2013) also

noted that the use of technology, such as AmI, would eventually transcend AIASGT

utilized by the military. Cook predicted an increased utilization of technology that will

enhance AI’s flexibility and adaptability. Likewise, Arkin (2011) provided

representational and design recommendations for the implementation of an ethical control

and reasoning system potentially suitable for constraining lethal actions in an

autonomous robotic system. Arkin’s findings also indicated that an unbiased assessment

that delineates the current use of autonomous technology is needed to evaluate embedded

ethical controls if possible. Furthermore, Cook and Epitropakis (2012) offered that there

are benefits to autonomous technology based on consideration of the need to incorporate

ethical control and reasoning into developing technology.

The problem is as large as the military and the advancement of technology. The

military is at the forefront of the employment of this kind of technology on the battlefield.

The military is in a transitional phase that is primarily dependent on the acquisition and

employment of semi or fully autonomous platforms (Arkin, 2011; Clarke, 2011; Cook,

2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013, Sparrow,

2012). Geis (2011) postulated that the single greatest theme to emerge from technology

horizons is the need, opportunity, and potential to dramatically advance technologies that

can allow the Air Force to gain the capability increases, manpower efficiencies, and cost

reductions available through far greater use of autonomous systems in essentially all

aspects of Air Force operations.

The specificity of the problem is as complex as the permutations associated with

autonomous technology, the employment of the systems on the battlefield, and the very

platforms themselves. The use of these systems is constantly changing as the technology

matures. Autonomous systems can be integrated in all aspects of existing C2ISR

activities (Mackey, 2013). The problem associated with the control, quantification, and

implications associated with the use of these systems is expanding as their popularity

increases (Singer, 2013).

The requirement that failed to be met will be and will continue to be the accurate

quantification of the involvement that autonomous or semi-autonomous systems will

have on current and future military conflicts (Arkin, 2011). An exacerbating factor is the

secretive nature of systems development and platform capability (Arkin, 2011; Clarke,

2011; Cook, 2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013,

Sparrow, 2012). Autonomous systems capable of making independent decisions are

transforming the battle space and will experience expanded use by military planners

(Mackey, 2013). Some major concerns are if the military ever lost control of these

systems or if the systems themselves were utilized against the very countries that employ

them.

Advanced Autonomous Technologies

Major advancements in the development of autonomous technology have been

primarily by the military and researchers have identified unregulated advanced

autonomous technology is being developed by the military (Arkin, 2011; Epitropakis,

2012; Geis, 2011; Mackey, 2013; Singer, 2013). An example of advanced AIASGT will

be identified by Epitropakis (2012) known as Particle Swarm Optimization (PSO). PSO

will be utilized for the non-deterministic navigation of UAVs to allow UAVs to work

cooperatively toward the goal of protecting a wide area against airborne attack.

Advanced autonomous technology developed for military applications, such as PSO,

could transcend into commercial applications as it could be applied to benefit the

commercial industry in cost savings and efficiency as well as revolutionize the

transportation industry (Singer, 2013). However, the increased sophistication of

platforms that utilize advanced autonomous technology, such as the particle swarm

guidance systems, also call for a new look at the acceptance of this technology when it is

applied to fully autonomous, artificially intelligent platforms designed to take human

lives (Epitropakis, 2012). The reason technology like PSO will need to be further

researched is because of the increased efficiency the incorporation of AI will have in

creating human casualties (Epitropakis, 2012). Autonomous technology is posited to

create variables previously not considered or regulated (Arkin, 2011; Epitropakis, 2012).

Extant research shows that advanced AIASGT, such as UAVs that utilize PSO or

similar technology, will expand because of the efficiency, effectiveness, and increased

reliability (Geis, 2011). Mackey (2013) research concerned the current and envisioned

utilization of autonomous technology by the military and the findings were parallel with

the findings of Geis (2011) and Epitropakis (2012). The utilization of autonomous

technology for the military is a top priority (Geis, 2011; Epitropakis, 2012). Geis

authored the Blue Horizons research study for which the purpose will be to determine the

capabilities of advanced autonomous technology in which the Air Force would need to

invest to maintain its dominant air, space, and cyberspace capability in the year 2030.

Geis noted that specific recommendations from his research were to focus scarce research

dollars into future-system concepts and critical technology areas.

The expanded use of AIASGT is the primary goal of the USAF (Geis, 2011).

Adaptable and flexible advanced AIASGT could provide time-domain operational

advantages over adversaries, who are limited to human planning and decision making

(Geis, 2011). Similar research from Mackey (2013) and Singer (2013) indicated the

military’s use of advanced autonomous technology, which were designed to mitigate or

eliminate the human element out of the planning, preparation, and execution of an act,

will be utilized.

Research of advanced autonomous technology showed that the consideration of

such technologies originally surrounded the need for a machine to have the ability of

independent action, which is especially useful for mundane or routine activities,

particularly in terms of economic advantages in commercial applications (Geis, 2011;

Singer, 2013). The consideration of how advanced autonomous technologies are used

would change how technology would be accepted, especially when the major

advancements are covertly developed by the military (Arkin, 2011; Singer, 2013). Arkin

further accentuated the necessity of implementing an ethical control and reasoning

system into advanced autonomous technology. Ethical controls would have to be suitable

for constraining lethal actions in an AI powered robotic machines so that they fall within

the bounds prescribed by the laws of war and rules of engagement (Arkin, 2011).

Current research has not considered the permutations inherent in AI platforms that

are a reality in today’s battle space (Epitropakis, 2012; Singer, 2013). Epitropakis (2012)

identified the major breakthroughs and advancements of autonomous technologies are

designed to specifically terminate human lives on the battlefield. The postulation in

regards to the acceptance of technology specifically designed for battlefield applications

that autonomously make life or death decisions is the main theme of this study. A

recalcitrant resistant nature of military personnel to provide information when questioned

about their employment of autonomous technology against a human target has been

recognized (Singer, 2013). In correlation with Singer’s (2013) findings, Arkin (2011)

and Geis (2011) stated that the world will be fast approaching autonomous systems as a

predominant means of persecuting war, which is exacerbated by the recalcitrant resistant

nature of military commanders to provide information. In the event that the systems are

circumvented, the reality is that these very systems could be utilized against the

governments that employ them (Epitropakis, 2012; Singer, 2013).

There are many more complexities associated with the utilization of autonomous

technology for civilian applications, such as drug enforcement (Mackey, 2013).

Mackey’s (2013) research findings showed the utilization of autonomous technology is

far less restrictive when used on the battlefield. When systems are adapted for peaceful

commercial applications, the sophistication of the associated technologies will transcend

into even more deadly applications for the military, which are in correlation with

published research findings (Arkin, 2011; Mackey, 2013).

Technology is moving rapidly and causing unexpected changes; clandestine cyber

wars are a current reality and will accelerate as technology advances (Brock, 2010).

When autonomous technology is employed on the battlefield, it is vulnerable to cyber-

attacks (Brock, 2010; Geis, 2011). There is the need to embed safeguards against cyber-

attacks in machines that use autonomous technology because all major advancements in it

is due to microchip technology (Brock, 2010; Geis, 2011).

Researchers, such as Brock (2010), observed that Moore’s law in respect to the

subsequent miniaturization of the microchip may be obsolete due to the reality that

miniaturization will reach its apex. The apex of miniaturization is limited by the laws of

physics as they are currently understood. Advancements in technology, particularly

regarding AI that uses a microchip, will benefit with increased functionality giving

unknown capabilities to everything that uses advanced chipsets (Brock, 2010). In the

past, ethical controls and considerations had to be incorporated into the machine at the

time of development to be effective (Arkin, 2011). The uncertainties of the capabilities

of micro technology identified by Brock indicates that the ethical constraints identified by

Arkin (2011) and Epitropakis (2012) should be incorporated into autonomous machines.

The military continues to be a major developer and procurer of information

systems, yet very limited research has been done to determine the factors that influence

technology acceptance by military personnel (Levy & Green, 2010). Mackey (2013),

Arkin (2011), and Geis (2011) identified the military as the major developer of

information systems, including autonomous technology.

AIASGT and Autonomous Technologies

Major development of autonomous technology has been primarily by the military

and the use of TAM and its current relevance has been researched and considered as the

benchmark criterion when evaluating the acceptance of military technology (Ghazizadeh

et al., 2011). Original thoughts concerning acceptance of technology is parsimonious and

includes perceived usefulness, which has consistently been used to explain more variance

in technology acceptance than other variables (Ghazizadeh et al., 2011). Another

contribution to the acceptance of autonomous technology will be when Lee (2009)

expounded on the theory of planned behavior, perceived risk, and perceived benefit to

understand the acceptance of technology, such as AI (Holden & Karsh, 2010).

The development of technology and its insertion into how the medical industry

accepts new technology is developed with the principles incorporated in autonomous

technology (Ghazizadeh et al., 2011). The acceptance of autonomous technology has

been used to measure consumer acceptance of health informatics technologies, such as

m-health, e-health, and telemedicine research (Or et al., 2011). Consideration of

autonomous technology for the evaluation of the overall acceptance of technology is used

when postulating the benefits of medical technology and there is a push for the

application of AI in more consumer health informatics studies (Keselman, Logan, Smith,

Leroy, & Zeng-Treitler, 2010).

Ethical Considerations and Autonomous Technology

The ethical concerns inherent to the use of autonomous technology are directly

related to the necessary safety measures and controls that can or cannot be incorporated

into the design of AI deployed by the military along with the overall public acceptance of

the technology. The major advancements in the development of autonomous technology

have been achieved primarily by the military (Geis, 2011) and user perceptions of

AIASGT may correlate to literature to advance knowledge on the use and acceptance of

such technologies. There are ethical considerations necessary before the decision is made

to deploy military centric autonomous technology (Arkin, 2011). However, there are

moral dilemmas associated with autonomous technology that are designed to take human

lives (Arkin, 2011).

The problem concerning the ethics associated with the programming that is

imbedded in the technology as well as the ability of the technology to differentiate the

moral or ethical differences of any action is a problem that has been identified as needing

clarification (Geis; 2011). However, the needed technology that will embed ethics into

AI is not currently available (Arkin, 2011). The field of “roboethics” is a relatively new

academic concept, and will be described at a 2004 conference of philosophers,

sociologists, and scientists in Italy as being a machine that has the ability to both think

and act like a rational human possesses AI whether it is physically anthropomorphized or

not, and this necessitates ethical discussion about the treatment of these devices, for they

may possess cognitive (if not conscious) capacity beyond that of a simple machine

(Howlader, 2011, p. G:5). Arkin (2011) noted that the lack of technology that governs

ethical boundaries will not allow the needed ethical constraints to be imbedded into a

machine that is driven by AI. Imbedded ethical constraining technology will be a

consideration when evaluating the overall acceptance of autonomous technology used by

the military.

Howlader (2011) noted, “to create machines that can make their own choices, be

aware of their existence, and at the same time subordinate that free will to the benefit of

humanity is frankly unethical” (p. G:5). The dichotomy described by Howlader suggests

that ethics in relation to how AIASGT operate is a challenge. The U.S. Navy’s Office of

Naval Research produced a report on military robot ethics, which envisages a fast-

approaching era where robots are smart enough to make battlefield decisions that will

preserve human life (Epitropakis, 2012). Eventually, robots will come to display

significant cognitive advantages over homo sapien soldiers (Epitropakis, 2012).

There is a rush to the finish line that seems to be taken by the U.S. military in

production and fielding of AIASGT, especially with weapons, in the fields before other

countries (Lewis, 2013). Lewis (2013) noted that this rush may create challenges that

were not considered during the development and implementation of autonomous

technology. Furthermore, the deployment of AIASGT may create other problems that

could have never been anticipated. Because of budget reductions, all governmental

agencies are cash strapped and may be forced to cut corners (Lewis, 2013). Cost

mitigation is an ethical consideration when it forces engineers to field equipment that will

be not tested 100% (Arkin, 2011). Geis (2011) noted that military commander’s attitudes

were that as long as the equipment works, problems can be fixed as they are found. “we

put it into combat without knowing everything about it. But the Air Force needed to get

the Reaper capability into the field,” said Col. Chris Chambliss, the 432nd Wind and Air

Expeditionary Wing commander at Creech Air Base, Nevada (Geis, 2011, p. 8)

Ethical concerns are further exacerbated when admittedly; the developers of

autonomous technology noted that with all engineering challenges, there will be glitches

in the hardware or software (Drew, 2013). There have been configuration mishaps along

with equipment being fielded that is barely operable. The poor condition of the

equipment has been noted as a cause of damages to UAVs, properties, and civilian’s

lives. “The context will be to do just the absolute minimum needed to sustain the fight

now, and accept the risks, while making fixes as you go along,” said Colonel Mathewson

of the USAF said (Drew, 2013, p. 12). Robots are now being used for more than

surveillance to help troops on the ground. They are being used for more than providing

instant imagery back to the operators as these semiautonomous robots have now been

promoted to being used as “true hunter-killer” robots (Drew, 2013). This brings the

spectrum of using robots in warfare into a completely different realm of ethical issues.

The software development that makes AIASGT possible is a complex

undertaking. In software engineering, there are millions of lines of a programming code

that are written by a group of programmers. In the software development cycle, there

were many who created the base code for autonomous technology enabled machines,

producing ambiguity when attempting to track a single developer (Epitropakis, 2012). In

the software development environment, there is no one person who will know the

outcome of the entire program. Ethical concern can be generated because of the fact that

the possibility of faulty or malicious code could make its way into AIASGT software.

This faulty or malicious code could allow hostile or unauthorized access of the AI system

(Lewis, 2013). There is a misconception that autonomous technology will do only as

they are commanded or programmed, which has proven to be a serious problem

(Debusmann, 2009).

Under many of the autonomous technology guidelines designed to protect humans

from harm, the (potential) moral agency of autonomous technology is often ignored or

contradicted (Howlader, 2011). An AIASGT fully moral agent would be at least partially

self-aware, have some sense of self-preservation, and therefore should be able to save

itself from harm for any given circumstance. Under the current guidelines, self-

preservation on the part of autonomous technology is often subsumed within a duty to

protect humanity (Howlader, 2011). The motivation for self-preservation would compel

AIASGT to protect itself from harm. The ethical concern would be if the mechanisms

that will allow AIASGT to protect itself may cause harm to a human being. There are

multiple ethical considerations when it comes to AIASGT defending itself. Some ethical

questions that derive from AI defending itself against a human would be if the type of

action could be ethically (and/or legally) justifiable, or permissible, and what conditions

generate and sustain such ethical attitudes (Howlader, 2011).

Howlader (2011) asked if an autonomous technology machine could be

considered to have ethical or legal attributes and be held accountable for moral actions, or

if ethics are pre-programmed. The latter case would seem to indicate that AIASGT

cannot be fully moral agents. A fundamental problem with treating AIASGT, in human

forms or not as mere machines, arises from the nature of AI. AI has been described as

forms of nonorganic cognitive systems that can both think rationally and humanly as well

as act rationally and humanly (Howlander, 2011).

While in general, a human remains in the loop for decision making regarding the

deployment of lethal force, the trend is clear that targeting and engagement decisions are

being moved forward onto these machines as the science of autonomy progresses (Arkin,

2011). The dangers of abuse of unmanned robotic systems in war, such as the Predator

and Reaper, are well documented, which occurs even when a human operator is directly

in charge (Adams, 2010). With the advancements in autonomous technology, targeting

decisions will become increasingly autonomous with the intention of the military planner

of having them become fully autonomous (Mackey, 2013). Full autonomy is the goal for

military planners because of factors, such as the ever increasing tempo of the battlefield.

Military planners also found that inherently, AI machines limit casualties and offer

economy of resources by reducing the need for human personnel (Geis, 2011). These

achievements allow the Air Force to do more with less and to project air power with

increased efficiency (Geis, 2011).

A well delineated research hypothesis will be written that indicated the

autonomous systems could ultimately operate more humanely than human warfighters are

able (Arkin, 2011). As part of the research to test this thesis funded by the Army

Research Office, an ethical architecture for an autonomous system will be developed with

the intent to enforce the principles derived from the laws of war, thus having the goal of

enhancing noncombatant safety and survivability (Arkin, 2011). The author Both a

deontological (rights-based) perspective will be encoded within the laws of war (e.g., the

Geneva Conventions) and a utilitarian (consequentialist) perspective (e.g., for the

calculation of proportionality relative to military necessity as derived from the principle

of double effect) were considered (Arkin, 2011).

While the ethical architectural design is believed generalizable to other non-lethal

domains, such as personal robotics, nonetheless, there is the current implementation in

the context of military operations (Arkin, 2011). This implementation is a constraint

satisfaction problem for military obligations with inviolable constraints for ethical

prohibitions. Proportionality determination, the use of an appropriate level of force on a

specific target given military necessity, can be conducted by running, if needed, an

optimization procedure after permission is received over the space of possible responses

(from none, to weapon selection, firing pattern, aiming; Arkin, 2011). The

proportionality determination strives to minimize collateral damage when given

appropriate target discrimination certainty in a utilitarian manner, while the ethical

governor ensures that the fundamental deontological rights of civilian lives and property

are respected according to international conventions (Adams, 2010). If the potential

target remains below the certainty threshold and is thus ineligible for engagement, the

robotic system could invoke specific behavioral tactics to increase the certainty of

discrimination instead of “shooting first and asking questions later” (Arkin, 2011). See

Figure 1 for a depiction of this architectural process.

Figure 1. High Level Schematic of the Ethical Hybrid-Deliberative Reactive

Architecture. Adapted from R. Arkin, 2011, IEEE Society on Social Implications of

Technology, 30, p. 12.

AIASGT Moral, Ethical Adaptation, and Data Collection Integrity

In order for autonomous technology to display true ethical behavior, emotional

content has to be engineered into the programming (Haidt, 2010). Haidt (2010) provided

a list of parameters in what he described as a taxonomy of moral emotions that are

necessary to achieve true ethical constraints. The emotional content necessary to achieve

these goals is characterized as other-condemning (contempt, anger, disgust), self-

consciousness (shame, embarrassment, guilt), other-suffering (compassion), and other-

praising (gratitude, elevation). The concept of guilt has to be present whenever there is a

violation of ethics. Guilt has been noted as a critical motivator of ethical or moral

behavior. Guilt is usually caused by the violation of moral rules and regulations,

particularly if those violations caused harm or suffering to others (Haidt, 2010).

Guilt should only result from unintentional effects of the AI agent, but

nonetheless, its presence should alter the future behavior of the system so as to eliminate

or at least minimize the likelihood of recurrence of the actions that induced this affective

state (Arkin, 2011). Guilt is characterized by its specificity to a particular act and

involves the recognition that one’s actions are bad, but not that the agent itself is bad

(which instead involves the emotion of shame; Arkin, 2011). The value of guilt is that it

offers opportunities to improve one’s actions in the future (Haidt, 2010).

Various Developed Branches of Artificial Intelligence

Logistical AI has been accepted as a program that knows about the world in

general and the facts of the specific situation in which it must act, and its goals are all

represented by sentences of some mathematical logical language (McCarthy, 2012). The

program decides what to do by inferring that certain actions are appropriate for achieving

its goals. Pattern recognition has been incorporated into artificially intelligent

algorithms. AI programs often examine large numbers of possibilities, for example,

moves in a chess game or inferences by a theorem proving program and discoveries are

continually made about how to do this more efficiently in various domains. When a

program makes observations of some kind, it is often programmed to compare what it

sees with a pattern. An example that McCarthy (2012) gave to understand this is from a

vision program that may try to match a pattern of eyes and a nose in a scene in order to

find a face. The more complex patterns, like those found in a natural language text, a

chess position, or the history of some event are also studied. These more complex

patterns require quite different methods than do the simple patterns that have been

studied the most. The representations were facts about the world that are to be

represented in some way; usually languages of mathematical logic are used for

representation (McCarthy, 2012).

Inference AI is a technique utilized in developing AI, some from facts and others

can be inferred (McCarthy, 2012). Mathematical logical deduction is adequate for some

purposes, but new methods of non-monotonic inference have been used to logic since the

1970s. The simplest kind of on-monotonic reasoning is default reasoning in which a

conclusion is to be inferred by default, but the conclusion can be withdrawn if there is

evidence to the contrary. For example, when one hears a bird, the inference may be that

it can fly, but this conclusion can be reversed when one hears that it is a penguin. It is the

possibility that a conclusion may have to be withdrawn that constitutes the non-

monotonic character of the reasoning. Ordinary logical reasoning is monotonic in that

the set of conclusions that can the drawn from a set of premises is a monotonic increasing

function of the premises. However, circumscription is another form of non-monotonic

reasoning (McCarthy, 2012).

Common sense knowledge and reasoning is a foundation and strong basis for AI

development and is the area in which AI is farthest from the human-level, in spite of the

fact that it has been an active research area since the 1950s (McCarthy, 2012). There has

been considerable progress in terms of developing systems of non-monotonic reasoning

and theories of action, yet more new ideas are needed (McCarthy, 2012).

Learning from experience is one of the most important advances in the maturation

of AI (McCarthy, 2012). McCarthy (2012) suggested that artificially intelligent

programs can actually learn from experience as the approaches to AI are based on

connectionism and neural nets that specialize in learning. There is also learning of laws

expressed in logic; programs can only learn what facts or behaviors their formalisms can

represent, and unfortunately, learning systems are almost all based on very limited

abilities to represent information (McCarthy, 2012).

Planning is an aspect of AI that is considered a benchmark of its development

(McCarthy, 2012). Planning programs start with general facts about the world (especially

facts about the effects of actions), facts about the particular situation, and a statement of a

goal. From these, a strategy is generated for achieving the goal. In the most common

cases, the strategy is just a sequence of actions (McCarthy, 2012).

Epistemology is a study of the kinds of knowledge that are required for solving

problems in the world and is utilized in the development of AI (McCarthy, 2012).

Similar to epistemology, ontology is the study of the kinds of things that exist instead of

the kinds of knowledge. In AI, the programs and sentences deal with various kinds of

objects, and a study is conducted of what these kinds are and what their basic properties

are (McCarthy, 2012). Heuristics is experienced based technology for discovery, problem

solving, and learning that has been assimilated into AI development (McCarthy, 2012).

Heuristics is a way of trying to discover something or an idea imbedded in a program.

Heuristic functions are used in some approaches to search to measure how far a node in a

search tree seems to be from a goal. Heuristic leads to the need to compare two nodes in

a search tree to see if one is better than the other (i.e., constitutes an advance toward the

goal, and may be useful; McCarthy, 2012). Genetic programming is an algorithm based

machine learning methodology that emulates biological evolution (McCarthy, 2012).

McCarthy (2012) noted that genetic programming is a technique for getting programs to

solve a task by mating random list programs and selecting the fittest in millions of

generations.

Applications of Artificially Intelligent Technology

Game playing is identified as one of the most well-known applications that utilize

artificially intelligent technology (McCarthy, 2012). The development of the gaming

industry allowed developers to produce super sophisticated algorithms designed to

challenge the human mind. Machines that can play master level chess contain some AI in

them, but operate well against people mainly through brute force computation—looking

at hundreds of thousands of positions. However, for the program to beat a world

champion by brute force, known reliable heuristics require being able to look at 200

million positions per second. In the case of game playing, AI can run many algorithms

that rival the human mind’s capacity. The goal of the developers is to challenge the mind

in a gaming scenario. This quest could lead the developer to create a machine that could

out think a human (McCarthy, 2012).

The development of speech recognition technology is a major advancement in the

maturation of AI (McCarthy, 2012). Since speech is the primary means of

communication between people, scientists have worked on maturing the technology for

practical usage. There are many reasons that speech recognition technology has received

so much attention. The maturation of the technology can be traced to simple curiosity

about the mechanisms that are necessary to produce artificial human speech-like

capabilities in machines. The technology can be connected to the human desire to create

a machine that will automate tasks that require human interactions. Early in the

development of the computer, it will be clear that the man-machine interface could be

bridged quickly if the computer keyboard will be eliminated. Speech recognition by

machines will expedite the man-machine interface allowing for more efficient

transactions. Research in the field of automatic speech recognition is a major aspect of

the development of AI (McCarthy, 2012).

There were major advances in the development of speech recognition technology

in the 1990s (McCarthy, 2012). Computer speech recognition reached a practical level

for limited purposes such that United Airlines replaced its keyboard tree for flight

information by a system using speech recognition of flight numbers and city names.

While it is possible to instruct some computers using speech, most users have gone back

to the keyboard and the mouse due to it being more convenient. The challenges

associated with the development of a machine that can function and interact like a human

is still maturing. At present, human accomplishments are only at the beginning of this

new and rapidly maturing technology, but scientists predict that in the near future, the

sophistication of the machines that utilize this technology will rival that of a human.

Speech recognition technology is synonymous with AI and the acceptance of a

technology that can outthink a human is something that will require continuous research

that is parallel with its development. This parallel research could afford humanity the

opportunity to put in safeguards to eliminate or control there obsolescence (McCarthy,

2012).

Computer vision will be a major component of the use AI that will need to be

developed further (McCarthy, 2012). The world is composed of three-dimensional

objects, but the inputs to the human eye and computers’ TV cameras are two

dimensional. While some useful programs can work solely in two dimensions, full

computer vision requires partial three-dimensional information that is not just a set of

two-dimensional views. At present, there are only limited ways of representing three-

dimensional information directly, and they are not as good as what humans evidently use.

Today’s autonomous platforms are programmed to follow a predefined trajectory and this

methodology is only sufficient when the platform is operating in a fixed or predetermined

environment. A fixed or predetermined environment will have objects in known or

predetermined locations. In a semi-autonomous environment, there will be a human in

the loop that will make any needed alterations of the platforms trajectory (McCarthy,

2012).

Artificial vision is essential for fully autonomous platforms, especially when

machines must make decisions independent of human intervention (Howlander, 2011).

Artificial vision has to evolve to the point that the AI that is part of the system can

differentiate between objects observed. The technology will have to evolve to the point

the system can recognize all known objects. This is especially critical when these

systems are employed with platforms that are designed to seek and terminate a human life

(Arkin, 2011).

Fully autonomous systems capable of making independent decisions are being

utilized on the modern battlefield (Epitropakis, 2012). The need to consider the impact

that AI will have on humanity when used to take human lives needs to be theorized

(Epitropakis, 2012). Because of the covert utilization of AI by the military, a postulation

can be made that the topic has not been fully socialized. The development of AIASGT

should be considered in light of the component of ethical and societal acceptance.

The acceptance of fully autonomous technology capable of taking a human life

has not been fully vetted in the development of AIASGT. Autonomous platforms and

their ever expanding capabilities to make independent decisions will have immeasurable

impact on the way technology is used and accepted. The consideration of autonomous

self-governance will expand the original opinions of technological acceptance and

propagate theory to test the permutations that the future plethora of AI will have on the

acceptance of technology

Up to date, many researchers utilize new variables based on the acceptance of

technology (Arkin, 2011; Clarke, 2011; Cook, 2013; Davis, 1989; Geis, 2011; Kroeker,

2011; Mackey, 2013; Singer, 2013, Sparrow, 2012). Lee (2009) united the then current

criteria associated with the acceptance of technology and the theory of planned behavior,

perceived risk, and perceived benefit to understand the adoption of AIASGT in banking.

AIASGT Software Development

The continued development of autonomous technology would depend on

incremental advancements in both hardware and software development (Cook, 2013).

Any autonomous technology developed for Future Combat Systems (FCS) would be a

System of Systems (SOS) with a complex architecture and set of responsibilities

(Mackey, 2013). Software is the nucleus of autonomous technology and this software

will be in manned and unmanned platform systems, networked software, stand-alone

training systems, and interface with many complementary programs and other external

systems (Geis, 2011). AIASGT FCS platform systems (which are also known as prime

items) include the Manned Ground Vehicle (MGV) platforms, Unmanned Ground

Vehicle (UGV) platforms, Unattended Ground Sensors (UGS) and Unmanned Air

Vehicles (UAV) (Mackey, 2013).

One major concern about the software developed for autonomous technology is

that it is becoming increasingly sophisticated and vulnerable to the insertion of malicious

logic (Singer, 2013). System and subcontractor developers of AIASGT are responsible

for developing the manned and unmanned platforms, integrating their software

components into the platform, integrating networked software and platform unique

software, integrating Buyer Furnished Equipment (BFE) systems into their platform, and

for incrementally testing platform functions for errors and disconnects with BFE systems

and components. Prime item specifications specify capabilities and performance

requirements of the platform (which reference BFE systems and components), interface

requirements for the BFE systems and components, and verification requirements for the

platform including some integrated verification of the total AIASGT integrated platform.

Critics of AI software development identify vulnerabilities of AI software noting that

some components of the software cycle are developed between many companies’

(Mackey, 2013). This distributed development can allow the possibility of malicious

logic insertion into AI software (Arkin, 2011). The prime item specifications are

decomposed to specifications for platform software components, platform hardware

components, and a set of requirements for the AIASGT Distributed Software System

(DSS).

The AIASGT networked software consists of software components which will be

installed in all platforms and will be networked together to achieve a network centric

SOS. These components include Battle Command (BC), Intelligence, Surveillance, and

Reconnaissance (ISR), Network Management System (NMS), Training Common

Components (TCM), and logistics components. These components will be built upon the

AIASGT Systems of Systems Common Operating Environment (SOSCOE). The

collective set of networked software requirements include functional and performance

requirements, requirements to interface with the platforms, and verification requirements

including some AIASGT SOS integrated verification.

Any AIASGT software development will implement products for Modeling and

Simulation (M&S) (Geis, 2011). The requirements, selection and schedule for

implementing M&S assets will be driven by the needs of Spiral Outs, Experiments, and

core program Integrated Phases. This M&S framework is designed to achieve the

objective of implementing and maintaining a credible autonomous technology equipped

UA SOS representation that enables execution of concurrent development of systems

engineering and software requirements; production and/or acquisition of software and

hardware products; execution of a broad range of test, experimentation, analysis, training,

and operational applications; and definition and assessment of products to be

incrementally spun out to current forces (Arkin, 2011).

The AIASGT Command, Control, Communications, Computers, Intelligent,

Surveillance, and Reconnaissance (C4ISR) Segment Team is responsible for developing

the common software infrastructure, Battle Command, ISR, and Network Management

System software and for integrating that software into the C4ISR software package (Geis,

2011). The C4ISR team is responsible for integrating the C4ISR software and the rest of

the networked systems software into a networked software package. The AIASGT

networked software package will have the capabilities and capacities to support the

interoperability of different platform variants into one integrated, networked software

system. The C4ISR segment team will integrate the various C4ISR installations at the

FCS SOS level, and incrementally test the integrated variants to find errors and

disconnect with respect to C4ISR functions across the set of variants and in a SOS

operationally relevant set of platforms (Mackey, 2013).

AIASGT Systems of Systems Common Operating Environment (SOSCOE)

The SOSCOE concept will be developed in conjunction with the legacy US

Army’s FCS program. All autonomous technology will utilize some form of SOSCOE

which allows subsystem standardizes component-to-component communications between

software applications, computers, vehicles, virtual private networks, and communications

within the Global Information Grid (GIG), to enable interoperability with legacy, joint,

coalition, government, and non-government organizations (Geis, 2011). AIASGT

subsystem provides the capacity to process, synthesize, and transform enormous amounts

of information from multiple disparate sources and provides fault tolerance that maintains

information integrity in an unpredictable network environment. SOSCOE will isolate

applications from the underlying operation system to allow portability as operating

systems and platforms change. The SOSCOE subsystem provides the integration

framework for the Command, Control, Communications, Computers, Intelligence,

Surveillance and Reconnaissance (C4ISR) System. In turn, the C4ISR System is the

framework for integrating the overall Future Combat Systems (FCS) SOS. As such, the

SOSCOE subsystem is key to ensure integrated SOS functionality and performance. The

SOSCOE subsystem provides a stable framework for the FCS evolutionary-based

application development process, facilitating block upgrades over time.

FCS systems will be configured by integrating new and reusable hardware and

software components that are implemented by different suppliers. The SOSCOE is

configurable so that any specific instantiation only needs to incorporate the SOSCOE

components that are needed for the specific mission and capabilities for which the

instantiation is

AIASGT Test and Evaluation

AIASGT will have to be thoroughly tested and evaluated in order to build

confidence in system reliability before it is set loose to perform its intended function

(Arkin, 2011). All AIASGT development would have to follow a Software Test

Procedure that will be delineated in the Software Test Description (STD) which specifies

the delta qualification test preparations and procedures (Elgabalawi, 2012). In the case of

the US Army, the AIASGT Battle Command System (BCS) Computer Software

Configuration Item will be originally designed for the FCS program. The STD is

developed in accordance with the Software Test Plan (STP), Software Development Plan

(SDP) in accordance with a Software Test Description Standard (STDDS) (Elgabalawi,

2012).

AIASGT that are utilized by the military will be incorporated into what is known

as the BCS (Elgabalawi, 2012). The BCS is defined as a suite of IT and software

packages that will enhance the operational capability of the Modular Brigade Combat

Team (MBCT) by increasing its situational understanding and lethality. AIASGT will

enable units to better see first, understand first, act first and finish decisively. AIASGT

will be deployed in commander-type current force vehicles and Tactical Operations

Centers (TOC) from the Brigade down through the Battalion to the Company level. The

operator will interact with AIASGT using a Future Combat System (FCS) like generated

Graphical User Interface (GUI). AIASGT will be placed in all future Intelligent

Munitions Systems (IMS) and Unattended Ground Sensors (UGS) under development by

the US Army. AIASGT devices will be deployed as platoon and company assets to

gather situational awareness of particular areas of interest, and as a limited persistent

means of engagement with scalable effects (Mackey, 2013).

AIASGT Software Integration and Test (SWIT)

SWIT software integration activity that systematically exercises integrated

software items for the purpose of error finding and risk reduction for later testing. It

occurs at several levels including Computer Software Configuration Item (CSCI)

integration, integration of CSCIs within a platform, integration of CSCIs which are part

of a networked system and integration with complementary software systems. AIASGT

software qualification test activity that verifies that the software item performs to its

allocated requirements in the target computing system or an equivalent system. AIASGT

has to follow known software developmental life cycles. In following the best practices

of software development, the target computer within which the AIASGT software will

run, is usually located in an engineering Development Unit (EDU). If an EDU is not

available, what is known as the brass board (a Commercial Off-The–Shelf [COTS] item),

which closely emulates the EDU, is usually an acceptable substitute.

The processes comprising the AIASGT applications (components) along with key

system software processes are usually allocated amongst the Integrated Computer System

(ICS) processors. For military applications, AIASGT software includes Command and

Control (C2), Intelligence, Surveillance and Reconnaissance (ISR), Network

Management Systems (NMS), or System of Systems Common Operating Environment

(SOSCOE). The incorporation of AIASGT will expand as the technology maturates. All

current AIASGT software components communicate with each other using a SOSCOE

environment as a foundation of autonomy. AIASGT will assist in what is known as

Command Mission Execution (BCME), which enables the operators (commanders, staff

and other users at all levels) to collaborate on decision-making, policy, report and order

generation and control of tactical and support resources. AIASGT will focus on the

commander’s interface, control of friendly battlefield elements, and the production of

orders autonomously. AIASGT will have to develop Situation Understanding (SU).

AIASGT SU supports the knowledge of objects within the battlespace environment. It

calculates the area of coverage for things like the Intelligent Munitions System (IMS) and

Unattended Ground Sensor (UGS) fields, updates the Battle Space Objects (BSOs) based

on priority and warfighter input, and forwards the BSOs, field locations and engagement

status alerts to current systems like the Force XXI Battle Command Brigade and Below

(FBCB2).

AIASGT used by the military use what is universally known as the Warfighter

Machine Interface Services (WMIS) or what is known as simply the interface. AIASGT

will incorporate Sensor Data Management (SDM). AIASGT will collect and disseminate

sensor data, status, and performance information from other AIASGT devices such as the

UGS and IMS fields and provides this information to consuming applications. Any

software / computer based system including those with AIASGT will incorporate some

form of Network Management System (NMS). A NMS is responsible for the

management of the FCS wireless and wired networks within an integrated network

management framework. It provides network planning, configuration management, and

performance management. These services provide Internet Protocol (IP)

Advertisementress planning for the network architecture, manual planning inputs for IP

and spectrum. They also provide planning, configuring, and monitoring platform status

of the network elements, including ICS router, and surrogate Soldier Radio Waveform

(SRW).

AIASGT components are built atop the System of Systems Common Operating

Environment (SOSCOE). AIASGT SOSCOE services provide a stable, well defined

interface between domain application software and the underlying operating system,

networks, and processors in a network-centric environment. AIASGT SOSCOE provides

a significant amount of infrastructure functionality, which has been rolled up into a high

level set of service families including administrative services, Task Integration Network

(TIN) services, and communication services. AIASGT automates data store services,

information assurance services, interoperability services, software support services.

AIASGT SWIT Integration

AIASGT software integration strategy focuses on production of an overall general

plan for each Integration Phase (IP) or release. When software is developed for any

application including AIASGT development, the integration test threads and cases run for

a software release, are intended to demonstrate software soundness and to reduce the cost

and schedule risk associated with determining this soundness during later subsystem and

system tests. This is especially important by the nature of full autonomy in any system.

AIASGT software simulations will be used to support build integration and system test

activities. Execution of AIASGT software integration is fundamentally different from

verification testing. The nature of any AIASGT software integration is to focus on error-

finding, informal control of the testing, flexible procedure sketches and a rapid

mechanism for incorporation of important fixes to the software. It is important to have

firm engineering level configuration control of the software article being integrated and

other laboratory hardware and software so that problems can be associated with the

correct software versions. It is important to capture and fully document problems

encountered so that fixes can be developed. Equally important is to capture and

document input and operational sequences to support characterization and corrective

action. Software Problem Reports (SPR) will be generated by the engineer who

determines the error in the AIASGT software. These reports will list all submitted

Software Problem Reports, which will then be reviewed to evaluate the reliability of the

AIASGT software.

AIASGT Software Defect Documentation, Correction and Reintegration

During any software integration process, defects found in a software end item will

require that the software defect be documented and corrected. An AIASGT software

defect that is unique to a specific supplier test environment and does not impact other

suppliers’ software will be documented and resolved in accordance with the procedures

of that particular supplier and in compliance with the AIASGT Software Configuration

Management Plan (SCMP). If an AIASGT software defect is found that is present in a

delivered software end item or will not be resolved prior to delivery of the software end

item, that defect must be documented, tracked, and assessed by the Software

Configuration Control Board (SCCB) for the organization that is receiving that software,

also in compliance with the SCMP.

The ability for any AIASGT to function is dependent on the software that gives

AI its autonomous capability. Software development, test and integration are the essence

of AIASGT technology. The governance and the quality of the software that empowers

AIASGT will follow software engineering best practices. When an AIASGT software

anomaly is discovered, a Software Problem Report (SPR) will be initiated as described in

the SCMP. A SPR may be an indication of the reliability of AIASGT and is an indication

of the quality of the software. In software engineering, the AIASGT SPR is then

determined to be valid by the submitter’s lead or manager. If the software product that

the SPR should be assigned to is unclear, the lead/manager can determine that the SPR

needs to go to the SCCB for review and disposition. If the AIASGT software anomaly is

determined to be out of scope of this SCCB, it needs to be transferred to the appropriate

SCCB for assignment. If the AIASGT software anomaly impacts hardware, if will be

forwarded to the Configuration Control Board (CCB) for disposition. Once the

appropriate integration group has been assigned the SPR, they will evaluate the AIASGT

software anomaly to determine if an error exists and what changes will be needed to fix

it. The detailed analysis will be recorded on the SPR.

Analysis of the AIASGT software anomaly can include impacts to cost, schedule,

technical baseline, computer resource utilization, performance, supportability and

impacts to other software baselines. This is critical for the development of AIASGT

software. During the software engineering process, the SCCB reviews each AIASGT

software anomaly to determine if a change should be implemented in a software product

or in procedures. If the SCCB that is developing the software approves the AIASGT

software anomaly fix for implementation, it will determine the software version/block in

which it will be implemented. The SCCB also determines applicability of verification

testing for each SPR in order to govern the development of AIASGT. If corrective action

is complete, the result of the test is recorded in the SPR and/or in a test problem report, as

applicable, and the SPR approved for release. Once the fix has been provided, a fix for a

software defect, the verification venue must verify the fix has not adversely impacted

previously working functionality. To accomplish this, the SPR verification venue must

not only test the specific fix but also perform the appropriate level of regression testing.

The AIASGT software anomaly verification venue is responsible for determining the

appropriate test cases that need to be exercised to verify that the implemented software

fix has corrected the defect and for determining the level of regression testing required

that insures previously working functionality has not been impacted. Regression testing

may be required at integration facilities other than the SPR verification venue. This will

be assessed on a case-by-case basis and if not specifically required by the SCCB will be

left to the discretion of those integration facilities. Depending on where the software is in

the integration flow, the software containing the fix can be delivered as an engineering

version or as a final version. The engineering versions of the software can be delivered

as a Quick Turn-around (QT), Alpha, or Beta version depending on the nature of the

defect and the need of the other integration facilities. Following software development

best practices, if the software has previously completed a qualification test, the software

must be re-qualified prior to delivery. The level of AIASGT re-qualification can vary

from re-running the complete qualification test to performing a subset of the qualification

test. When evaluating the impact of a AIASGT software anomaly, the SCCB will

consider whether or not re-qualification is required and if so to what extent. After new

software has been delivered with the software defect corrected and verified at the

assigned verification venue, the software would usually proceed through the normal

integration testing flow. If the AIASGT SCCB determines that a software defect is

critical to the completion of a test in a higher-level integration environment such as the

System of Systems Integration Lab (SoSIL) or in Field Test, the normal integration or

qualification testing flow (exclusive of safety testing) may be accelerated. In these cases,

the AIASGT SCCB will determine which integration facilities can be skipped in the

reintegration testing process. The integration facilities that are receiving the software that

has skipped an integration level must agree to accept the risk associated with skipping

intermediate integration test levels. As the Integration and Lab Focal Point Engineers for

a given phase will be key SPR and SCCB players, they will need to be very

knowledgeable of both the content in the FCS Software Problem Report Database Users

Guide and the Software Configuration Management Plan (SCMP) D786-10251-1.

AIASGT and Cyber Security

The governance and protection of the data that AIASGT processes is a paramount

concern for civilian and military planners (Arkin, 2011; Clarke, 2011; Cook, 2013; Davis,

1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013, Sparrow, 2012). The

national and economic security of the United States depends on the reliable functioning

of critical infrastructure. To strengthen the resilience of this infrastructure, President

Obama issued Executive Order 13636 (EO), “Improving Critical Infrastructure

Cybersecurity” on February 12, 2013. This Executive Order calls for the development of

a voluntary Cybersecurity Framework (“Framework”) that provides a “prioritized,

flexible, repeatable, performance-based, and cost-effective approach” for assisting

organizations responsible for critical infrastructure services to manage cybersecurity risk.

Executive Order 13636 notes that critical infrastructure is defined as systems and assets,

whether physical or virtual, so vital to the United States that the incapacity or destruction

of such systems and assets would have a debilitating impact on security, national

economic security, national public health or safety, or any combination of those matters.

AIASGT used by the military is a system and asset that uses cyberspace for various

functions that the EO delineates a national security concern. Due to the increasing

pressures from external threats, organizations responsible for critical infrastructure need

to have a consistent and iterative approach to identifying, assessing, and managing

cybersecurity risk (EO). AIASGT technology utilizes what is known as five framework

core functions.

The first function of mitigating any AIASGT cybersecurity related threat is

associated with its identification (EO). For the development of AIASGT, the developers

of each system have to manage their cybersecurity risk. The risks to deployed AIASGT

systems are extensive and can cause impact to the collection or dissemination of data.

To properly identify the function of cyber security and how it affects AIASGT,

the total development cycle of the AIASGT technology has to ensure that adequate cyber

security is incorporated into the management of the asset to include the business

environment, governance, risk assessment, and overall management strategy (EO). The

activities in the identification of the cyber security risks associated with AIASGT are

foundational for effective implementation of the cyber security framework necessary to

protect AIASGT. Understanding the business context, resources that support critical

functions and the related cybersecurity risks enable an organization to focus its efforts

and resources. Defining a risk management strategy for AIASGT enables risk decisions

consistent with the business needs or the organization. To protect AIASGT, there will be

a need to develop and implement the appropriate safeguards, prioritized through the

organization’s risk management process, to ensure delivery of critical AIASGT

infrastructure services. The protect function of AIASGT cybersecurity includes access

control, awareness, training, data security, information protection processes and

procedures, and protective technology.

The protect activities are performed consistent with the organization’s risk

strategy defined in the identify function. The detection phase will allow AIASGT

developers the ability to develop and implement the appropriate activities to identify the

occurrence of a cybersecurity event. The detect function includes following anomalies

and events, security continuous monitoring, and detection processes. The detect function

enables timely response and the potential to limit or contain the impact of potential cyber

incidents. In the respond phase of AIASGT cybersecurity, the challenge would be to

develop and implement the appropriate activities, prioritized through the organization’s

risk management process (including effective planning), to take action regarding a

detected cybersecurity event. AIASGT is autonomous by nature however the foundation

of AIASGT is computer automation. The respond function includes response planning,

analysis, mitigation, and improvements. The respond function is performed consistent

with the business context and risk strategy defined in the Identify function. The activities

in the respond function support the ability to contain the impact of a potential

cybersecurity event.

The AIASGT cybersecurity recover phase will be to develop and implement the

appropriate activities, prioritized through the organization’s risk management process, to

restore the capabilities or critical infrastructure services that were impaired through a

cybersecurity event. The recover function includes recovery planning, improvements,

and communications. The activities performed in the recover function are performed

consistent with the business context and risk strategy defined in the identify function.

The activities in the recover function support timely recovery to normal operations to

reduce the impact from an AIASGT cybersecurity event.

Summary

AIASGT is experiencing the scrutiny and associated political problems with the

employment of semi or fully autonomous systems on the battlefield (Epitropakis, 2012).

The problems are increasing for the United States Government and are in direct

correlation to the effectiveness and the increased popularity and utilization of

autonomous systems on the battlefield (Howlander, 2011). The use of these systems is

expanding primarily because of the inherent mitigation of loss of life of the personnel that

employ these systems along with the increased efficiency and endurance of the systems

(Geis, 2011). The argument has been made that the efficiency of these systems

transcends to the reduction of collateral damage against the hostile forces they are

employed against. TAM specifically related to intent to use needs to be further expanded

and tested to delineate the effect of the artificial, intelligent thinking nature of

autonomous platforms specifically employed to take human lives (Singer, 2013).

Autonomous platforms with imbedded AI are the most popular among military

commanders and they are employing these systems as soon as they are available (Geis,

2011). Donley (year) noted the demand for autonomous or semi-autonomous platforms

far exceeds the availability currently in the inventory (Geis, 2011). Geis (2011) also

suggested that the persecution of conflicts has been forever transformed by the use of

UAVs. Futurists have predicted that conflicts will be persecuted primarily by intelligent

self-governing autonomous systems (Sparrow, 2012).

The question of culpability for the destruction that autonomous platforms can bear

upon the enemy is a major topic of debate and research (Arkin, 2011; Clarke, 2011;

Cook, 2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013,

Sparrow, 2012). Many political and national leaders often debate as to how quickly

military planners gravitate towards the clean hands approach when it comes to the

employment of autonomous systems against human targets (Arkin, 2011; Clarke, 2011;

Cook, 2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013,

Sparrow, 2012). The lack of direct human involvement and the implications associated

with this detachment compels a need for theological research of artificially intelligent

autonomous systems that make decisions to terminate a human life.

Chapter 3: Research Method

The rapid proliferation of autonomous, self-governing platforms and the secret

nature of their employment by the U.S. military have increased pressure to fully quantify

and delineate their capabilities (Singer, 2013). Arkin (2011) and Geis (2011) both noted

that new technologies used by the military have historically made killing more efficient

than the longbow, artillery, armored vehicles, aircraft carriers, or nuclear weapons. The

problem is that autonomous self-governance has not been identified as a technology that

requires comprehensive policies and procedures to mandate compliance within known

and acceptable parameters (Arkin, 2011; Geis, 2011; Singer 2013). Therefore, the

purpose of this qualitative case study was to explore the perceptions of ex-military

personnel who have administrated artificially intelligent self-governing technology

designed to take human lives and the following research questions were used to guide the

study.

Q1. How do ex-military personnel perceive the usefulness of AIASGT?

Q2. What are the perceptions of ex-military personnel regarding the attitude to

use AIASGT?

The content of this chapter consists of the research method and design

justification, including details regarding the population and sample. The developments of

the survey questions are also discussed. The data collection, processing, and analysis

presented along with the assumptions, limitations, and delimitations as well as the ethical

assurances related to this study. The chapter ends with a summary.

Research Method and Design

A qualitative study was appropriate in this type of research as the quantitative

instruments needed to measure the impact of autonomous systems do not yet exist (Yin,

2013). A qualitative case study design was used in order to develop an in-depth depiction

of the participants’ perspectives (Yin, 2013) regarding the acceptance and use of the

AIASGT on the battlefield. This case study consisted of a survey as the main data

collection method as quantitative methods rely on objective questionnaires as the

research instrument (Yin, 2013). A case study offers rich, detailed information on a

particular phenomenon (Stake, 2008). Stake (2008) suggested a holistic approach be

used to answer the questions of how the collaborative system works, and what effect

collaborative communication can have on the franchised relationship.

A case study design was appropriate and chosen because it fits the criteria

described by Yin (2013) as the data from this study were applied to answer why and

when questions along with focusing on a contemporary phenomenon within a real-life on

text. A case study was appropriate when the researcher has little control over events

(Yin, 2013). The goals of a case study are to be exploratory, descriptive, or explanatory

(Yin, 2013). The case study procedures that need to be followed are planning, design,

preparation, collection, analyzation, and sharing stages (Yin, 2013).

A case study is an empirical inquiry used to investigate a contemporary

phenomenon in-depth and within its real-life context, especially when the boundaries

between phenomenon and context are not clearly evident (Yin, 2013). The case study

inquiry allows the researcher to cope with the technically distinctive situation in which

there will be many more variables of interest than data points and benefit from the prior

development of theoretical propositions to guide data collection and analysis (Yin, 2013).

This case study involved asking questions to ex-military personnel regarding their

opinion concerning the current and future use of AIASGT for military purposes. This

gave an indication as to the general acceptance of the technology if the civilian

population were given the opportunity to provide an educated opinion on its use.

The opinions of ex-military personnel who have used or are educated about

AIASGT may promote opinions that could dictate that AIASGT be strictly controlled or

banned. Question one was: Have you ever used AIASGT on the battlefield? This

question will help to quantify the degree of involvement an ex-military person had using

AIASGT. Question two was: Have you ever supported a mission that utilized AIASGT

on the battlefield? This question was also used to quantify the much larger pool of

personnel that were part of an operation that utilized AIASGT, but may not have had

direct involvement or development. The responses to these questions as well as others

allowed for the formation of a clear description of the research questions. Employing a

quantitative method or some other qualitative design would not allow answering the

research questions nor be in alignment to the purpose of this study.

Population

The ideal population from which participants were drawn was anyone who will

develop or use AIASGT. Reaching a population so vast would be a challenge so the

population was limited to ex-military personnel who are members of the American

Legion (AL) Post 346 located in Neptune, New Jersey. The AL participants chosen were

a subset of the population that is generalizable to the ideal population of those who

develop or use AIASGT.

The AL was chartered by Congress in 1919 as a patriotic veteran’s organization.

Focusing on service to veterans, service members, and communities, the Legion evolved

from a group of war-weary veterans of World War I into one of the most influential

nonprofit groups in the United States (AL history, 2015). Membership swiftly grew to

over 1 million with local posts across the country (AL, 2015). Today, membership stands

at over 2.4 million in 14,000 posts worldwide (AL, 2015). The posts are organized into

55 departments, one in each for the 50 states along with posts in the District of Columbia,

Puerto Rico, France, Mexico, and the Philippines (AL, 2015). Over the years, the Legion

has influenced considerable social change in America, won hundreds of benefits

for veterans and produced many important programs for children and youth (AL, 2015).

Therefore, drawing participants from this well-known and recognized entity allowed for

ample diversity and ensure saturation of data was possible. In terms of characteristics of

the population, the veterans range from pre-computerized technology usage in the

military to those who recently left the service (AL, 2015). Due to the focus of the study,

it was necessary to recruit from a subset of the overall population regarding those with

more recent military experience.

Sample

There is no past research from the user or developer’s perspective concerning the

assimilation of AIASGT by members who have served in the military and are no longer

governed by the Uniform Code of Military Justice (UCMJ). The members in this study

could give their opinion freely without concern for permission by the Department of

Defense (DOD) because they were no longer active duty or affiliated with the military.

The case study research plan involved use of a purposeful sample of 10 personnel who

have used AIASGT (Guest, Bunce, & Johnson, 2006). Purposeful sampling was used as it

is best suited to construct multi-perspectival, emancipatory, participatory, and

deconstructive interpretations for qualitative data collection (Harsh, 2011), and gather a

purposeful sample from the ex-military personnel from the AL Post 346. Purposeful

sampling from a group that is both knowledgeable and familiar with AIASGT was the best

methodology to employ for this study because this was a specialized case wherein in-depth

understanding was being sought.

The sample size was predominantly a discretionary judgmental choice with the

objective to detect a measureable effect with the approach (Yin, 2013). According to

Guest, Bunce, and Johnson (2006), saturation is possible upon the completion of six

surveys, therefore, a purposeful sample size of 10 participants is considered optimal.

However, it is noted that should saturation not be reached after the completion of 13

surveys, participants would have been recruited until such time saturation was reached.

An invitation to participate in the study was placed on the AL Post 346 website

and monthly paper that included a web-link to the post website that allowed participant

participation in the survey. The AL Post Commander made the announcement to the AL

members requesting participation in the study. The names and participation of the 600

members of the AL Post 346 located in Neptune, New Jersey were requested by both

email and conventional mail invitations. Screening was in the form a questionnaire that

quantified their suitability for the study. The first 10 participants who passed screening

were selected for the study. Suitability was determined if they had been exposed to

AIASGT while in the military. All participants were ex-military who were not governed

by any military protocol because they are civilians and have no secret or official use only

knowledge of any current operational use of AIASGT.

Materials/Instruments

The research problem of the acceptance and willingness to use AIASGT was

explored. The instruments (see Appendix D and E) was used to focus on the research

questions that are in direct correlation with the research problem. The questions were

straight forward to mitigate and ambiguities in the interpretation of results. The study

assimilated legacy content contained in non-classified public records. Current peer

reviewed information that already exist provided an outstanding resource and was easily

accessed. Peer reviewed information concerning the current use of AIASGT were

industry accepted as factual and facilitated the comparison of information concerning the

acceptance and use of AIASGT. The use of public records data was easily obtained and

cost effective. This study utilized a web-based questionnaire that delineated the study’s

purpose with the participants.

The instruments (see Appendix D and E) were comprised of questions that were

designed to allow each participant the ability to articulate their position concerning the

acceptance and use of AIASGT. The questions were accessible to all the participants via

email and Google Docs. Yin (2009) noted that an email survey methodology is effective

in mitigating any prejudice that may manifest itself during a face to face survey. Each of

the survey questions was designed to evoke the opinions of the participant’s willingness

to use or accept the use of AIASGT. Surveys of the participants in this case study

consisted of questions designed to gain a perspective of the suitability and overall

knowledge of the participants on the employment or operational aspects of AIASGT used

by the military as well as screening demographic questionnaire (see Appendix C) to

ensure the individual is qualified to participate. The general survey questions were

developed based on an extensive review of related literature to quantify the participants

past military involvement with AIASGT as well as garner the perspectives of AIASGT

employment. The survey instrument questionnaires (Appendix D and E) were centered

on the opinions of a military professional concerning the use and acceptance in their

previous role in operating AIASGT. All 15 questions (Appendix D) and 3 open ended

questions (Appendix E) were designed to answer the research questions concerning the

overall use and acceptance of AIASGT.

A field test was utilized in accordance with the validation process described by

Yin (2013), in that the survey questions used in the field test were designed to validate

the research design methodology along with the associated data collection

instrumentation for the actual case study. The process utilized in the field test was in

accordance with Northcentral University’s (NCU) policies, procedures, and subsequent

approval. The field test was conducted with the post commander and the deputy who

were not part of the overall study. The survey questions were validated by the post

commander and the deputy, who were non participants. Changes were made at the

request of the post commander or the deputy during the validation. A field test was

conducted in accordance with the data collection methodology necessary to make a claim

(Yin, 2013). Electronic data collection in the form of questionnaires delivered via

Google Docs was an optimal solution and the methodology used to assemble and process

all data. The AL website hosted a link to the data collection portal. The AL monthly

newspapers were also Advertisement a link to the data collection portal (Appendix F).

There was an announcement to all the members during a membership meeting that

directed participants to the web portal for data collection.

All materials including the online informed consent form were presented to the

participants simultaneously for their convenience. The Google Docs website was used

for data collection. Many participants were not expected to participate in the survey at

the time it was distributed because many work a variety of shifts. Again, the web-based

Google Docs distribution proved to be optimal because it allowed participants working

various shifts to complete the survey. Questionnaire responses were analyzed and

categorized for similarities and differences, and coded, leading to emergent themes (Merriam,

2009). The questionnaires were saved on the Google Docs portal set up specifically for this

data collection. There was no need for follow-up or revalidation of the questions because the

questions were validated by the initial field test (Yin, 2013).

Data Collection, Processing, and Analysis

Data collection was conducted through the recording of questionnaire results

utilizing Google Docs, a web-based survey host along with the in-depth analysis of peer

reviewed documents pertinent to and in direct correlation with the study. Data were

collected from the AL members using a survey questionnaire that included questions

focusing on their exposure to AIASGT. Yin (2009) suggested that each of the survey

questions was in direct correlation to one or more of the main research questions. The

study data were derived from the direct questions asked making the process expeditious

for any survey. Each question was asked by the questionnaire to ascertain the

participants’ experiences with AIASGT and their views on the acceptance and

willingness to use the technology. Prospective participants found using the AL

newsletter, website along with the post commander’s endorsement and public

announcement requesting maximum participation. The researcher made contact with the

participant via email when the participant returned the survey questionnaire and informed

consent forms. After this is accomplished, the researcher introduced himself via email

and invited the AL member to participate in the study. Only after the AL member agreed

to participate in the study, survey questions returned, forms signed, and secured by the

researcher, did the participant receive the research questions. The e-mail survey structure

served to eliminate any prejudice that the researcher might have shown in a face-to-face

survey (Yin, 2009).

The elimination of statistical analysis allows qualitative researchers to be more

creative in dealing with sampling issues. They do not have to endure the strenuous

randomization process of sampling procedure because the results cannot be generalized

to a bigger population, and only analytical generalization can be conducted where a

particular set of results is generalized to a broader theory (Yin, 2009). Flick (2009)

suggested that the individuals or cases are selected as participants for a qualitative study

not because they represent their population (and therefore, the issue of generalizability)

but owing to their relevance to the research topic. Inevitably, the idea of randomization

outmoded the idea of nonprobability sampling or nonrandom sampling.

The data collected from the responses received during the surveys were compared

to TAM and its five factors (external factors, perceived usefulness, perceived ease of use,

intention to use, and attitude to using) using a pattern-matching technique presented to

align the TAM constructs to the responses by the participants. The research data needed

for this study were obtained from information obtained from explanatory content analysis

of archival records and the analysis of questioners. The content analysis helped uncover

cause and effect, identify core consistencies, patterns, and themes (Krippendorff, 2012).

Constructs were used for coding of the data, and analysis of the data was consistent with

the approach of relying on theoretical propositions as the first and most preferred strategy

for data analysis (Yin, 2009). Coding was used as an initial step towards more in-depth

analysis, arranging things in a systemic order, and making it part of a system or

classification to categorize (Saldana, 2009). The data was categorized and coded using a

process called open coding (Strauss & Corbin, 2008). Open coding is a process by which

the data is categorized by themes that appear to describe the phenomenon under study

(Strauss and Corbin, 2008). The study results offer developers of AIASGT effective

criterion to judge autonomous capabilities against the willingness to use and accept AI.

This study serves as a foundation to continued research into the necessary constraints

developers of AIASGT have to employ. The research was coded for patterns when

reviewing survey questioners from the participants. All documents were aligned and in

correlation with the study to extrapolate the emerging themes. This case study used a

survey as the main data collection method; quantitative methods rely on objective

questionnaires as their research instrument, which were collected for analysis (Yin,

2009). This study is descriptive in that the goal was to document a phenomenon and the

real life situations from where the phenomena originated (Yin, 2009).

Assumptions

It was assumed that all participants were active AL Post 346 members at the time

the study was conducted and that the participants were operators or developers of

AIASGT. All AL participants signed a consent form. All participants were informed of

the confidential nature that all information will be treated with, and the intentions of the

researcher with regards to the reason and nature of the research study. The participants

were made aware that their participation would be mutually beneficial for all and how

important their opinions were concerning the use of AIASGT. Transparency was

achieved by making available, and propagating all results to the participants. Credibility

and ease of any trepidation associated with the participants was achieved by the

researcher being a civilian who also was honorably retired ex-military after 21 years of

service. It was assumed that all participants responded truthfully and accurately to all

survey questions. The only assumed limitation was availability of the participants. There

was no expectation that there would be any abnormalities in any interpretation or

subsequent synthesis of the data.

The expectation was that many participants would not be able to participate in the

survey at the time it was distributed because many work a variety of shifts. It was

assumed that the web-based Google Docs distribution would be optimal because it

allowed participants working various shifts to complete the survey. The anonymous

nature and the inherent security features of the web-based Google Docs data collection

methodology allowed many of the participants the desired anonymity enabling them full

freedom to participate in the study. Questionnaire responses were analyzed and

categorized for similarities and differences, and coded, leading to emergent themes (Merriam,

2009). An email survey methodology is effective in mitigating any prejudice that may

manifest itself during a face-to-face survey (Yin, 2009). This data collection

methodology enabled the participants to take as much time as desired to fully articulate

the acceptance and willingness to use AIASGT.

Limitations

The expectation was that the primary limitation of this qualitative descriptive case

study would be the multiple types of AIASGT and the permutations associated with the

technology associated with its rapid development and insertion into operation. This

limited the study to existing or legacy technology. Limitations were that the AL

population would not respond to the request for participants. AL personnel respond best

with a questionnaire data collection methodology rather than face-to-face interviews.

According to Guest, Bunce, and Johnson (2006), saturation is possible upon the

completion of six surveys, therefore, a purposeful sample size of 10 participants is

considered optimal. An alternate plan was that should saturation not been reached after

the completion of 10 surveys, AL participants would be recruited until such time

saturation was reached. Another expected limitation included the problems of inference

which relate to the possibility of drawing acceptance and use of AIASGT due to current

global war or near war events. Current use of AIASGT is highly publicized and may

cause participants to draw conclusions about the use of AIASGT; meanwhile there may

be underlying constructs such as motives, attitudes, or norms regarding that the use of

AIASGT that may obscure rational thought. An expected limitation to the study was the

focus on only ex-military personnel. Many civilian uses of AIASGT have proven to be

gaining acceptance and use and the expectation is that civilian uses of AIASGT will

outpace the military’s utilization of the technology. Yin (2009) noted that the case study

methodology does not measure a phenomenon’s prevalence, but is a valuable method for

better understanding the phenomenon in its totality. A limitation may have been due to

the sample being restricted to ex-military personnel that has experience with AIASGT

and not involving or considering civilian application of the technology. An expected

limitation was due to the survey questionnaire methodology used for data collection. To

mitigate these limitations, adherence to proven methodologies of survey based data

collection implemented. Yin (2009) noted that the e-mail survey could be an

advantageous way of data collection. The e-mail survey structure served to eliminate any

prejudice that the researcher might have shown in a face-to-face survey (Yin, 2009).

Data collection was in the form of web-based questionnaires to conduct a systematic

content analysis for emergent themes (Harsh, 2011; Yin, 2013).

Despite these limitations, the expectation was that the strength of the study of

AIASGT would easily be extended to a larger sample population within the civilian

arena, with the same methodological approach. AIASGT will give machines full

autonomy, allowing them to make independent decisions without any human

intervention. No empirical research has been found regarding the impact of what has

been described as technology creep associated with the assimilation of sophisticated

artificial intelligence (AI) into autonomous systems. The focus of this study was upon

the assimilation of AI into self-governing systems that are utilized for the specific

purpose of terminating human lives. The technology acceptance model (TAM) and its

five factors (external factors, perceived usefulness, perceived ease of use, intention to

use, and attitude to using) were used as the theoretical lens of this study and to establish

the relationship between the findings generated from the data collected and the literature

review.

Delimitations

Delimiters include the study of AL that has experience with AIASGT. Their

experiences with the technology may be perceived differently since reverting to civilian

status. This delimitation allowed the research to concentrate on AL personnel

exclusively. This delimitation allowed the research to concentrate on ex-military

personnel that had experience with AIASGT technology. The delimitation delimited the

findings that emerged from the study to the AL personnel that frequent the AL. The

survey collection methodology for data collection including primary and secondary data

selected for this study further delineated the scope of the findings solely towards the

acceptance and willingness to use AIASGT by AL affiliated personnel. The data

collected form AL members narrowed the scope by focusing solely on AL operators of

AIASGT. This delimits the transferability of the findings of the study to AIASGT use

primarily by the military. Subsequent research should replicate this study in other

industries that develop or operate AIASGT.

The expectation was that the utilization of data collection from Google Docs

would allow the convenience and automation of the data collection (survey) process,

however, this methodology adversely impacted the opportunity for the researcher to have

the benefit of first person observation of the participants. The Google Docs data

collection methodology was utilized despite the limitations based on the convenience of

electronic data collection for both the participants and the researcher. Yin (2009)

identified electronic collection of survey documents as advantages by noting that an

email survey methodology is effective in mitigating any prejudice that may manifest

itself during a face to face interview.

Ethical Assurances

In accordance with established procedures, the Northcentral University

Institutional Review Board (IRB) reviewed all of the required documents to verify that

there was no safety or security risks associated with any of the participants at any time

during the data collection process. The researcher complied with the rules established by

the institution as the important cornerstone of ethical principles (Lensink, 2011). AL

members were provided complete details about the survey in advance. The members in

this study were able to give their opinion freely without concern for permission by the

Department of Defense or any other governmental entity because they are no longer

active duty, affiliated with, or know of any military operation besides what is publicly

disclosed. The data in the study does not include AL member’s names, ensuring that the

AL member’s personal information and answers to the questionnaires remain

confidential. Full disclosure concerning the purpose of the research given to all of the

AL members and an informed consent form (see Appendix A) was sent to each

participant in advance of the study. Participants were informed that their participation in

the study was voluntary and that they could drop out of the study at any time.

There were no deceptive methodologies used to access information in the study.

Any information concerning this study will be maintained confidentially. AL members

received the contact details of the researcher and had full access during the entire process.

Only after the AL member agreed to participate in the study, survey questions returned,

forms signed, and secured by the researcher did the participant receive the research

questions. All information was submitted to a secure web-based Google Docs data portal

giving flexibility and security the online survey. Results of the survey were provided to

all AL members at no charge and information regarding the administration of the surveys

was available to all AL members.

Summary

An in-depth, qualitative research of ex-military personnel that have had

experience with AIASGT was assimilated. The areas of consideration were based on the

theoretical framework for this study which is the Technology Acceptance Model (TAM;

Davis, 1989), which is rooted in the TRA developed by Ajzen and Fishbein (1973). The

Davis (1989) TAM and its five factors (external factors, perceived usefulness, perceived

ease of use, intention to use, and attitude to using) were used as the theoretical lens of this

study and to establish the relationship between the findings generated from the data

collected and the literature review. TAM specifically related to intent to use needs to be

further expanded and tested to delineate the effect of the artificial, intelligent thinking

nature of autonomous platforms specifically employed to take human lives (Singer,

2013). The outcome of this study expands the body of knowledge of the Davis (1989)

TAM and all of the various iterations of it that did not specifically the acceptance of AI.

One outcome of the findings of this research may result in the development of

multinational agreements limiting and controlling the assimilation of AI. The findings of

this study allow constructs, patterns, themes, and determinations that will assist future

planners in considering the possible impact that AIASGT can have if the proper

constraints are not put into place.

Finally, the results of this study can help those that who are responsible for

technology refresh or insertion to question the capability of the AI that is embedded into

the technology they are responsible for. The opinions of ex-military personnel who have

experience with the operation and development of AIASGT had not been consulted

concerning its actual employment. The rapid proliferation of autonomous, self-governing

platforms and the secret nature of their employment by the U.S. military have increased

pressure to fully quantify and delineate their capabilities (Singer, 2013). Questions

concerning the specificity of the constraints implemented by the manufactures of

AIASGT technology can be mitigated by utilizing the results of this study to develop

tools to better understand the concerns of researchers and civilian personnel affected by

the development and assimilation of AIASGT. Once AIASGT is inserted into the

society, it may be too late to put the needed safeguards in place.

100

Chapter 4: Findings

The impetus of this qualitative, purposeful case study was to explore the

relationship between TAM and AIASGT. The researcher set out to explore the ways that

the acceptance and use of technology in TAM would permutate with the postulation of

AIASGT. A purposeful case study methodology explored the use and acceptance of

AIASGT by utilizing existing archival data and research participants by conducting a

hard copy questionnaire. This chapter delineates the results of each research question.

Archival and legacy research utilized to validate, solidify and codify the findings.

Demographic data collected from all participants qualifying them prior to the collection

and analysis of the research data. The conclusion of the chapter consists of a summary of

findings obtained from the research questions in correlation to emergent themes and their

overall evaluation.

After receipt of NCU IRB approval on 26June16, the research began with the

recruitment of AL members to become participants in the study. The participants were

all recruited by the AL post website, newsletter, and a public endorsement from the post

commander with the recommendation to participate in the study. Identified participants

selected and the informed consent form was populated and signed. Once the forms

secured, the ability of the participant to access the survey instrument obtained. For

convenience of the participant, an electronic means to conduct the survey proved to be

optimal. All of the survey questionnaires were coded and responses to the research

questions were organized in correlation to the responses. Every participant provided

robust and complete answers to each of the questions. Flick (2009) suggested that the

individuals or cases are selected as participants for a qualitative study not because they

101

represent their population (and therefore, the issue of generalizability) but owing to their

relevance to the research topic.

Yin (2009) suggested that a purposeful case study could be the appropriate design

under several circumstances. Yin (2009) noted that this type of study represents the

critical case in testing a well-formulated theory. The theory probably has some specified

sets of proposition or assumption that are contained in specified circumstances (Yin,

2009). Data was arranged in categories, and placing collected data within categories (Yin,

2009). Data from such study can contribute significantly to knowledge and theory-

building process (Baker & Ishak, 2014). Baker & Ishak (2014) noted that qualitative

researchers can choose to have a case design, when the case represents an extreme or

unique case such as AIASGT.

During data collection, identifying the order of problems by urgency will be

important to identify the issues that need immediate intervention and devise a plan of

action for them before the rest of the issues (Yin, 2009). To employ such a design, the

researcher has to be certain that the phenomenon under study is very rare and the

participant that has the specific characteristics is very few and far in between (Baker &

Ishak, 2014). If other researchers have similar opportunities and can cover some

prevalent phenomenon previously inaccessible to social scientists, such conditions justify

the use of single-case study on the grounds of its revelatory nature (Yin, 2009). A case

study can be the design of choice when it involves revelatory case (Baker & Ishak, 2014).

Interview responses were analyzed and categorized for similarities and

differences, and coded, leading to emergent themes (Merriam, 2009). After an

examination of the textual data the researcher defined an initial set of codes with which to

102

organize the data (Yin, 2009). The research utilized open coding that analyzed the data

and reduced it into identifiable emerging themes. Coding was utilized as the initiating

step towards a robust in-depth analysis, arranging things in a systemic order, and making

it part of a system or classification to categorize (Saldana, 2009). Yin (2009) suggested

that a researcher would manually review the textual data for coding and qualitative

analysis, and this was accomplished. The data coding process identified the emergence

of meaningful patterns among the words or phrases in the textual data that matched the

initial set of codes (Yin, 2009). The author elaborates by noting that coding provided a

formal mechanism for organizing the data. In accordance with Yin (2009), several codes

were created based on emerging themes and patterns that emerged from the interview

data, and data were organized according to the appropriate category (Yin, 2009).

All participants provided demographic data as part of the study. All participants

were associated with the AL Post 346. As part of the data collection process, the

demographic survey was populated along with all informed consent forms. All informed

consent forms were secured before the electronic propagation of the questionnaires. The

survey questions were specifically designed to extrapolate the data necessary for a robust

study sample. Data was obtained by AL Post 346 whose members are located within the

New Jersey area. All of the participants are from the New Jersey area. All participants

were exposed to AIASGT while in or around the military. The average time that a

participant spent in the military and exposed to AIASGT was 4 years, however there

were personnel who participated in the study that have been exposed to AIASGT for over

20 years. All participants exposed to the development of AIASGT utilized for both

military and civilian applications for over 8 years. The question concerning the total

103

exposure to AIASGT by each participant requested by the researcher to determine if there

was an exposure to both civilian and military applications of AIASGT. The difference in

the opinion concerning the use and acceptance of AIASGT was determined as negligible

with minute differences in the responses based on the age of the member. All participants

were ex-military or civilians that were part of or serviced the US Armed Forces. All

members held a college degree. All but one of the participants lived within 5 miles of the

AL facility. The average exposure to AIASGT was 14 years, but there were participant

that had exposure that ranged from eight to twenty years. No participants exposed to

AIASGT for less than eight years. Nine out of ten of the participants were male and there

was one female in the study. All participants were college graduates.

Table 1

Participant Demographic Information (Listed in order of experience)

Results

The chapter reiterates the questions that were created to guide the research of

AIASGT along with the process and methodology is used to gather the necessary

Age Gender Years of College Degree

63 M 20 Y

62 M 20 Y

63 M 18 Y

60 M 15 Y

58 M 15 Y

57 M 10 Y

55 M 10 Y

53 M 10 Y

50 F 8 Y

104

information for comprehensive analysis. The results from this case study will be

delineated and categorized under separate subjects. The chapter concludes with a

summary of the findings extrapolated from the salient points. The study employed a

comprehensive well-delineated survey that contained both demographic along with open-

ended questions used by participants. All documents related to the accomplishment of the

study were retrieved, reviewed, and properly secured. The questions proposed to the

participants were open-ended and easily comprehendible. All of the responses collected

from the participants were meticulously categorized and coded. The data further

categorized into intelligible themes extrapolated from the raw data of the participants.

There were 29 pages of data collected. Each participant produced and average of 2-4

pages of data.

How do ex-military personnel perceive the usefulness of AIASGT? The basis of

the study was encompassed in the first research question and that was how do ex-military

personnel perceive the overall usefulness of AIASGT. The perceived usefulness of

AIASGT by those that have used it was the impetus of the study.

This research question was on how the operator of AIASGT perceived its

usefulness. All participants answered this question unanimously and perceived AIASGT

as useful. One theme that emerged was concerning the mitigation of friendly force or first

responder casualty’s. These findings support Arkins (2011) assessment that the intent of

the military is to assimilate autonomous technology into its war making capabilities in an

auspicious attempt to mitigate friendly force casualties (Arkin, 2011). Mackey (2013)

and Arkin (2011) noted a key goal for military commanders is to mitigate friendly-force

105

casualties by utilizing AIASGT that is fully independent of human interaction. The

perceived usefulness of AIASGT was in direct correlation to its ability to be autonomous.

The survey questions that examined the perceived usefulness were questions 1

through 6. Most of the research participants responded to the survey questions with

simple yes or no answers. There was no distinction given concerning the willingness to

uses AIASGT between civilian or military applications of the technology. Everyone that

participated in the survey answered in the affirmative. The willingness to use AIASGT

was not only supported but also considered essential.

The second survey question was would you willingly use AIASGT in combat?

Like all of the survey questions, this question was designed to give the survey participant

the opportunity to elaborate. Simple yes answers given proving the willingness to use

AIASGT in combat, which was universally supported. Comments from 3 of the

participants noted that their willingness to use was to minimize friendly casualties. The

rapid development and adaptation of technology was considered essential.

Survey question 3 explored if the participant felt that enough transparency and

regulations are in place. The major theme gathered was that there was not enough

transparency and that development of AIASGT is considered compartmentalized. The

dichotomy was that even with the perceived lack of transparency, participants universally

supported the development of AIASGT. Along the lines of transparency, survey question

4 asked do you think that you can ethically use AIASGT? All participants answered in

the affirmative. Participants noted that ethics is considered paramount in western

development and use. In contrast, ethics was considered a low or no priority for foreign

nationals that were developing AIASGT.

106

Survey question 5 asked the participant if they would advocate for the use of

AIASGT for the military. Every participant answered in the affirmative. All participants

expressed their willingness to use AIASGT in combat and universally supported. A

strong theme concerning the willingness to use AIASGT in an effort to minimize friendly

casualties. Rapid development and adaptation of technology considered essential. Survey

question 6 asked were you ever asked on your willingness to use AIASGT? A common

theme was that no opinion was ever requested from any of the participants concerning

their willingness to use or not use AIASGT. AIASGT was part of daily environment and

the willingness to use AIASGT considered simply part of the position.

Table 2

Summary of Survey Themes in Relation to the Research Questions

Survey Question Themes

Q1. Would you willingly use AIASGT? • Willingness to use AIASGT was supported.

• Willingness to use considered essential.

Q2. Would you willingly use AIASGT in

combat? • Willingness to use AIASGT in combat

universally supported.

• Willingness to use to minimize friendly casualties.

• Rapid development and adaptation of technology considered essential.

Q3. Enough transparency and regulations are

in place? • Not enough transparency • Development of AIASGT considered

compartmentalized.

Q4. Do you think that you can ethically use

AIASGT? • Ethics considered paramount in western

development and use.

• Ethics considered a low or no priority to adversaries.

Q5. Would you advocate for the use AIASGT

for the military? • Willingness to use AIASGT in combat

universally supported.

• Willingness to use to minimize friendly casualties.

• Rapid development and adaptation of technology considered essential.

107

Q6. Were you ever asked on your willingness

to use AIASGT? • Willingness to use or not use AIASGT

was never asked.

• AIASGT was part of daily environment. • Willingness to use AIASGT considered

simply part of the position.

What are the perceptions of ex-military personnel regarding the attitude to

use AIASGT? The second research question was the operator’s attitude to the actual use

of AIASGT. Participant’s attitude to the use of AIASGT was unanimous acceptance.

Participants considered development of AIASGT is considered compartmentalized

resulting in inadequate oversight. Participant's responses described their participation in

providing any feedback concerning the assimilation of AIASGT as non-participatory.

Exposure to AIASGT and the associated system upgrades were described as cyclic

technology refresh. The pace of systemic upgrades to the existing AIASGT were only

limited to what was described as budgetary constraints and contractual challenges.

System upgrades along with increased capability of AIASGT were considered mission

essential and a top priority.

The survey questions that examined the attitude to using AIASGT were survey

questions 7 through 12. In all cases, most of the research participants responded to the

survey questions with simple yes or no answers. Themes did arise as the data collection

continued to support a positive attitude to using AIASGT. Survey question 7 asked are

there enough regulations of AIASGT for civilian use? This question was designed to

solicit a response from the participant in delineating the differentiation of military and

civilian use of AIASGT. The data collected noted that participants were indifferent

between the civilian or military use of AIASGT. Participants felt that current regulations

considered adequate. Participant’s comments indicated concerns over the preservation of

existing privacy regulations.

108

Survey question 8 asked is there enough information to accept AIASGT for

civilian use? All participants responded positively. Comments were that there is enough

information that exists with western development of AIASGT. Major themes were

concerns over the development of AIASGT by adversaries. Participants indicated that the

development of AIASGT for civilian use is generally transparent and acceptable. As with

question 8, question 9 asked if there enough transparency and regulations for AIASGT

for civilian use. Again the participants did not distinguish civilian from military use of

AIASGT and there were themes developed that suggested that there was enough

transparency that exists with western development of AIASGT.

Question 10 asked if you would advocate the use of AIASGT for the civilian use.

The use of AIASGT for civilian use was universally accepted. A theme that was present

was for the avocation and acceleration of AIASGT capabilities. The concern was not

concerning the development of AIASGT; it was to accelerate these capabilities.

Participants were concerned that we would fall behind on the development of AIASGT

by being over regulated. Question 11 asked, “Would you advocate strict regulations of

the use of AIASGT?” The themes generated were decidedly against any additional

regulations. Current regulations were considered adequate and robust. Concerns over the

preservation of existing privacy regulations were common. Question 12 was designed to

affirm question 11 which was “are you concerned about regulation of AIASGT?” The

affirmation was decidedly that current regulations were adequate and robust. One theme

that surfaced again were for the preservation of existing privacy regulations.

109

Table 3

Summary of Survey Themes in Relation to the Research Questions

Survey Question Themes

Q7. Are there enough regulations of

AIASGT for civilian use? • Regulations considered adequate

and robust.

• Concerns over the preservation of existing privacy regulations were

common.

Q8. Is there enough information to accept

AIASGT for civilian use? • Enough information exists with

western development of AIASGT.

• Concerns over the development of AIASGT by adversaries.

• The development of AIASGT for civilian use is generally transparent

and acceptable.

Q9. Is there enough transparency and

regulations for AIASGT for civilian use? • Enough transparency exists with

western development of AIASGT.

• The development of AIASGT for civilian use is generally transparent.

Q10. Would you advocate the use of

AIASGT for the civilian use? • Enough information exists with

western development of AIASGT.

• Concerns over the development of AIASGT by adversaries.

• The development of AIASGT for civilian use is generally transparent.

Q11. Would you advocate strict

regulations of the use of AIASGT? • Regulations considered adequate

and robust.

• Concerns over the preservation of existing privacy regulations were

common.

Q12. Are you concerned about regulation

of AIASGT? • Current regulations considered

adequate and robust.

• Concerns over the preservation of existing privacy regulations were

common.

Evaluation of Findings

The evaluation of the findings of this study were provided from the results of the answers

given to the twelve survey questions designed to answer the two research questions that

110

were the impetus of the study. Research of AIASGT showed that the consideration of

such technologies originally surrounded the need for a machine to have the ability of

independent action, which is especially useful for mundane or routine activities,

particularly in terms of economic advantages in commercial applications (Singer, 2013).

Utilizing archival along with participant involvement, the comprehensive analysis of the

research questions are as follows. The research questions were designed to codify the

user’s perspective of acceptance and use of AIASGT. Members that participated in the

study provided delineated and comprehensive responses. All responses to the survey

questions categorized and reviewed. The researcher expropriated themes, perceptions,

and the frequency of exposure to AIASGT. The research questions were designed to

solicit a robust response from the participant’s. The responses provided data concerning

the operator’s perspective on the acceptance and actual use of AIASGT.

The researcher conducted a study with 10 participants. According to Guest,

Bunce, and Johnson (2006), saturation is possible and a purposeful sample size of 10

participants considered optimal. When conducting data analysis, the frequency of similar

answers suggested data saturation was obtained. The results suggested that if further

research were conducted, the results of the continued research would most likely produce

the same or similar results. The researcher observed and recorded salient themes codified

from the careful analysis of the collected data. The impetus of the study was to

determine how the operator of AIASGT perceived the usefulness and the attitude to using

the technology.

The first research question was "How do ex-military personnel perceive the

usefulness of AIASGT?” The data collected to answer the research question was

111

extrapolated from the results of survey questions one threw six. Similar research from

Mackey (2013) and Singer (2013) indicated the use of advanced AIASGT, which were

designed to mitigate or eliminate the human element out of the planning, preparation, and

execution of an act. Further reinforcement of the analysis obtained from the remaining

questions. All survey data collected obtained from the ten participants of the study. The

researcher utilized questions one threw six of the survey instrument to assemble, analyze

and extrapolate salient themes to draw conclusions on the perceived usefulness of

AIASGT by the operator. While the intent of AIASGT is to assist the operator, the

perception of the operators concerning the usefulness was universally accepted. The

governance and protection of the data that autonomous technology processes is a

paramount concern for civilian and military planners (Arkin, 2011; Clarke, 2011; Cook,

2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013, Sparrow,

2012). The Davis (1989) TAM and its factor (usefulness) were used to establish the

relationship between the findings generated from the data collected and the literature

review. While the producers and purchasers of AIASGT propagate the capabilities to the

field, the perception of the usefulness by the operators of the technology were rarely

considered. Despite the compartmentalization surrounding the development of AIASGT

and limits the input from the operators of the technology has in its development, the

technology was universally accepted as useful.

The second research question, “What are the perceptions of ex-military personnel

regarding the attitude to use AIASGT?” The data collected to answer the research

question extrapolated from the results of survey questions seven threw twelve. Further

reinforcement of the analysis obtained from the remaining questions. All survey data

112

collected obtained from the ten participants of the study. The researcher used questions

seven threw twelve of the survey instrument to assemble, analyze and extrapolate salient

themes to draw conclusions on the attitude to using AIASGT by the operator. Research of

advanced autonomous technology showed that the consideration of such technologies

originally surrounded the need for a machine to have the ability of independent action,

which is especially useful for mundane or routine activities, particularly in terms of

economic advantages in commercial applications (Geis, 2011; Singer, 2013).

The Davis (1989) TAM and its factor (attitude to using) were used to establish the

relationship between the findings generated from the data collected and the literature

review. Participant’s attitude to the use of AIASGT was unanimously accepted. Themes

emerged that codified the position that participants considered development of AIASGT

as essential. Despite the compartmentalization surrounding the development of AIASGT

and limits the input from the operators of the technology has in its development, the

technology received a positive attitude to use by every participant. Some themes

emerged that reveled trepidation concerning the erosion of privacy rights. Davis’ (1989)

TAM and its five factors (external factors, perceived usefulness, perceived ease of use,

intention to use, and attitude to using) was used as the theoretical lens of this study.

Singer (2013) noted that TAM intent to use needs to be further expanded and tested to

hypothesize the effect of the artificial, intelligent thinking nature of autonomous

platforms specifically employed to take human lives. The outcome of this study will

expand the body of knowledge of the Davis (1989) TAM and all of the various iterations

that did not specify the acceptance or use of AI and the subsequent changes that it would

produce to the model.

113

Summary

The study was designed to answer the research questions concerning the

acceptance and use of AIASGT from the operator’s perspective. In order to ascertain an

answer to the research questions, the testimony, feedback, and associated

recommendations from the actual operators of AIASGT were obtained. The research

concentrated on and extrapolated policy’s that concentrated specifically on the

introduction of AIASGT. The researcher noted the perceptions concerning the use and

acceptance of these capabilities by the operators along with the permutations associated

with AIASGT development. The researcher compared the publicly published policies of

AIASGT insertion, perceptions, and the recommendation of process owners and

operators. The correlation between the developers of AIASGT and the perceptions of the

real world operators of AIASGT is necessary in order to formulate the appropriate

response to a rapidly expanding and permutating technology. The specificity concerning

the effects that AIASGT has on the operator and the correlation between the producers of

AIASGT is needed.

All aspects of AIASGT development needs to be quantified and exposed in order

to adequately regulate, detect, and prevent what is known as technology creep. Currently

the public has little access to the actual use, advancements, or development of AIASGT.

The results contained in Chapter 4 and an evaluation of this study findings were

supported by previous study findings (Arkin, 2011; Clarke, 2011; Cook, 2013; Davis,

1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013, Sparrow, 2012), and

autonomous self-governance (Arkin, 2011; Brock, 2010; Clarke, 2011; Cook, 2013;

Erlebach 2013; Geis 2011; Ghazizadeh et al., 2011; Phalp, 2008; Singer, 2013; Sleeman,

114

2008, Sparrow, 2012; Valavanis, 2007). Questionnaires were obtained from 10

participants from AL Post 346 were analyzed for the study.

The overall common theme was the support of the development and use of

AIASGT. Along with the willingness to use and acceptance of AIASGT, a common

theme that emerged was that no operator of AIASGT was ever consoled on their

willingness to accept the capabilities of AIASGT. Additional themes that emerged were

that operators of AIASGT felt that their opinion concerning the use or acceptance of

AIASGT capable platforms considered insignificant or irrelevant by developers.

115

Chapter 5: Implications, Recommendations, and Conclusions

The purpose of this qualitative study was to assess the willingness to use and

accept AIASGT by operators of the technology. The rapid development and assimilation

of AIASGT have necessitated the need to consult with the real world operators of the

technology. The necessity of acquisition policies of AIASGT that include operator

feedback and its effect on adoption, acceptance, assimilation, and use from the lens of

TAM which is a component of the theoretical perspective for this study. The problem is

the opinions of the personnel who have experience with the operation and development

of AIASGT have not been consulted concerning its actual employment. The purpose of

this study was to analyze the perspectives of experienced operators who can express and

articulate the usefulness of AIASGT and their attitude towards the use of the technology.

The rapid proliferation of autonomous, self-governing platforms and the secret nature of

their employment by the U.S. military have increased pressure to fully quantify and

delineate their capabilities (Singer, 2013). Autonomous self-governance has not been

identified as a technology that requires comprehensive policies and procedures to

mandate compliance within known and acceptable parameters (Arkin, 2011; Geis, 2011;

Singer 2013).

This chapter begins with a brief reiteration of the discussions concerning the

problem and purpose and the methodology used to codify this study. In this chapter I

reiterated the limitations, ethics, and overall challenges necessitated for data collection.

The implications of the findings were delineated to address the implications and impact

of AIASGT. In the final section, the findings are discussed in direct correlation with the

theoretical framework and the research questions. In this chapter, my research

116

expectations were explored and compared to existing literature and data. All questions

that were answered from the responses of survey participants were orchestrated to

extrapolate the necessary data to answer the research questions. All results extrapolated

from the findings of this research address the problem statement which is the impetus of

the study. The findings of this case study contribute to all present literature, and finally

culminate into recommendations for the continued research of AIASGT.

Since perceived usefulness is such a fundamental driver of usage intentions, it is

important to understand the determinants of this construct and how their influence

changes over time with increasing experience using the system (Venkatesh & Davis,

2000). Perceived ease of use, TAM’s other direct determinant of intention, has shown a

less consistent effect on intention across studies (Venkatesh & Davis, 2000). A better

understanding of the determinants of perceived usefulness would enable the designing of

organizational interventions that would increase user acceptance and usage of new

systems. Therefore, the goal of the present research is to extend TAM to include key

determinants of TAM’s perceived usefulness and usage intention constructs and to

understand how the effects of these determinants change with increasing user experience

overtime with the target system (Venkatesh & Davis, 2000).

Current TAM models do not include or consider the impact of AIASGT used for

the military or law enforcement purposes. The integration of AI systems that make

independent human-free decisions into the existing TAM has not been fully considered or

theorized (Arkin, 2011). The assimilation of AIASGT for military applications that will

ultimately take human lives as an aspect of the acceptance of technology would expand

the existing TAM models. The findings presented in this qualitative study utilized the

117

collected participant survey information extrapolated from similar case studies. Well

categorized and delineated information was assembled for the final study. The collected

data was comprised of the survey questions that were presented to the participants along

with the researcher’s observations. The study will be propagated and made available to

researchers to assist in the documentation and creation of knowledge of AIASGT. The

findings will be available to any researcher for transferability and knowledge sharing.

All research documented and meticulously delineated to optimize authenticity and

continuity of data collection. The data were collected from a purposeful sample of 10

personnel who have used AIASGT (Guest, Bunce, & Johnson, 2006). Purposeful

sampling will be used as it is best suited to construct multi-perspectival, emancipatory,

participatory, and deconstructive interpretations for qualitative data collection (Harsh,

2011), and gather a purposeful sample from the personnel associated with the AL Post

346.

Purposeful sampling from a group that is both knowledgeable and familiar with

AIASGT is the best methodology to employ for this study because this will be a specialized

case wherein in-depth understanding is being sought. Surveys are an effective tool that will

allow the researcher to gather all needed data for a study (Yin, 2013). The sample size is

predominantly a discretionary judgmental choice with the objective to detect a

measureable effect with the approach (Yin, 2013). According to Guest, Bunce, and

Johnson (2006), saturation is possible upon the completion of six surveys, therefore, a

purposeful sample size of 10 participants is considered optimal. However, it is noted that

should saturation not be reached after the completion of 10 surveys, participants recruited

and saturation reached. The sample size chosen was optimal to complete the study

allowing the opportunity to expand the sample size during future research. AL Post 346

118

sponsored the study and the researcher does not have any interrelationships or prior

contact with any of the participants of the study. All participants conducted the survey

and submitted the data for collection electronically. The anonymity afforded by

electronic data collection facilitated the ability of the participants to be open, honest and

confident that their privacy be maintained. This chapter delineates related permutations

inherent in the development and deployment of AIASGT. All findings are in direct

correlation with the research literature. All subsequent recommendations derived from

the study will improve future research and use of AIASGT by the transfer of operator

knowledge. Chapter five has presented the implications of the study related to usefulness

and actual utility (use) of AIASGT along with the impact on operators and how this study

relates to previous studies.

Implications

Determining the impact that autonomous systems will have on society has to be a

priority for developers (Sharkey, 2012). After researching acquisition policies, an

understanding of the permutations and rapid development of various types of AIASGT,

an overview of current governmental regulations designed to cope with the phenomenon

emerged. After reviewing relevant documentation, the researcher found that there were

challenges with the development of AIASGT which included social, political, economic,

and overall governance because of the rapid assimilation of AIASGT. The areas of

consideration were based on the theoretical framework for this study which is the

technology acceptance model (TAM; Davis, 1989), which is rooted in the TRA

developed by Ajzen and Fishbein (1973). The intent of the military is to assimilate

autonomous technology into its war making capabilities in an auspicious attempt to

119

mitigate friendly force casualties (Arkin, 2011). Both social influence processes

(subjective norm, voluntariness, and image) and cognitive instrumental processes (job

relevance, output quality, result demonstrability, and perceived ease of use) significantly

influenced user acceptance (Venkatesh & Davis, 2000). These findings were applied to

advance theory and contribute to the foundation for future research aimed at improving

understanding of user adoption behavior (Venkatesh & Davis, 2000).

The researcher for this case study collected data form participants that operated

AIASGT and were able to articulate their positions by delineating clear and concise

answers on the survey. There were challenges experienced during data collection

mitigated by utilizing the survey methodology of data collection. The participants were

willing to answer surveys because of the ease surveys presented. This case study consists

of a survey as the main data collection method as quantitative methods rely on objective

questionnaires as the research instrument (Yin, 2013). A case study offers rich, detailed

information on a particular phenomenon (Stake, 2008). Stake (2008) suggested a holistic

approach be used to answer the questions of how the collaborative system works, and

what effect collaborative communication can have on the relationship. The sample size is

predominantly a discretionary judgmental choice with the objective to detect a

measureable effect with the approach (Yin, 2013). According to Guest, Bunce, and

Johnson (2006), saturation is possible upon the completion of six surveys, therefore, a

purposeful sample size of 10 participants is considered optimal.

A second challenge was that the nature of the data collection process was internet

based and did not yield a large number of connected personnel. This challenge was

mitigated by collecting the required number of participants for the study despite the

120

challenges. Many known limitations with data collection from the participants were not a

factor because of the data collection methodology utilized. All of the data was collected

within days with the added convenience of the use of electronic data collection that

required no direct researcher presence during the process. All research participants were

well educated and were cognizant and comprehensive concerning the requirements and

willingness to participate in the study. There were some nuances of bias amongst the

participants that may be reflected in their unanimous avocation for the use of AIASGT.

All of the participants were encouraged to elaborate on their answers, but most of the

responses were short and precise. This may be because all of the participants who

possessed experience with AIASGT considered their involvement limited by

compartmentalization because of federal and industrial sensitive government civilian

background.

The researcher sought to explore the methods that the participants wanted

AIASGT utilized. With the postulation on the methodologies that AL personnel wanted

to employ AIASGT there may be an opportunity to improve the development and

regulation of the technology. By understanding the user’s perspective of the technology,

there may also be an opportunity for the developers of AIASGT to foster and develop a

better relationship with the users of the technology. With the involvement of the

AIASGT user community, human factors engineers can work closer with the community

and engineer the capabilities and agreed upon limitations into the technology. The

facilitation of collaboration between the developers of AIASGT and the users of the

technology would prove to be advantageous to the rapid development of the technology.

With all factors considered, the lack of participation concerning the development of

121

AIASGT was the most salient point of consternation of the participants. 10 participants

from AL Post 346 participated by answering a survey. All participants were exposed to

AIASGT and the average time that a participant was exposed to AIASGT was 4 years,

and there were personnel who participated in the study that have been exposed to

AIASGT for over 20 years. There were no physical challenges experienced by any of the

participants. All ethical procedures were adhered to with no compromise in

confidentiality. Yin (2013) noted that the case study consists of a survey as the main data

collection method as quantitative methods rely on objective questionnaires as the

research instrument. The data collected from the participant’s answers were in direct

correlation with previous study’s along with a comprehensive review of the literature.

All themes that developed from data collection have identified commonality of

suggestions from amongst the participant’s perspectives. All themes produced by this

study delineated in the literature review as considerations for the permutations inherent in

the continued development of AIASGT. This case study created by the researcher

allowed the exploration of the themes associated from the process owner’s perspective on

the useful ness and willingness to use AIASGT. The study research will expand the

body of knowledge of the Davis (1989) TAM and all of the various iterations of it that

did not specifically the acceptance of AI. TAM specifically related to intent to use needs

to be further expanded and tested to delineate the effect of the artificial, intelligent

thinking nature of autonomous platforms (Singer, 2013). The amount of exposure that

the average participant had with AIASGT was robust and adequately sufficient. The

absence of participatory communication amongst the developers of AIASGT and the

actual operators was a common theme. Interventions, such as user participation, peer

122

support, and management support, will be particularly relevant (Venkatesh & Bala,

2008).

Managers can decide on resource allocation for interventions based on the impact

of interventions on different determinants of IT adoption and type of systems (Venkatesh

& Bala, 2008). First, managers will now have a framework to decide what interventions

to apply during pre- and post-implementation stages and for what types of system, such

as for a complex system, perhaps interventions that will create favorable ease of use

perceptions will be relevant design characteristics, user participation, training, and peer

support; for a voluntary system, interventions that will influence the determinants of

perceived usefulness will be important to implement design characteristics, user

participation, incentive alignment, training, organizational and peer support; and for inter

organizational systems that affect organizational business processes or a customer

relationship management system that is critical to service delivery (Froehle, 2006). For

example, if design characteristics cannot be changed in a system, managers can allocate

more resources to training and user participation to make employees familiar with the

systems (Venkatesh & Bala, 2008). Davis (1989) postulated on what he identified as the

technology acceptance model (TAM) theory, and identified five elements for any

technology to be accepted overall. These five elements are (a) external factors, (b)

perceived usefulness, (c) perceived ease of use, (d) intention to use, and (e) attitude to

using (Davis, 1989). The current iterations of TAM, there are no specific elements that

consider AI, AIASGT, and its effect on the Davis (1989) acceptance model for perceived

usefulness and attitude to using.

123

The first research question was on how the operator of AIASGT perceived its

usefulness. The survey questions used were specifically worded to extrapolate themes

which in this case ultimately supported the research question. The survey questions that

supported research question one are questions one threw six of the survey. Consistent

with the intent of the developers of AIASGT, successful integration and assimilation of

the technology was accepted as beneficial to the recipients (Arkin, 2011; Clarke, 2011;

Cook, 2013; Davis, 1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013,

Sparrow, 2012). Although there was some consternation on who controlled AIASGT, the

participants perceived AIASGT as useful. In correlation with previous studies, the

contention is that there is not enough oversight concerning the development of AIASGT

(Arkin, 2011; Brock, 2010; Clarke, 2011; Cook, 2013; Erlebach 2013; Geis 2011;

Ghazizadeh et al., 2011; Phalp, 2008; Singer, 2013; Sleeman, 2008, Sparrow, 2012;

Valavanis, 2007). In a time when the public has shown concern for the deployment of

autonomous systems such as drones, the government has increased its use of these

technologies (Singer, 2013). These findings supported the theme for strict oversight of

the development and use of AIASGT. All respondents were given questions to determine

the saturation with AIASGT by soliciting responses from participants that would quantify

their exposure to AIASGT. Each respondent indicated that they experienced robust

exposure with AIASGT consisting of hands on or developmental involvement of the

technology. The participants indicated a very personal desire to utilize AIASGT. There

was some consternation amongst the participants on how AIASGT is currently developed

and subsequently deployed. The perception of being disconnected between the operator

and the developer of AIASGT was a constant theme. The developers of AIASGT

124

compartmentalize all aspects of the development of AIASGT. The consensus is that this

compartmentalization is to protect industrial security concerning the actual use, location,

and overall capabilities of AIASGT. The developers of AIASGT are under the asepsis

that all aspects of AIASGT have to be secured in this highly competitive environment.

This environment is perceived by the process owners in the field as a deficiency due to

the absence of robust operator feedback in the actual development and use of AIASGT.

Developers of AIASGT are concerned about the protection of intellectual

property causing receptacle feedback challenges between the operators of AIASGT and

its developers. It is the opinion of the operators of AIASGT that the developers monitor

the compartmentalization of every aspect of its development in order to keep positive

control of the technology. This allows for both the securing of industrial intellectual

property and technical advancements from competitors. While operator involvement

with the actual development of AIASGT is one of the best methods of designing the

desired capabilities into the technology, this type of involvement is limited to the ability

to find people with the desired real world field experience. This is exacerbated by the

reality that AIASGT is revolutionary and does not have a history of deployed

performance to adequately postulate the capabilities, constraints or limitations that would

be best practices to engineer into the technology. Operators noted that in the field that

there have been configuration mishaps along with equipment being fielded that is barely

operable. The poor condition of the equipment has been noted as a cause of damage to

AIASGT, properties, and civilian’s lives. “The context will be to do just the absolute

minimum needed to sustain the fight now, and accept the risks, while making fixes as you

go along,” said Colonel Mathewson of the USAF said (Drew, 2013, p. 12). These real

125

world operating conditions further disconnect and subsequent discontentment of the

operators of AIASGT in the field from the feedback loop to the acquisitions community.

Ethical concerns are further exacerbated when admittedly; the developers noted

that with all engineering challenges, there will be glitches in the hardware or software

(Drew, 2013). AIASGT are now being used for more than Intelligence, Reconnaissance,

or Surveillance (ISR) and is being developed and rapidly assimilated providing

revolutionary capabilities to the operators. The usefulness of AIASGT as with any

technology is dependent upon its accessibility and usefulness to the operator. AIASGT

utility accepted as a valuable asset by all respondents. The concerns on how AIASGT is

deployed were a constant theme that has to be mitigated for the perception of usefulness

of TAM to be a considerable factor.

The usefulness of the software that allows AIASGT to operate autonomously is

the foundation to its capabilities. The software development that makes AIASGT possible

is a complex undertaking.

Figure 2. Software Development Life Cycle

126

In software engineering, there are millions of lines of a programming code that

are written by a group of programmers. In the software development cycle, there were

many who created the base code for AIASGT enabled machines, producing ambiguity

when attempting to track a single developer (Epitropakis, 2012). In the software

development environment, there is no one person who will know the outcome of the

entire program.

Ethical concerns can be generated because of the fact that the possibility of faulty

or malicious code could make its way into AIASGT software. This faulty or malicious

code could allow hostile or unauthorized access of the AIASGT system (Lewis, 2013).

There is a misconception that AIASGT will do only as they are commanded or

programmed, which has proven to be a serious problem (Debusmann, 2009). The

importance of software development, its security, and the ability to eliminate malicious

code that can compromise its security is paramount (Drew, 2013).

The second research question was based on the operators’ attitude to use

AIASGT. The survey questions used were specifically worded to extrapolate themes,

which in this case ultimately supported the research question. The survey questions that

supported research question two are questions seven threw twelve. In correlation with the

findings of question ones determination of the usefulness, the developers of AIASGT and

the willingness to use, successful integration, and assimilation of the technology were

accepted as beneficial to the recipients (Arkin, 2011; Clarke, 2011; Cook, 2013; Davis,

1989; Geis, 2011; Kroeker, 2011; Mackey, 2013; Singer, 2013, Sparrow, 2012). The

findings were in direct correlation with the researcher’s expectations that the operator’s

attitude to using AIASGT was positively and universally accepted by all participants.

127

Since the development of AIASGT is highly compartmentalized, the participants

assimilated the equipment and the use of AIASGT as simply part of the duty position.

As in the first research question, there was some consternation on who controlled

AIASGT. In correlation with previous studies, the contention is that there is not enough

oversight concerning the development of AIASGT. A common theme was that all

participants considered development of AIASGT essential and highly

compartmentalized. The hands on operators of AIASGT described their participation in

providing any feedback concerning the assimilation of AIASGT as non-participatory.

Participants have a different opinion because of their experience and chance to reflect

upon the current and continued use of AIASGT. Although there was the lack of operator

input to developers, or regulatory constraints, all the participants’ attitude concerning the

utilization of AIASGT was universally accepted and positive. The Davis (1989) TAM is

not specific to the development of autonomous self-governing technology; however,

Venkatesh and Davis (2000) noted that TAM2 focused on a number of related criteria

that determine technology acceptance. There are no current models that specify the

consideration of AIASGT. The consideration of the changes that AIASGT will have on

the overall acceptance of technology would expand TAM , specifically the attitude to

using. The self-governance and ability of technology to make independent, artificially

intelligent decisions could affect the way technology is accepted.

Recommendations

The overall design of this study fit in well with the purpose of this research which

was to postulate the methods of determining the acceptance and use of AIASGT. The

128

problem is the opinions of AL personnel who have experience with the operation and

development of AIASGT have not been consulted concerning its actual employment.

The purpose of this qualitative case study is to explore the perceptions of AL members

who have developed or used some form of artificially intelligent self-governing

technology designed to take human lives. The findings were a synthesis of primary and

secondary resources. Based on the study’s objective of determining the acceptance and

use of AIASGT, the following recommendations are in direct correlation and congruent

with the findings.

The results and recommendations of this study may be used for the development

of AIASGT. The results of this study can be utilized for scholarly and practical

applications and as a foundation for subsequent research. This study was significant in a

multitude of ways. The primary contribution is to understand the user’s perspective of

utilizing the specific capabilities of AIASGT, and its acceptance and use by the operator.

The research of AIASGT contributed to all present literature on TAM. The findings of

this study may be utilized to further the relationship on autonomous capabilities and the

willingness to use AIASGT for both military and civilian applications. Firstly, the

findings of this study also contributed to the literature on the acceptance of AIASGT for

all applications to include both military and civilian applications. Secondly, every aspect

of AIASGT and how it would permutate the willingness to use and the perceived

usefulness of the technology can be added to collective collaboration with both the

developer and user communities. Thirdly, the findings of the study can positively

influence the future development and subsequent assimilation of AIASGT. The

developers of AIASGT may benefit from collaborative communion between the user and

129

developer communities. The developers of AIASGT may find that by utilizing some of

the recommendations from the findings that development of AIASGT will be influenced

positively with recommended ethical controls ever present.

This study contributed to the TAM theory framework that was the impetus of the

study. The postulation of the willingness to use and the perceived usefulness of AIASGT

and how it correlates into the present TAM was the basis of the study. This relationship

and the fostering of these relationships are the greatest concern and the foundation of the

user-developer communities. The problem concerning the ethics associated with the

programming that is imbedded in the technology as well as the ability of the technology

to differentiate the moral or ethical differences of any action is a problem that has been

identified as needing clarification (Geis; 2011). However, the needed technology that

will embed ethics into AI is not currently available (Arkin, 2011). By focusing on how

AIASGT was perceived by a controlled group from the AL, the researcher was able to

extrapolate form the data collected a well delineated and eclectic of the operators

experience with AIASGT and the interaction with developers.

This dissertation focused primarily on the actual hands on operator’s perception

of the usefulness and willingness to use AIASGT. The limited time associated with data

collection exacerbated by the infrequency of meetings at the AL, created few challenges

for the researcher. While the impetus of the study was to delineate the response data

from the participants, the researcher had a desire to reach out to the developers of

AIASGT in an attempt to foster the need for closer collaboration between both the user

and developer communities. This collaborative environment can manifest itself into

future research responses from all sides (user, regulatory and developer) optimizing

130

future development of AIASGT. By recording, delineating, and collaborating between

all communities, the analysis of feedback data could create effective and alternative ways

for the future development of AIASGT.

This study utilized theory such as TAM as a way to explain the relationship

between the user operator and the developer of AIASGT. Subsequent research could

utilize the theoretical base of this study intricately into the developmental relationship

between the industry developers of AIASGT and the operators along with additional

considerations considering the inevitable assimilation and maturation of AIASGT.

Original thoughts concerning acceptance of technology is parsimonious and includes

perceived usefulness, which has consistently been used to explain more variance in

technology acceptance than other variables (Ghazizadeh et al., 2011). Operators of

AIASGT as a norm will postulate that all aspects well thought out in reference to its

development. Another contribution to the acceptance of autonomous technology will be

when Lee (2009) expounded on the theory of planned behavior, perceived risk, and

perceived benefit to understand the acceptance of technology, such as autonomous

technology (Holden & Karsh, 2010). The users of AIASGT are concerned by ethical

controls. Howlader (2011) asked if an AIASGT machine could be considered to have

ethical or legal attributes and be held accountable for moral actions, or if ethics are pre-

programmed. Consternation concerning the assimilation and maturation of AIASGT

were control centric.

Subsequent research could incorporate one or more themes generated from the

data extrapolated from participants. The themes collected identified the desire for close

collaboration between the user and developer communities of AIASGT. This

131

collaboration could yield radical permutations in the current theory and assist in its future

development. The findings will assist future researchers in finding additional ways of

categorizing, developing, controlling, and optimizing the maturation of AIASGT. Based

on the study’s objective on determining the opinions of AL operators of the usefulness

and willingness to use AIASGT and the positive effect on its future development sparked

by collaboration, the additional recommendations are for consideration:

Conduct additional qualitative and quantitative analysis of collaborative

environments between the user, regulatory, and developer communities of AIASGT. Yin

(2009) noted that while the data collected by the researcher is valid, the only way to

establish reliability is to have the study replicated many times over. To compare the

themes observed by the researcher and to codify the validity of the findings, a subsequent

and similar study utilizing participants from another industry saturated with AIASGT

may be orchestrated for this purpose. The question of whether other operators of

AIASGT have similar elation or trepidation for the use may be explored with various

developers internationally.

Another approach to consider for the formulation of a qualitative study could be

to quantify the necessity for collaboration between the developers and the user

communities who operate AIASGT. One of the challenges of this study was that the

population selected for participation in the study was affiliated with the military in the

past; who are now civilians. If the researcher was able to find participants from multiple

industries and varying experiences, the expectation would be that robust data could have

been optimized for collection. The researcher would have been able to ensure that all

aspects of the current use and development of AIASGT were considered.

132

Additional challenges exist within the AIASGT user communities. The addition

of additional studies that would complement this effort, such as an AIASGT regulatory

commission’s effort, could create a model for the successful development of AIASGT. A

robust plethora of qualitative and quantitative analysis between AIASGT developers and

users other than the AL would assist in its maturation and acceptance. While the data

reviewed by the researcher is valid, the only way to stablish reliability is to have the

study replicated many times over (Yin, 2009). A different AIASGT community can be a

viable option to collect data from. Another approach to a similar study could be to

determine what degree of participation between the user community and the developers is

a factor for the future development of AIASGT. An additional approach to the

quantification on the degree that collaboration between the user and developer has on the

overall development of AIASGT.

The consideration of additional theories used to explain the use and acceptance of

AIASGT can be utilized to formulate a similar study. By utilizing a variation of the

accepted theoretical base that is used to delineate the current development of AIASGT,

various perspectives can be represented. This research can optimize the successful

development and assimilation of AIASGT internationally. Given that IT is becoming

increasingly complex and pertinent to employees’ decision making and work processes,

this research has implications for broad IT-enabled organizational decision making (e.g.,

collaborative forecasting, inventory management, replenishment, service delivery;

Venkatesh & Bala, 2008). A future study could be based on the developer and user

collaboration with regard to the effectiveness of the effort. All participants mentioned the

desire to collaborate and influence the development of AIASGT. The desire for

133

transparency and the ability to influence the development of AIASGT was the

overwhelming theme extrapolated from the data. A detailed study on the involvement of

the user community and collaboration of the developers of AIASGT may assist in the

contribution to the literature of the use and usefulness of AIASGT and provide a method

of positively influencing its technological maturation and assimilation.

A greater comprehension was obtained as to the satisfaction experienced and the

effectiveness on the collaboration between the user and developer communities. The

result of the research has shown that there is a positive correlation between the successful

use and perceived usefulness of AIASGT based on the degree of collaboration.

Trepidation concerning the introduction of AIASGT can be mitigated with the robust

involvement between impartial users and the developers. One method that will optimize

collaboration between the developer and user communities is to create effective

regulations and oversight of AIASGT development. Subsequent research could be used to

determine if the current development methodologies are affected by the feedback of the

user communities. If both the developer and user communities have the ability to

influence the development of AIASGT, then the developers will be satisfied with the

acceptance of the technology. The users of the technology will be similarly satisfied

allowing for the acceptance of the technology. This collaborative successful outcome

will aid in its acceptance, propagation increasing use and sales.

Future research could benchmark industry best practices ensuring that

communication between regulators, users, and developers are facilitating collaboration.

The themes and data collected from the various entities may produce additional

information as to whether collaboration between the user and developer has an influence

134

on perceived use of AIASGT. If the development of AIASGT can obtain transparency

and the perceived use and usefulness of AIASGT can be collaboratively developed, any

perceived challenges can be effectively mitigated. One challenge that the researcher

experienced was the lack of the preservation of human capital involved in the current

development of AIASGT. There were no participants in the study that were involved in

the operation of AIASGT that had any involvement in its development.

Past studies have explored the theoretical framework for this study which is the

Technology Acceptance Model (TAM; Davis, 1989), which is rooted in the theory of

reasoned action (TRA) developed by Ajzen and Fishbein (1973). Davis (1989)

postulated on what he identified as the Technology Acceptance Model (TAM), and

identified five elements for any technology to be accepted overall. The emphasis of this

study was to study perceived usefulness and attitude to using AIASGT. Original

thoughts concerning acceptance of technology is parsimonious and includes perceived

usefulness, which has consistently been used to explain more variance in technology

acceptance than other variables (Ghazizadeh et al., 2011). The respondents to the study

were in correlation to accepting all aspects of AIASGT. All expressed a high degree of

confidence that AIASGT is currently regulated and assimilated into the user

communities. The positive responses from the participants expressed confidence in the

current efforts of development and maturation of AIASGT. All respondents expressed

optimism and value added associated with the expanding use of AIASGT. The

respondents were disappointed, confused, and frustrated that the developers of AIASGT

would miss the opportunity of collecting valuable data from the user community. The

desire to collaborate and have influence with the decision making process for the

135

development of AIASGT was evident. A transparent collaborative, reciprocal

environment would mitigate any trepidation associated with AIASGT.

Further research is needed on organizations that have regulatory control over

AIASGT. The findings may assist developers collaborate with users to increase the

perceived usefulness and the willingness to use AIASGT. Respondents believe that

developers of AIASGT are utilizing industry best practices. What would happen if the

developers of AIASGT were to use the operators of the technology more effectively?

The notion that operators of AIASGT considered their input to its development beneficial

may affect the acceptance and pace of technology assimilation than initially thought.

Additional research may determine if the findings of the current research is a singularity

or a reoccurrence amongst operators of AIASGT.

Conclusion

Davis (1989) postulated on what he identified as the Technology Acceptance

Model (TAM), and identified elements for any technology to be accepted overall. The

problem is that when inserting the element of AIASGT into TAM, the permutations that

AIASGT causes will possibly alter all of the known factors identified and accepted as

being governed by the current TAM. In particular, AIASGT will probably alter the intent

to use as well as the attitude of using in unpredictable ways. The intent to use will

possibly be permutated by the unknown variables introduced by the self-governing

properties of AIASGT. The main finding was that there is a deficiency in collaboration

between the users and developers of AIASGT. AIASGT is given to the operator as a

finished working technology often without operator input. The operator of AIASGT has

136

to believe in the technology for them to perceive the usefulness and have a positive

attitude to using AIASGT.

It is in the best interest of the developers of AIASGT to incorporate the feedback

from the operators. This along with accurate projections and controls of the propagation

of AIASGT will help to maturate the technology and mitigate any trepidation related to

its rapid assimilation amongst the user community. The use of autonomous systems is on

the rise, especially with law enforcement and for agricultural applications (Singer, 2013).

There is a deprivation of collaboration between the developers of AIASGT and the user

community. The underlying motivations of the community’s may not be totally

understood or proportional to the degree necessitated to properly influence the

development of AIASGT. Robust collaboration between the user and developmental

communities will be beneficial to every aspect of future AIASGT development. Current

regulations of AIASGT are incorporating collaboration between stakeholders to

formulate comprehensive policies. The rationale of current rules and policies rapidly

permutate in correlation with the maturation of AIASGT. The current theory’s and

subsequent policies that govern TAM are not in pace with the maturation and

assimilation of AIASGT.

The current governorship of AIASGT has been described as not clearly defined

and subject to confusion. Because of the pace of AIASGT development, it is a challenge

to observe the impact of AIASGT. Current measures of monitoring, regulation, and

control are imperfect and may lead to imperfect conclusions. What is described as stove

pipe solutions to AIASGT are increasing because developers are concentrating their

efforts on advancing and proliferation of technology. These efforts directly correlate to

137

the advancement of the technology, and profit margins by the proliferation of the

capability. Without effective regulation of AIASGT, conflict may exist between the user

and developer communities. Past research has proposed that TAM be expanded in a

multitude of ways. The main reason for this is that the original TAM had to be updated

to keep in correlation with the capabilities of technological advancements. A multitude

of factors are present in a dynamic user developer collaborative environment. As

examined threw this study of AIASGT, there are dynamic factors that can positively

impact the development and maturation of AIASGT. Collaboration between all decision

makers can mitigate ambiguity in the regulations necessary to develop AIASGT and

positively impact its development.

Including the perspectives of the user communities of AIASGT exposed multiple

additional factors of consideration that can impact its development. This will enable all

stakeholders with additional information that will assist in their evaluation, development,

and regulation of AIASGT. The themes that emerged from the participants can serve to

shape higher level theory directly impacting the regulation and maturation of AIASGT.

Subsequent future research should include other stakeholder groups, specifically

communities that operate AIASGT not affiliated with the AL.

138

References

Adams, J. (2010). US defends unmanned drone attacks after harsh UN Report. Christian

Science Monitor. Retrieved from http://www.csmonitor.com/

Arkin, R. (2011). Governing lethal behavior: Embedding ethics in a hybrid

deliberative/reactive robot architecture. IEEE Society on Social Implications of

Technology, 30(4), 7-11. doi:10.1109/MTS.2011.943307

Asimov, I. (1986). Robot dreams. New York, NY: Ace Books.

Bagozzi, R. (2011). Measurement and meaning in information systems and

organizational research: Methodological and philosophical foundations. MIS

Quarterly, 35(2), 261-292. Retrieved from http://www.misq.org/

Baxter, P. (2008). Qualitative case study methodology: Study design and

implementation for novice researchers. The Qualitative Report, 13 (4), 544-559.

Retrieved from http://www.nova.edu/ssss/QR/QR13-4/baxter.pdf

Botvich, D. (2010). BiRSM: bio-inspired resource self-management for all IP-networks.

Network, IEEE, 24(4), 20-25. doi:10.1109/MENT.2010.5464223

Brock, D. (2010). The uncertain future of Moore’s law. Journal of Chemical Education,

84(8), 1278. doi:10.1021/ED084p1278

Beaudry, A., & Pinsonneault, A. (2005). Understanding user responses to information

technology: A coping model of user adaptation. MIS Quarterly, 29, 493–525.

Retrieved from http://www.misq.org/

Carter, A. (2012). Autonomy in weapon systems. Department of defense directive

3000.09. Retrieved from https://www.fcsace.com/Windchill/servlet/

WindchillAuthGW/ wt.enterprise.URLProcessor/URLTemplateAction?

action=ObjProps&oid =VR%3 Acom.boeing.ids.doc.

FCSProgramDocument%3A57918465&u8=1

Clarke, J. (2011). Identification and control. International Journal of Modeling,

12(3), 208-209. doi:10.1504/IJMIC.2011.039699

Cook, D. (2013). Ambient intelligence: Technologies, applications, and opportunities.

Journal Elsevier, 6(2), 277-298. doi:10.1016/j.pmcj.2013.04.001

Corbin, J., & Strauss, A. (2008). Basics of qualitative research: Techniques and procedures for developing grounded theory (3rd ed.). Thousand Oaks, CA: Sage.

Crispin, B. (2008). What killed the robot soldier? Retrieved from

http://www.strangehorizons.com/2008/20081110/crispin-a.shtml

139

Darley, W. (2005). Clausewitz’s theory of war and information operations. Retrieved

from http://www.au.af.mil/au/awc/awcgate/jfq/4015.pdf

Davies, S. (2013). It’s war but not as we know it. Engineering and Technology, 4(9), 40-

43. doi:10.1049/et.2013.0907

Davis, F. (1989). Perceived usefulness, perceived ease of use, and user acceptance of

information technology. MIS Quarterly, 13(3), 319-340. Retrieved from

http://www.misq.org/

Debusmann, B. (2009, April). Killer robots and a revolution in warfare. Retrieved from

http://www.reuters.com/article/2009/04/22/idUSLM674603

Desjarlais, O. F. (2008). UAS – Unseen, unheard, unstoppable. Airman Magazine,

September/October 2008, 4-11. Retrieved from http://airman.dodlive.mil/

Desjarlais, O. F. (2008). UAS – Unseen, unheard, unstoppable. Airman Magazine,

September/October 2008, 4-11. Retrieved from http://airman.dodlive.mil/

Drew, C. (2013, March 17). Drones are weapons of choice in fighting Al Qaeda.

Retrieved from http://www.nytimes.com/2013/03/17/business/17uav.html?

sq=drones&st=cse&scp=1&pagewanted=all

Donley, M. (2010, May). Technology horizons: A vision for Air Force science &

technology. Retrieved from http://www.af.mil/shared/media/document/

AFD-101130-062.pdf

Elgabalawi, N. (2012). Software integration and test. Retrieved from

https://www.fcsace .com/Windchill/servlet/WindchillAuthGW/

wt.content.ContentHttp/viewContent/3762448.doc?u8&HttpOperationItem=wt.co

ntent.ApplicationData%3A8905455&ContentHolder=wt.doc.WTDocument%3A8

905452

Ellul, J. (1962, Autumn). The technical order, technology and culture. Proceedings of the

Encyclopedia Britannica Conference on the Technological Order, 3, 394-421.

Retrieved from: http://ellul.org/wp-content/uploads/2015/08/Elluls-1962 -

Article1.pdf

Flick, U. (2009). An introduction to qualitative research: Fourth edition. London: Sage.

Froehle, C. (2006). Service personnel, technology, and their interaction in influencing

customer satisfaction. Decision Sciences, 37, 5–38. http://dx.doi.org/10.1111/

j.1540-5414.2006.00108.x

140

Gattone, S. (2011). Adaptive cluster sampling with a data driven stopping rule.

Statistical Methods & Applications, 20(1), 1-21. doi:10.1007/s10260-010-0149-5

Geis, J. (2011). Blue horizons study assesses future capabilities and technologies for the

United States air force. Interfaces, 41(4), 342. doi:10.1287/inte.1110.0556

Gorenc, S. (2006, April). Air Force committed to unmanned vehicles. Retrieved from

http://www.af.mil/news/story.asp?storyID=123018987

Guest, G., Bunce, A., & Johnson, L. (2006). How many surveys are enough? An

experiment with data saturation and variability. Field Methods, 18(1), 59-82.

http://dx.doi.org/10.1177/1525822X05279903

Haidt, J. (2003). The moral emotions. In R. Davidson, K. R. Sherer, & H. H. Goldsmith

(Eds.), Handbook of affective sciences (pp. 852-870). Oxford: Oxford University

Press.

Jones, T. (2008). Artificial intelligence: A systems approach. Sudbury, MA: Jones &

Bartlett.

Klinger, J. (2006). The social science of Carl von Clausewitz. Retrieved from

http://www.carlisle.army.mil/USAWC/parameters/Articles/06spring/klinger.pdf

Krippendorff, K. (2012). Content Analysis: An introduction to its methodology. Thousand

Oaks, CA: Sage Publications.

Lee, M. (2009). Factors influencing the adoption of internet banking: An integration of

TAM and TPB with perceived risk and perceived benefit, Electronic Commerce

Research and Applications, 8(3), 130-141. Retrieved from

http://www.sciencedirect.com/science/article/pii/S1567422308000598

Lensink, R. (2011). Peeters Online Journal. Ethical Perspectives. 18(4), DOI:

10.2143/EP.18.4.2141848http://poj.peeters -leuven.be/content.php

?id=2141848&url=article&download=yes

Lewis, R. (2013). Robots must learn to obey a warrior code. Retrieved from

http://technology.timesonline.co.uk/tol/news/tech_and_web/article5741334.ece

Mackey, J. (2013). Recent US and Chinese Antisatellite Activities. Air and Space Power

Journal, 23(3), 82–93. Retrieved from http://www.airpower.maxwell.af.mil/ airchronicles/apj/apj09/fal09/mackey.html

Manfreda, K. (2008). Web surveys versus other survey modes. International Journal of

Market Research, 50 (1), 80. Retrieved from https://www.mrs.org.uk/ijmr

McCarthy, J. (2007). What is artificial intelligence? [Electronic version]. Retrieved from

Stanford University, computer science department website:

141

http://www-formal.stanford.edu/jmc/whatisai.pdf

Merriam, S. (2009). Qualitative research: A guide to design and implementation. San

Francisco, CA: Jossey-Bass.

Phalp, K. (2008). Particle swarm guidance system for autonomous, unmanned aerial

vehicles in an air defense role. Journal of Navigation, 61, 9-21. http://dx.doi.org/

10.1017/S0373463307004444

Popkin, G. (2015). Moore’s law is about to get weird, never mind tablet computers. Wait

till you see bubbles and slime mold. Nautilus, 21. Retrieved from http://nautil.us

/issue /21/information/moores-law-is-about-to-get-weird

Saldana, J. (2009). The coding manual for qualitative researchers. London: Sage.

Sharkey, N. (2008). Cassandra or false prophet of doom: AI robots and war. Retrieved

from IEEE Computer Society database:

http://search3.computer.org.ezproxy.umuc.edu/search/results

Sharkey, N. (2012). Drone race will ultimately lead to sanitized slaughter. Retrieved

from: http://www.theguardian.com/world/2012/aug/03/

drone-race-factory-slaughter

Sharkey, N. (2012). The evitability of autonomous robot warfare. International review of

the red cross, 94, 866. doi:10.1017/S1816383112000732

Singer, P. (2013). The global swarm. Foreign Policy. Retrieved from

http://www.foreignpolicy.com/articles/2013/03/11/the_global_swarm

Sinkovics, R., Penz, E., & Ghauri, P. (2010). Analyzing textual data in international

marketing research. Qualitative Market Research: An International Journal, 8(1),

9-38. doi:10.1108/13522750510575426

Sleeman, R. (2008). Systems approach to unmanned air vehicle development and

certification. Improvements in Systems Technology, 6, 209-221. doi:10.1007/

978-1-84800-100-8_13

Sparrow, R. (2012). Just say no to drones. IEEE Technology and Society, 31(1) 56-63.

doi:10.1009/MTS.2012.2185275

Stake, R. (2008). Qualitative case studies. In N. K. Denzin, Y. S. Lincoln (Eds.),

Strategies of qualitative inquiry (3rd ed.). Thousand Oaks, CA: Sage.

Taiwo, A., & Downe, A. (2013). The theory of user acceptance and use of technology

(UTAUT): A Meta-analytic review of empirical findings. Journal of Theoretical

and Applied Information Technology, 49(1), 48-58. Retrieved from

http://www.jatit.org

142

The AL. (2015). History. Retrieved from http://www.legion.org/history

Valavanis, K. (2007). Advances in unmanned aerial vehicles: State of the art and the

road to autonomy. New York, NY: Springer.

Venkatesh, V. (1999). Creating favorable user perceptions: Exploring the role of

intrinsic motivation. MIS Quarterly, 23, 239–260. Retrieved from

http://www.vvenkatesh.us/Downloads/Papers/fulltext/pdf/

2000(2)_MS_Venkatesh_Davis.pdf

Venkatesh, V. (2013). User acceptance of information technology. MIS Quarterly, 36 (1),

157-178. Retrieved from http://www.vvenkatesh.com/Downloads

/Papers/fulltext/pdf /Venkatesh_Thong_Xu_MISQ_forthcoming.pdf

Venkatesh, V., Morris, M., Davis, G., & Davis, F. (2003). User acceptance of

information technology: Toward a unified view. MIS Quarterly, 27(3), 425-478.

Retrieved from http://www.vvenkatesh.us/Downloads/Papers/fulltext/

pdf/2000(2)_MS_Venkatesh_Davis.pdf

Venkatesh, V., & Davis, F. (2000). A theoretical extension of the technology acceptance

model: Four longitudinal field studies. MIS Quarterly, 46(2), 186-204. Retrieved

from http://www.vvenkatesh.us/Downloads/Papers/fulltext/pdf/

2000(2)_MS_Venkatesh_Davis.pdf

Venkatesh, V., & Bala, H. (2008). Technology acceptance model 3 and a research agenda

on interventions. Decision Sciences, 39 (2), 273-315. Retrieved from

http://www.vvenkatesh.com/downloads/papers/fulltext/pdf/

venkatesh_bala_ds_2008.pdf

Winner, L. (1978). Autonomous technology: Techniques-out-of-control as a theme in

political thought. Cambridge, MA: MIT.

Yin, R. (2013). Case study research: Design and methods (5th ed.). Thousand Oaks, CA:

Sage.

143

Appendixes

144

Appendix A: Informed Consent Form

145

Appendix B: Screening Demographic Questionnaire

Background: The military is continuously modernizing and increasing the use of

Artificially Intelligent Autonomous Self-Governing Technology (AIASGT) into combat

units. This technology is assimilated into the combat units without the consideration of

acceptance, approval, disapproval, or willingness to use by the operator.

Have you ever had any direct experience in using AIASGT for the military? _______

Do you feel qualified to give an opinion on the future or continued use of

AIASGT?_______

Demographic Questions:

1. Which of the following best represents your main area of work in support of AIASGT?

_______Instructor _______Management/Supervisor

_______Operator _______Technician

_______Hardware Engineer _______Software Engineer

_______Student _______Retired _______Other:__________________________

2. Indicate your highest completed level of education from the choices below.

_______Diploma _______Associate’s Degree

_______Bachelor’s Degree _______Master’s Degree

_______Doctoral Degree

3. What is your gender? _______Male _______Female

146

Appendix C: Survey Questionnaire

Thank you for participating in the survey concerning the use of Artificially

Intelligent Autonomous Self-Governing Technology (AIASGT). Once you

have completed the questionnaire, please email it back to

[email protected]

147

Appendix D: AIASGT Acceptance and Use Survey

1. Would you willingly use AIASGT on the battlefield?

2. Would you willingly accept the capabilities of AIASGT against others in

combat?

3. Do you think that enough transparency and regulations are in place for

you to accept AIASGT for military applications?

4. Do you think that enough transparency and regulations in place for the

ethical use of AIASGT for military applications?

5. Would you advocate for the use AIASGT for the military?

6. Were you ever consulted on your willingness to accept the military

capabilities of AIASGT?

7. Do you think that there are enough regulations of AIASGT for civilian

applications?

8. Do you think that there is enough information available to accept the

capabilities of AIASGT for civilian applications?

9. Do you think that enough transparency and regulations are in place for

you to accept AIASGT for civilian applications?

10. Would you advocate for the use AIASGT for the civilian applications?

11. Would you advocate for strict regulation concerning the use of

AIASGT.

12. Do you have any concerns regarding the regulation or future

development of AIASGT?

148

Appendix E: Survey Participation Advertisement

An invitation to participate in the study will be placed on the AL Post 346 website and

monthly paper.

Please participate in the survey concerning the use of Artificially Intelligent Autonomous

Self-Governing Technology (AIASGT). Please email Robert Carmack at

[email protected]. An email will be sent back to you with a link to access the survey

questions. You can access the survey directly at:

https://drive.google.com/folderview?id=0B2W5Zu97xjCxcUZqSGpOXy1ISGs&usp=sha

ring. The link will also be available on the AL Post 346 website

http://www.al346neptune.org/.

Once you have completed the questionnaire, please email it back to [email protected]