Essay 11

ch11.pdf

Home >Computer Science homework help >Essay 11

Online Communities, Virtual Reality, and Artificial Intelligence

 This chapter examines ethical and social issues affecting three diverse (and seemingly unrelated) aspects of cybertechnology:

1) online communities, including social networking services (SNSs);

2) virtual environments (VEs), including virtual reality (VR) applications;

3) artificial intelligence (AI).

 A unifying theme that brings together this disparate cluster of topics is the impact they have for our notions of community and (personal) identity in the digital age.

Online Communities, Virtual Reality, and Artificial Intelligence (Continued)

 The two main themes, community and (personal) identity, are analyzed via the three distinct aspects of cybertechnology in the following ways:

a) SNSs enable social interactions that challenge our traditional notion of community;

b) some VEs and VR applications allow users to construct new and alternate (personal) identities;

c) AI-related developments invite us to reassess our sense of self and, ultimately, question what it means to be human in a world that we share with non- human “intelligent” agents and entities.

1. Communities in Cyberspace

 What is a Community?

 Webster's New World Dictionary of the American Language defines a community as “people living in the same district, city, etc., under the same laws.”

 Note that this (traditional) definition of community – with its emphasis on people living in the “same district” or “same city,” etc. – stresses the importance of geography and physical space.

 But our traditional conception of community has evolved significantly because of our interactions in online forums and social networking sites (SNSs) such as Facebook.

Online Communities vs. Traditional Communities

 Rheingold (2001) defines online communities as “computer-mediated social groups.”

 He describes what it was like when he joined the WELL (Whole Earth ‘Lectronic Link), an early online community, in 1985.

 Rheingold believes that the WELL became a community because of the social contracts and collaborative negotiations that happened when members met online.

Online Communities (Continued)

 Rheingold points out that in the WELL, norms “were established, challenged, changed, re- established, etc., in a kind of “speeded up social evolution.”

 He also notes that when WELL members decided to get together occasionally at physical locations in the greater San Francisco Bay area, the WELL became a hybrid community because it then spanned both physical space and virtual space.

Forming Online Communities

 In the past, many people tended to identify themselves as members of a community by way of categories such as their national heritage, religious affiliation, etc.

 Today, common interests, as opposed to traditional categories, often bring people together to form online communities.

Forming Online Communities (Continued)

 Many people, both young and old, now interact in Web- based social networking sites (SNSs), such as Facebook and Twitter, as well as in professional-oriented networking sites such as LinkedIn, where they:

 send instantaneous “news feeds” to friends in the forms of “tweets” via a popular online service called Twitter.

 participate in one or more blogs (web logs).

 communicate through electronic messenger services provided by Web-based organizations such as America Online’s Instant Messaging (AIM), or Voice Over Internet Protocol and video messaging services such as Skype and Apple’s Facetime.

Satisfaction for Those in Online vs. Traditional Communities

 Do people find as much enjoyment and satisfaction in participating in online communities as they do in traditional ones?

 Parsell (2008) cites a survey showing that 43% of members of online communities claimed to feel “as strong” about their online communities as their traditional or “real world” communities.

Satisfaction in Online Communities (Continued)

 Parsell also believes that this satisfaction may be due to the enhanced nature of the Web – what some now refer to as “Web 2.0.”

 Web 2.0 is very different from the early Web (i.e., “Web 1.0”), primarily because of the interactions it makes possible in social networking services (SNSs).

 Analysts disagree on exactly which criteria differentiate Web 2.0 from Web 1.0.

 But most agree that the kinds of services made possible by SNSs and blogging sites have significantly altered the way users interact on the Web.

Blogs as Online Communities

 A very popular mode of online communication for is a forum called the blog (or Web log).

 What, exactly, is a blog?

 According to the (online) Merriam Webster Dictionary, a blog is “a Web site that contains an online personal journal with reflections, comments, and often hyperlinks provided by the writer.”

Blogs as Online Communities (Continued)

 How do blogs function as online communities?

 While some blogs function as online diaries (i.e., personal blogs), others provide commentary on a particular topic or news story and are categorized as:

 political blogs,

 corporate blogs,

 health blogs,

 literary blogs,

 travel blogs.

Blogs as Online Communities (Continued)

 Review Scenario 11-2 in the textbook, which describes some controversies affecting Jessica Cutler’s personal blog (the Washingtonienne case).

 Among the controversies raised in Cutler’s (now infamous) blog were questions about:

 privacy,

 anonymity,

 defamation.

 Recall also our discussion of some controversial aspects of political blogs in Chapter 10.

Pros and Cons of Online Commu- nities: Identifying some Pros

 Many see online communities as positive, because people can:

 meet new friends and future romantic partners;

 form medical support groups by joining Internet- based social networking groups;

 join discussion forums to disseminate material to like-minded colleagues;

 communicate with people they might not otherwise communicate with by physical mail or telephone.

Pros and Cons of Online Commu- nities: Identifying some Cons

 Many also believe that online communities have had some negative effects because they threaten traditional community life and minimize face-to-face communications.

 Other critics point out these communities are problematic because they have:

 facilitated social polarization,

 threatened our traditional notion of friendships,

 facilitated deception.

 We examine each of these criticisms.

Online Communities and Social Polarization

 Online communities, whose focus tends to be on topics and issues that are divisive and narrow, easily contribute to social polarization.

 For example, Parsell (2008) argues that “extremely narrowly focused” online communities can be dangerous because they “can polarize attitudes and prejudices,” which can lead to increased social division.

Social Polarization (Continued)

 Parsell worries that the “narrow focus” of many online communities presents us with cause for concern, which he articulates in the following argument:

1. People tend to be attracted to others with like opinions.

2. Being exposed to like opinions tends to increase our own prejudices.

3. This polarizing of attitudes can occur on socially significant issues…

4. Where the possibility of narrowing focus on socially significant issues is available, increased community fracture is likely.

Online Communities and the Impact for Friendships

 To what extent, if any, is physical interaction between individuals necessary for true friendships to develop and flourish?

 At one time, the notion of “disembodied friends” might have seemed strange.

 But today, we hear about so-called “friends” who communicate regularly online but who may have never met in physical space.

Friendships in Online-Only Communities

 What implications do online-only communications between individuals have for our traditional understanding of friendship.

 Is it possible for people who interact only in virtual (or purely online) contexts to be “real friends”?

Friendships in Online-Only Communities (Continued)

 Cocking and Matthews (2000) argue that the kinds of close friendships we enjoy in physical space are not possible in pure virtual environments (i.e., in contexts that are solely computer-mediated).

 They note that online-only friendship occurs in “a context of communication dominated by voluntary self disclosure, enabling and disposing me to construct a highly chosen and controlled self-presentation.”

 As a result, Cocking and Matthews believe that we “miss the kind of interaction between friends that seems a striking and commonplace feature of a close friendship.”

Friendships in Online-Only Communities (Continued)

 Cocking and Matthews also argue that in off-line contexts, we involuntarily disclose aspects of ourselves.

 We do this through indicators or “cues” in our in- teractions with others because of “non-voluntary self disclosure,” which means that one has less control over the way he or she presents oneself to others.

 So, Cocking and Matthews conclude that important aspects of our true personalities are involuntarily revealed, which makes close friendships possible in off-line contexts but not in virtual ones.

Online Friendships (Continued)

 Briggle (2009) disagrees with Cocking and Matthews.

 Briggle considers factors such as “sincerity” and “distance” in making his case for why friendships in purely virtual contexts can be “initiated and flourish.”

 He points out that communications among friends in off- line contexts, which are based largely on “oral exchanges,” are not always candid or sincere.

 Briggle also notes that the distance involved in typical computer-mediated communications can give friends the courage to be more candid with one another than in typical face-to-face interactions.

Online-Only Friendships vs. Face- to-Face Interactions

 Does it matter that online communication has minimized the kinds of face-to-face interactions that define traditional friendships?

 Is that necessarily a negative thing?

 On the one hand, being able to send an email or a text message, or to write on a Facebook friend’s “wall” or “time-line,” is far more convenient than having to meet face to face in physical space to communicate with them.

 On the other hand, some worry that critical interpersonal skills may decline, or be lost altogether, in excessive on- line communications that avoid face-to-face interactions.

Online-Only Friendships vs. Face-to-Face Interactions (Continued)

 Examine Scenario 11-1 in the textbook, involving “Ralph.”

 On the one hand, Ralph finds it very convenient to communicate only with his online or virtual “friends.”

 What impact could this behavior have for Ralph if at some future point, he decides that he would enjoy face-to-face interaction with one of those friends?

 For example, Will Ralph’s interpersonal skills be sufficiently diminished, as a result of his prolonged (and exclusive) online-only interactions?

Online Communities and Deception

 Online communities can also reveal a “darker side” of the Internet because people can, under the shield of anonymity, engage in behavior that would not be tolerated in most physical communities.

 For example, individuals can use aliases and screen names when they interact in online forums, which makes it easier to deceive about who actually is communicating with them.

 Review Scenario 11-3 in the textbook, which des- cribes a cyberbullying incident on MySpace resulting in a teenager’s suicide (the Megan Meier case).

Online Communities and Deception (Continued)

 A (now) classic case of deception in an online community involved an incident that has come to be described by some as a “virtual rape in cyberspace.”

 Users interacting (anonymously) in an online forum called LambdaMOO used “screen names” to represent the virtual characters they portrayed in that forum.

 One user programed his character to “perform” sexually offensive actions on two of the virtual characters that also interacted in the LambdaMOO forum (see Dibbell, 2001).

 For an interesting analysis of incident involving this “virtual rape,” see Søraker (2010).

2. Virtual Environments and Virtual Reality

 We should note that online communities are also often referred to as virtual communities.

 But online communities comprise only a subset of virtual environments (or VEs).

 These environments are “virtual” in the sense that they exist in cyberspace, which is usually equated with virtual space.

 But what, exactly, is meant by “virtual”?

Defining “Virtual” in Virtual Environments

 At least three different senses of “virtual” can be distinguished, because the term virtual can be:

1) contrasted with “real,” as in cases where virtual objects are differentiated from “real objects”;

2) contrasted with “actual” – e.g., where a person claims that she is “virtually finished” her project;

3) used to express a feeling that one has “as if” he or she were physically present – i.e., the feeling of presence that one can have in a telephone conversation or an online messaging exchange.

Virtual Environments (Continued)

 As already noted, VEs exist in virtual space.

 What is meant by “virtual space”?

 Virtual space can be viewed as a range of computer-generated environments that could not exist without computers and cybertechnology.

 These environments, in turn, include online (or virtual) communities, as well as (three- dimensional) virtual reality (VR) applications.

Virtual Environments (Continued)

 We have already noted that virtual environments (VEs) are both similar to and different from the kinds of online/virtual communities described earlier.

 While both kinds of virtual forums are “computer- generated,” and therefore could not exist without cybertechnology, VEs also provide contexts in which users can do more than merely interact with other users.

 As Søraker and Brey (2015) point out, VEs can “visualize imaginary environments” as well as simulate real ones.

VEs and Virtual Worlds

 Not only do VEs subsume virtual (or online) communities, they also subsume (what some authors call) virtual worlds.

 What are virtual worlds, and how are they similar to and different from VEs.

 Søraker and Brey describe virtual worlds as a type of VE in which users typically:

a) are “represented by avatars”;

b) have the “illusion of perceiving a three dimensional world consisting of virtual objects.”

VEs and Virtual Worlds (Continued)

 Like Søraker and Brey, Wankel and Malleck (2010) also describe virtual worlds as environments that are “three dimensional.”

 But we have already noted that for our purposes, a VE need not have a three-dimensional aspect.

 We have also seen that VEs can include online communities that are two-dimensional environments, while VR applications are always three dimensional.

 Our focus will be mainly VR applications and (some of ) their ethical impacts.

Defining Virtual Reality and VR Applications

 What, exactly, is meant by the expressions “virtual reality” and “VR application”?

 Brey (1999) defines virtual reality (VR), as well as a VR application, as

“a three-dimensional interactive computer-generated environment that incorporates a first-person perspective.” [Italics Added]

Brey’s Definition of VR (Continued)

 Three important features in Brey's definition of VR (and VR applications) can be found in its incorporation of:

1) interactivity; 2) three-dimensional graphics; 3) a first-person perspective.

Virtual Reality vs. Augmented Reality

 Some now confuse VR with augmented reality (AR).

 Both AR and VR are “computer-mediated” realities, as well as three-dimensional environments.

 But Søraker and Brey (2015) note that in VR applications, a user’s “field of vision is substantially replaced by the computer-generated visual output.”

 So, VR applications typically provide the user with a “simulated world,” thereby replacing the real world.

 AR technology, on the contrary, enhances (or augments) a user’s view of the real world.

Virtual Reality vs. Augmented Reality (Continued)

 With AR equipment, the user gains either direct or indirect access to the physical (real) world in a way that is enhanced or augmented by computer-generated sensory inputs.

 These typically include video inputs (including eyewear such as Google glass) and sound inputs.

 So, with AR technology, users still interact with their physical environments in real time.

 But our main concern is not with AR applications that enhance our perception of the real (physical) world we inhabit but with VR applications that either simulate the real world or create completely artificial worlds.

 In particular, we focus on ethical aspects of some VR applications.

Ethical Aspects of VR Applications

 How are ethical issues involving controversial behav- ior in (VEs and) VR applications different from issues associated with our everyday physical environments.

 Consider, for example, morally controversial acts displayed on television or played out in board games.

 Brey points out that television programs sometimes display violent acts and some board games allow participants to act out morally controversial roles.

 But Brey notes that in VR applications, users are act- ively engaged, while television viewers are passive.

Ethical Aspects of VR Applications (Continued)

 Brey also notes that VR users are not spectators, but rather are more like actors, i.e., as are the board game players who also act out roles in certain board games.

 But Brey also points out that VR applications, unlike board games, simulate the world in a way that gives it a much greater appearance of reality.

 In VR applications, the player has a first-person perspect- ive of what it is like to perform certain acts and roles, including some that are criminal or immoral, or both.

 We next examine some ethical aspects of one type of VR application: online video games.

Ethical Aspects of Online Video Games

 We should note that some analysts differentiate single- player video games from multi-player online games, which are commonly referred to as MMORPGs (Massively Multiplayer Online Role-Playing Games).

 MMORPGs include games such as Second Life and World of Warcraft.

 Second Life (designed by Linden Lab), which includes members called “Residents,” (has 1million regular users as of 2014, and 13 million registered accounts as of 2008).

 World of Warcraft (WOW), one of the most popular MMORPGs, has 7 million subscribers as of Aug. 2014.

Violent and Sexually-Offensive Acts in MMORPGs

 Some critics claim that Second Life facilitates child pornography because virtual characters who are adults in real life can have sex with virtual characters who are children in that MMORPG (Singer, 2007).

 Cases of virtual prostitution on Second Life have also been reported – i.e., where some Residents were “paid to (use their avatar to) perform sex acts or to serve as escorts” (Brey, 2008).

 If these reports are correct, there are clearly some forms of sexually-offensive acts that take place in MMORPGs, which would not be tolerated outside these environments.

Violent and Sexually-Offensive Acts in MMORPGs (Continued)

 In addition to concerns about sexually offense behavior in online games, many worry about the kinds of violent acts that are also carried out in these environments.

 Wonderly (2008) suggests that some forms of violence permitted in online games be “more morally problematic” than pornography and other kinds of sexually offensive behavior in virtual environments.

 She points out, for example, that relatively few video games “permit sexual interaction between characters”, and even fewer allow “deviant sexual conduct.”

 Wonderly notes that many popular games permit and ev- en require “copious amounts of wanton graphic violence.”

Violent and Sexually-Offensive Acts in MMORPGs (Continued)

 Luck (2009) notes that while most people agree that murder is wrong, they do not seem to be bothered by virtual murder in MMORPGs.

 He points out that some might see the virtual murder of a character in a video game as no different from the “taking of a pawn in a chess game.”

 Luck notes that people have different intuitions about acts in virtual environments that involve morally-objectionable sexual behavior, such as child pornography and pedophilia.

 He worries that the kind of reasoning used to defend virtual murder in games could, unwittingly, be extended to defend virtual pedophilia.

Violent and Sexually-Offensive Acts in MMORPGs (Continued)

 Luck believes that the following argument may unintentionally succeed in doing this:

1. Committing acts of virtual murder does not significantly raise the likelihood of committing actual murder.

2. Committing acts of virtual pedophilia does significantly raise the likelihood of actual pedophilia.

3. Therefore, it is immoral to commit virtual pedophilia, but it is not necessarily immoral to commit virtual murder.

 While this argument may appeal to many, Luck notes that it is also difficult to defend because of the lack of empirical evidence needed.

Assessing “Harm” Resulting from Acts in VEs and VR Applications

 Is it wrong to perform acts in VEs and VR applications that would be considered immoral in “real life”?

 Can these acts cause real harm?

 We might think that since no one can be physically harmed in a VE, any harm caused in the virtual realm is not “real harm” but only virtual harm.

Assessing the Harm in VEs (Continued)

 Because a harm caused in a VE might not result in physical harm to a person, does it follow that no real harm resulted?

 For example, if you receive an insulting e-mail message (in virtual space), is the harm you suffer any less real than the harm that would be caused by the same message written on paper in a letter sent to you via physical mail?

 Review again Scenario 11-1, which shows why harm caused in a virtual space is not unreal harm or harm that is merely limited to virtual characters in a virtual environment.

Arguments for Evaluating Harm in Virtual Environments/VR Applications

 Brey describes two kinds of arguments that can help us to evaluate harm in virtual environments:

1. The argument from moral development, which reasons that the way we treat virtual characters can affect the way we treat real-life people.

2. The argument from psychological harm, which reasons that the way we interact with virtual characters can cause psychological harm to people in real-life situations who have suffered harm (e.g., as in the case of Lambda-Moo and real-life rape victims).

Virtual Economies and MMORPGs

 Some MMORPGs have their own monetary currencies.

 For example, Second Life uses the Linden Dollar (linden, or L$), which its Residents can both use in in-game transactions and exchange outside the game for real currencies such as the US dollar or the euro.

 As a result, virtual economies have emerged.

 Brey (2013) believes that the emergence of these economies can also increase the likelihood that moral controversies will arise in those virtual environments.

Virtual Economies and MMORPGs” (Continued)

 Virtual property, as in the case of virtual money, can be acquired and exchanged with players in MMORPGs.

 It can also be sold and exchanged outside the game to interested parties (in the physical world).

 In some cases, the virtual property has become so desirable that it has led to violent acts in the real world.

 In an incident in China, a person who had stolen some- one’s virtual sword in a MMORPG was murdered in real life by the “sword’s” owner (Warner and Raiter, 2005).

 So, virtual economies can have real-life implications and can result in physical harm to individuals in the real world.

Virtual Economies and “Gold Farming”

 One controversial activity associated with virtual economies in gaming environments is a form of labor and economic exchange called “gold farming.”

 Kimppa and Bisset (2008) define gold farming as “playing an online computer game for the purpose of gaining items of value within the internal economy of the game and selling these to other players for real money.”

 These “virtual” items can include “desirable items” as well as in-game money (where the rules defining the game’s internal economy permit this).

 The items can also include “highly developed” game characters.

Virtual Economies and “Gold Farming” (Continued)

 These “virtual” items can also be sold via online auctions or designated Web sites.

 Kimppa and Bisset point out that the 2009 “in-game gold market” globally was estimated at 7 billion dollars

 They note that the practice of gold farming is most popular in countries like China and Mexico that have both low-average income- levels and “relatively good” Internet access.

Virtual Economies and “Gold Farming” (Continued)

 Gold farming has also raised concerns about working conditions in the real world.

 Warner and Raiter describe a situation in rural China where people who participated in World of Warcraft were paid to work 12-hour shifts of gold farming.

 The workers acquired “virtual gold” within the game and then sold it outside the game to interested players.

 The business became profitable because many players who could afford to purchase the virtual gold” preferred to buy it rather than to do the work necessary to earn it in the game (Warner and Reiter).

Virtual Economies and “Gold Farming” (Continued)

 In some cases, gold-farming practices have become advantageous to the “gold farmers” themselves, who were able to earn more money obtaining and selling virtual gold than they could than in traditional agricultural work.

 But gold-farming practices have also led to reported cases of Chinese sweatshops, where laborers can work “day and night in conditions of practical slavery” to acquire the virtual gold and virtual resources (Brey 2013).

3. Artificial Intelligence (AI) and its Impli- cations for What it Means to be Human

 We can ask: Which criteria distinguish humans from other kinds of creatures and entities?

 Is rationality a plausible criterion?  The view that only humans are rational has

recently been challenged on two separate fronts: 1) research in animal behavior/intelligence

suggests that many primates, dolphins, and whales are capable of demonstrating skills we typically count as rational;

2) developments in artificial intelligence (AI) have shown that certain forms of "rational activity" can also be attributed to computers.

What is AI?

 John Sullins (2015) defines AI as “the science and technology that seeks to create intelligent computational systems.”

 He notes that AI research has aimed at building computer systems that can duplicate, or at least simulate, the kind of intelligent behavior found in humans.

AI: Background and Some Key Historical Developments

 Early AI research focused on developing software programs that could play chess, do calculus, and “solve problems” that require a high level of human intelligence.

 The first AI programs were “problem solvers” comprising software code, but included little or no hardware.

 Many believed that AI could be achieved without having to copy “nature’s way of thinking” (i.e., via a physical brain).

 AI research in that era, which focused on constructing a kind of “disembodied intelligence,” is now typically referred to by expressions such as “classical AI,” “symbolic AI,” or “good old fashioned AI” (GOFAI).

AI Background/Developments (Continued)

 The classical AI approach was criticized by those who argued that human intelligence cannot be reduced to a manipulation of symbols (captured in software programs).

 One group argued that an artificial brain with neural networking (that could “perceive” and “learn” its environment) was also required for a machine to learn and understand the world and thus potentially duplicate the way that humans think.

 This scheme in AI is often described as a “bottom-up” (or inductive) approach to machine learning.

 The classical/symbolic AI model, on the other hand, is typically viewed as a “top-down” (or deductive) approach.

AI Background/Developments (Continued)

 Another division in AI arose when a group of AI research- ers argued that it was not critical to build machines that were as intelligent as humans (or even to build machines that thought in the same way humans do).

 They believed that a legitimate goal for AI would be to develop systems that were “expert” in performing specific tasks that required a high level of intelligence in humans.

 For example, a system such as an “expert doctor” could be highly competent in diagnosing medical diseases, although it would be unable to perform any tasks outside that very narrow domain.

 See also the discussion of expert systems in Chapter 10.

AI Background/Developments (Continued)

 But many AI researchers believed that it was still possible to achieve the original goal of emulating (general) human intelligence in machines.

 Some, including researchers working on the CYC Project, use an approach that builds on classical/symbolic AI by designing software programs that manipulate large databases of factual information.

 Others, such as “Connectionists,” have designed neural networks that aim at modeling the human brain, with its vast number of neurons and arrays of neural pathways, which exhibit varying degrees of “connection strengths.”

AI Background/Developments (Continued)

 One concern that arose early in AI research could be viewed as more sociological than technological in nature,

 This concern or question asked how we (as humans) might react to a world where machines would be as intelligent, or possibly even more intelligent, than us.

AI Background/Developments (Continued)

 John Weckert (2001) articulates this kind of (sociological) concern when he asks:

Can we, and do we want to, live with artificial intelligences? We can happily live with fish that swim better than we do, with dogs that hear better, hawks that see and fly better, and so on, but things that can reason better seem to be in a different and altogether more worrying category…What would such [developments mean for] our view of what it is to be human?

Can We build (Genuinely) Intelligent Machines? An Ongoing Debate

 We can ask whether it is possible, even in principal, to build “machines,” i.e. either software programs or (physical) artificial entities, that are “genuinely intelligent” and whose intelligence could rival and possibly exceed that of humans.

 Some critics argue that, at best, AI researchers would be able to build machines that merely simulate, rather than replicate, human intelligence.

 But this is still an “open question”; the debate is ongoing.

 Each side has presented a series of arguments and “thought experiments” to defend its position.

The Turing Test as a “Thought Experiment” for Machine Intelligence

 In 1950 Alan Turing predicted that by the year 2000, a computing machine would be able to pass a test, which has come to be called the Turing Test, demonstrating machine intelligence.

 Turing envisioned a scenario in which a person engaged in a conversation with a computer (located in a room that was not visible to the human) was unable to tell – via a series of exchanges – whether he or she was conversing with another human or with a machine.

 He believed that if the person could not be sure that this entity was a human or a computer, then we would have to attribute some degree of intelligence to the computer.

The Turing Test, Deep Blue, and Watson

 While AI researchers would concede that Turing's proph- ecy has not yet been fully realized, they also point to the significant progress and achievements in AI thus far.

 For example, in 1997 an IBM computer program called Deep Blue defeated Gary Kasparov, then reigning champion, in the competition for the world chess title.

 Also, in 2011, another IBM computer program, called Watson, defeated two human opponents in the TV game show Jeopardy in a championship match (a human- computer competition that was viewed by millions of people around the world.)

The Turing Test, Deep Blue, and Watson (Continued)

 Watson, like Deep Blue, is a disembodied AI, i.e., a highly sophisticated set of computer programs.

 Unlike Deep Blue, which could be viewed as an “expert system” that is highly skilled at playing chess (but not necessarily competent in other areas), Watson is capable of answering a wide range of questions that are posed to it in natural language.

The Turing Test and Watson (Continued)

 Some believe that Watson’s skills at least simulate human intelligence in the broad or general sense of “intelligence.”

 But did Watson, in defeating its human challengers, also exhibit the skills necessary to pass the Turing test?

 Even if Watson could pass the Turing test, would that necessarily show that Watson possessed (human-like) intelligence.

The Turing Test and Searle’s “Chinese Room” Argument

 Was Watson merely acting in a manner similar to the individual in John Searle’s (now) classic “Chinese Room” thought experiment?

 In that scenario, a human who is a native English speaker but who understands nothing about the Chinese language is able to perform tasks that require manipulating Chinese symbols to produce correct answers to questions posed in Chinese.

The “Chinese Room” Argument (Continued)

 The person in the “Chinese room,” who is not seen by anyone outside that room, receives questions from someone who passes them to him through an opening or slot.

 That person then consults a series of instructions and rules located inside the room, all of which are written in English.

 The person is able to substitute the incoming Chinese symbols for other Chinese symbols in such a way as to produce correct answers to the questions asked.

The Chinese Room Argument (Continued)

 Once the person inside the room has completed the task, he passes the answers through the slot to a person waiting outside.

 That person might assume that the individual who returned the correct answers understood Chinese.

 But Searle (1980) argues that it is possible that the person (inside the room) understood nothing about the semantic meaning of the questions he received and the answers he returned.

The Chinese Room Argument (Continued)

 Searle points out that the person in the room could merely have been following a set of syntactic rules (written in English) that were designed to manipulate symbols that happened to be in Chinese.

 In fact, it is possible that the English- speaking person (inside the room) may not even know that the symbols involved are elements of the Chinese language.

Watson, the Turing Test, and the Chinese Room Argument

 Was Watson’s behavior in the Jeopardy game show analogous to that of the person in Searle’s Chinese room?

 Did Watson actually “understand” the meaning of the symbols (in the questions and answers) involved?

 Or did Watson simply use a series of syntactic rules and cross-checking algorithms to manipulate the information stored in Watson’s vast database to get the correct answers?

Watson, the Turing Test, and the Chinese Room Argument (Continued)

 It is not clear to what extent, if any, Watson could be said to have any understanding of natural language?

 So, even if Watson is capable of passing the Turing test (as it was originally posed), it would not necessarily follow, using Searle’s argument, that Watson possesses human-like intelligence.

Watson, SIRI, and the Chinese Room Argument

 Some skeptics might argue that Watson is nothing more than a kind of (very broad) expert system, or perhaps some combination of expert systems, that behaves like a more sophisticated version of (Apple’s) SIRI.

 Although SIRI is capable of responding to many questions posed to “her” with correct answers, it is doubtful that people would be willing to describe SIRI as possessing human-like intelligence.

Watson, SIRI, and the Chinese Room Argument (Continued)

 So, just as we need not ascribe human-like intelligence to SIRI, one could argue that it is not necessary to attribute genuine (or human-like) intelligence to Watson.

 Nevertheless, we can still see why many humans would feel uneasy about the fact that a computer, or AI entity, had defeated two highly intelligent human beings in a championship match involving intelligence (even if it was only in a game show contest).

Cyborgs and Human-Machine Relationships

 Consider some questions pertaining to the nature of the human-machine relationship that can result from the development of cyborgs and related AI entities)

 We have already seen how the prospect of machine intelligence in AI research can affect our sense of what it means to be human.

 The development of cyborgs (and related AI entities) may have a similar effect on us.

Cyborgs and Human-Machine Relationships (Continued)

 We examine this controversy via two distinct, but related, questions:

a) Are humans becoming more computer- like?

b) Are computers becoming more human- like?

 We look at (a) from the perspective of AI-induced bionic chips.

Cyborgs and (AI-induced) Bionic Chip Implants: Are Humans Becoming More Computer-Like?

 Future chip plants made possible by AI could be designed to make a normal person “super human.”

 Weckert (2001) notes that “conventional” implants designed to “correct” deficiencies have been around and used for some time.

 The purpose of conventional, or “therapeutic” implants, has been to assist patients in their goal of achieving “normal” states of vision, hearing, heartbeat, etc.

AI and Controversies Involving Bionic Chip Implants (Continued)

 Moor (2005) notes that because the human body has “natural functions,” some will argue that implanting chips in a body is acceptable as long as these implants maintain and restore the body’s “natural functions.”

 He also notes that while therapeutic implants will likely be accepted, “enhancement implants” will likely be controversial.

 Moor believes that many will find a policy based on a therapeutic-enhancement distinction to be appealing.

AI and Chip Implants (Continued)

 In other words, Moor believes that:  therapeutic chip implants such as

pacemakers, defibulators, and bionic eyes that maintain and restore natural bodily functions will most likely be accepted;

 enhancement implants, such as giving patients added arms or infrared vision, will most likely be prohibited.

AI and Chip Implants (Continued)

 Additionally, Moor believes that such a policy would likely endorse the use of chip implants that:

 reduced dyslexia;  assist memory of Alzheimer patients.  But Moor also believes that this kind of policy

would not endorse the implanting of either a:  “deep blue” chip for superior chess play;  miniature digital camera that would record

and playback what a person had just seen.

AI, Chip Implants, and What it Means to Be Human

 Arguably, we need to assess now some of the advantages and disadvantages of bionic implants that AI will make possible.

 Weckert (2001) asks us whether we:

want to be ‘superhuman’ relative to our current abilities with implants that enhance our senses, our memories, and our reason- ing ability? What would such implants do to our view of what it is to be human?

Are We (Becoming) Cyborgs?

 Some now worry that with bionic parts, humans and machines could soon begin to merge into cyborgs.

 Ray Kurzweil (2000) suggests that the distinction between machines and humans may no longer be useful.

 James Moor (2005) believes the question we must continually reevaluate is not whether we should become cyborgs, but rather: What sort of cyborgs should we become?

The Challenge in Distinguishing AI Entities from Humans: Are Computers Becoming More Human-Like?

 Some AI entities (i.e., “bots”) in the form of avatars already assist users in organizing their work schedules, reminding them of important scheduled meetings, arranging travel, and so forth.

 Also, personal digital assistants, and smart phones using voice-recognition programs (such as iPhone’s SIRI), now interact with humans on a daily basis.

 Even though they are merely virtual entities, some already exhibit human-like features when viewed on screens or when listened to on electronic devices.

Are Computers Becoming More Human-Like (Continued)?

 After interacting with one’s human-like agent (avatar) over a long period of time, is it possible that one might begin to act as if he/she is conversing with a real person?

 Is it also possible that one might, after an extended period of time, begin to confuse some virtual entities (with whom one interacts) with actual flesh-and-blood characters that those entities represent.

 For example, as virtual entities become increasingly more human-like in appearance, it may also become more difficult to distinguish between our interactions with some (physical) person’s screen avatar and with an actual human represented by that avatar or virtual entity.

Are Computers Becoming More Human-Like (Continued)?

 Confusion resulting from interacting with artificial entities may become more exacerbated as we move from virtual entities on screens (of computers/devices) to interacting more regularly with physical AI entities, such as robots.

 Sophisticated robots of the future may not only look more human-like, but may also exhibit sentient characteristics.

 These robots, like humans and animals, could (arguably) have some sentience (and thus be capable of simulating the experiences of sensation, feeling, and emotion).

 Robots and other kinds of AI entities of the near future may also exhibit, or appear to exhibit, consciousness.

Are Computers Becoming More Human-Like (Continued)?

 The movie 2001: A Space Odyssey includes a computer named HAL, who exhibits higher- order thinking functions and behavior that also resembles human consciousness.

 In addition to carrying out his ordinary computational tasks, HAL engages in sophisticated conversations with members of the space ship’s crew, plays chess, criticizes art, etc. – all of which seem to exhibit some human-like intelligence on HAL’s part.

Are Computers Becoming More Human-Like (Continued)?

 Even if AI entities like HAL do not succeed in achieving full consciousness, they could still be capable of exhibiting (or simulating) other human-like characteristics such as rationality; other AI entities may simulate sentience.

 Although HAL seemed to simulate human consciousness, he was not sentient, since ‘he” had no body (and thus could be viewed as a form of disembodied intelligence).

 So it is possible that some AI entities (like HAL) may exhibit (or at least simulate) rationality and consciousness but not sentience, while other AIs may exhibit (or simulate) rationality and sentience but not consciousness.

Are Computers Becoming More Human-Like (Continued)?

 So, independently of whether these AI entities will ever achieve (human-like) consciousness, they could still exhibit (or simulate) some other human-like characteristics.

 Critics worry whether we will be prepared to meet the kinds of social and ethical challenges these kinds of AI entities will likely pose.

 Review Scenario 11-2 (in the textbook), which illustrates one kind of moral concern that could arise if we develop human-like AI entities.

Do (At Least Some) AI Entities Deserve Moral Consideration

 Three questions worth considering from Scenario 11-4 are:

1) Should the artificial boy (or any artificial creatures like “him”) ever have been developed in the first place?

2) Is it morally permissible for the “boy’s” adopted parents to discard “him” as they would some other kind of “computer resource”?

3) Does this “boy” deserve (at least some) moral consideration (especially from the human parents who adopted “him”)?

Expanding the Sphere of Moral Consideration for AI Entities

 Will we need to expand the domain of moral consideration to include artificial-intelligent entities (such as the artificial boy, “David,” in the film AI, described in Scenario 11-4)?

 If the answer to that question is “yes,” two additional questions arise:

A. Which kinds of entities or “things” deserve moral consideration?

B. Why do they warrant it?