Big Developments in E-Business - 6 pages - APA - 5 references - Due 31 October

ORI GIN AL ARTICLE

The rise of cyberinfrastructure and grand challenges for eCommerce

Susan J. Winter

Received: 10 December 2009 / Revised: 20 November 2010 / Accepted: 20 November 2010 /

Published online: 21 January 2011

� Springer-Verlag 2011

Abstract Advances in ICT have enabled the transformation of commerce into eCommerce. The eCommerce revolution is well under way, but what grand chal-

lenges are on the horizon? This paper extrapolates from the past to understand the

forces that shape the future. It focuses on cyberinfrastructure growing out of the

Cold War and highlights the role of government, Big Science and of the National

Science Foundation in that story. Big Science and eCommerce will continue to

shape Web activities and to react to advances in ICT. Both are involved in the co-

evolution of physical, social, organizational, economic and legal factors and face

analogous issues arising from similarly disruptive transformations, though the

manifestations of these transformations differ in their surface features. Each may

lead in addressing a particular issue at a given time, but can learn from the other.

Finally, this paper identifies three Extreme Grand Challenges for Big Science and

eCommerce that represent important steps toward an even brighter future for both.

Keywords Grand challenge � Cyberinfrastructure � Internet history � Infrastructure � eCommerce

1 Introduction

Information and communication technologies (ICT) have enabled the transforma-

tion of commerce into eCommerce with the rise in globalization, flattened

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the

author and do not necessarily reflect the views of the NSF. Some ideas in this paper were drawn from an

unpublished talk by John L. King, and are used with permission from the author.

S. J. Winter (&) Office of Cyberinfrastructure, National Science Foundation, 4201 Wilson Blvd,

Arlington, VA 22230, USA

e-mail: [email protected]

123

Inf Syst E-Bus Manage (2012) 10:279–293

DOI 10.1007/s10257-011-0165-5

hierarchies, and the growth of networked organizations, not to mention doing

business ‘‘on-line.’’ These have depended on improved ‘‘cyberinfrastructure,’’

meaning advances in network capacity, connectivity, interoperability, processing,

and effective handling of large quantities of data (Atkins 2003). Early barriers to

eCommerce such as low public trust of online payments have been largely

addressed. The eCommerce revolution is well under way, but what issues remain?

To use contemporary parlance from Big Science, ‘‘What grand challenges are on the

horizon?’’ This paper extrapolates from the past to understand the forces that shape

the future. It focuses on cyberinfrastructure growing out of the Cold War, highlights

the role of the National Science Foundation in that story, and suggests steps toward

an even brighter future for Big Science and eCommerce working together to address

emergent cyberinfrastructure grand challenges.

The term ‘‘Big Science’’ as used in this paper encompasses all science and

engineering research that meets one or both of two conditions. First, it is expensive

due to the need for equipment, large numbers of expert personnel, and institutional

support. The Manhattan Project and the Apollo Program are often cited as examples

of Big Science, but there are earlier examples including the mobilization of the

National Foundation for Infantile Paralysis (polio) and its hugely successful March

of Dimes campaign in the 1930s that led to the creation of the Salk and Sabin

vaccines and the virtual elimination of polio in many parts of the world. Second, it

addresses a problem of great scale and/or social importance requiring the

engagement of a large number of experts from across different domains. A

contemporary example of this is the challenge of climate change, but similar

examples can be seen in efforts to clean up major bodies of freshwater such as the

Great Lakes. Big Science is a nickname, and as such is intended merely as a

shorthand description of much more complicated, underlying activities.

World War II and the subsequent Cold War altered the US national climate for

Big Science and ushered in an era of close partnership between the federal

government, academia, and private industry. That climate of collaboration was

essential in the development of the Internet that enabled eCommerce. The Internet

would not have been created only by commercial interests pursuing profits; the role

of government pursuing a vital national interest was essential, as was the role of

academics working at the forefront of science (King et al. 1997). Nevertheless, the

thing we now call the Internet was shaped dramatically by commercialization, a

process undertaken by and overseen by NSF in the 1990s because of the network’s

growing importance to Big Science. The importance of the Internet to commerce of

all kinds grew out of original investments made on behalf of national defense and

the role of Big Science in such defense, but the story of the Internet in the era of

commercialization is far from over.

2 NSF, big science and the cold war

The National Science Foundation was born in the midst of the Cold War. It arose

from the vision of Vannevar Bush, Dean of Engineering at MIT and an advisor to

US leaders in WW II. In July of 1945 Bush published an influential article titled ‘‘As

280 S. J. Winter

123

We May Think’’ in The Atlantic Magazine (Bush 1945a) that foreshadowed many scientific advancements yet to come and helped build public support for science.

The same month, in his role as Director of the US Office of Scientific Research and

Development, Bush issued a report to President Harry Truman titled ‘‘Science, The

Endless Frontier’’ that proposed something called a National Research Foundation

(Bush 1945b). This document was the impetus for the creation of the NSF to

‘‘…promote the progress of science… advance the national health, prosperity and welfare… secure the national defense…’’ (NSF 1950).

The Cold War that ran between 1948 and the early 1990s saw much innovation in

information technology, based in research conducted at the intersection of military,

industrial and academic interests. For example, the Semi-Automatic Ground

Environment, Project SAGE, was designed and deployed to detect Russian bombers

entering the US from Canada and determine whether they represented a surgical

strike or a large scale attack (Hughes 1998; Naughton 2000; King 2010). SAGE

consisted of radar stations, command centers, and air defense positions connected

through a data communication network and modern computers operating 24/7.

Project SAGE had many consequences, not the least involving Vinton Cerf, later

dubbed a ‘‘father of the Internet.’’ Cerf reports that he became interested in

computers as a teenager when he saw a SAGE computer at the System Development

Corporation in Santa Monica, California (Lambert et al. 2005).

The Internet was also a product of the Cold War, but its evolution is more

complicated than most people know. The Internet had origins in the ARPANet, the

first successful packet-switched network developed by the Department of Defense

Advanced Research Projects Agency between the late 1960’s and the late 1980’s

(Edwards 1996; Abbate 1999; Naughton 2000). In 1962 J.C.R. Licklider, a prime

instigator of the ARPANet, moved to ARPA from an advanced Boston-area

research company called Bolt Beranek and Newman (BB&N). BB&N had been

founded in 1948 by two MIT professors and a graduate student, and had established

a strong reputation for its work for the military. Licklider convinced DARPA to

invest in an advanced computer-based communications infrastructure that could

survive multiple nuclear strikes (the strategic objective) and bring scientists working

for the DoD to computer resources instead of having DoD buy each scientist an

expensive computer (a practical objective).

In 1968 ARPA awarded a contract to BB&N with work to be done by various

universities on packet switching. Vinton Cerf, at this time a graduate student at

UCLA, worked on Telnet and the File Transfer Protocol (FTP) as part of this project

(Crocker 1987). As a faculty member at Stanford, Cerf co-designed with Robert

Kahn the Transfer Control Protocol (TCP) and the Internet Protocol (IP) which

together formed TCP/IP, the basis of the ARPANet. Cerf subsequently became an

ARPA program manager, funding research that enabled key features of the

emerging network that proved particularly useful for distributed communities of

scientists such as remote login and file transfer (Cerf 1993; Lynch 1993).

As noted above, providing computer support to scientists was an important early

objective for the ARPANet. As part of the Cold War, the US invested heavily in

provision of scientific computing for US scientists between 1950 and the early

1970’s (Aspray and Williams 1994). A significant portion of this investment was

The rise of cyberinfrastructure and grand challenges for eCommerce 281

123

through the National Science Foundation, and NSF became the first-tier provider of

computing support of US ‘‘open science.’’ NSF was not alone in providing

computing support: the Atomic Energy Commission (later the Department of

Energy), the Department of Defense, and specialized agencies such as the National

Aeronautics and Space Administration and the National Weather Service also made

investments. But NSF’s role as a supporter of ‘‘open science’’ proved vital as the

Internet evolved.

During the 1980s NSF’s destiny in computing was transformed by the rise of the

supercomputer movement, spurred largely by computational scientist Kenneth

Wilson (Fisher 2006). Wilson was an elementary particle physicist who had joined

the faculty at Cornell in 1963 and won the Nobel Prize in Physics in 1982 for his

work on quantum field theory. Wilson worked on problems with finite but high

degrees of freedom that could be solved only with very powerful computers. He

successfully lobbied Congress and the White House to make a major national

investment in what became known as supercomputing, and NSF was given the task

of implementing a plan to build and support supercomputing at a national scale. Six

supercomputer centers were founded in this effort, of which three survive to this

day: the San Diego Supercomputer Center at UCSD, the National Center for

Supercomputer Applications at the University of Illinois at Urbana-Champaign, and

the Pittsburgh Supercomputer Center run jointly by the University of Pittsburgh and

Carnegie-Mellon University. Support for these centers and related programs became

part of the new NSF Directorate for Computer and Information Science and

Engineering (CISE) in 1986.

The DoD withdrew from the rapidly evolving ARPANet project in the late 1980s

to pursue defense-only applications and NSF took leadership of the project because

the emerging network had become an important tool for Big Science. NSF

continued supporting the academic programs related to the ARPANet, and

embarked on the creation of a new network called NSFNet through the non-profit

MERIT Network, Inc. in Ann Arbor, Michigan. MERIT and the academic centers

worked with IBM and MCI to improve network capability (T1 speeds of 1.5 mb/

second in 1988 and T3 speeds of 45 mb/second in 1991), opened the network to

commercial use in 1993, and officially named the network the Internet in 1995 just

before shutting NSFNet down and turning the enterprise over to a mix of non-profit

and commercial interests. The initial ‘‘testbed’’ of the rapidly growing network were

links between the NSF supercomputer centers, and the network topology today

clearly shows the role of those centers.

NSF remains engaged with advanced networking. NSF’s CISE Directorate

continued to support supercomputing until 2005 when the NSF-wide Office of

Cyberinfrastructure (OCI) was created, and continues to support research in

computer architectures, data management and high-performance networking. OCI

supports the development and deployment of high end computing power (e.g.,

high performance computers, clusters, grids, clouds), simulation and visualization

methods and technologies, large scale data acquisition, processing, and storage

techniques, distributed and networked sensors, instrumentation, and remote work,

linking people and information to form virtual organizations (Atkins 2007).

282 S. J. Winter

123

eCommerce today runs on a foundation built in significant measure by the NSF,

driven by the powerful forces of the Cold War and the needs of Big Science. The

Cold War wound down in the early 1990s at about the time NSFNet was taking the

network toward commercial development. Since the end of the Cold War,

commerce has become a powerful driver of Internet development, replacing the role

played by defense in the earlier period. Big Science remains an important factor in

the development of cyberinfrastructure technologies. Perhaps more important, Big

Science depends upon advancements both in enabling technologies and in the social

conventions of work. eCommerce does, as well. The next section describes the way

the work of Big Science is changing, with implications for eCommerce.

3 Cyberinfrastructure and big science at work

Big Science is increasingly organized around grand challenges: goals that push the

boundaries of existing knowledge, established disciplines, and available capabilities

(Hoare and Milner 2005). Grand challenges embrace multiple intellectual fields as

well as societal concerns such as competitiveness, security, economic development,

or well being. Examples include reconciling quantum physics and relativity, making

solar energy economical, reverse-engineering the human brain, and achieving

universal and effective primary education. Progress on grand challenges requires

long time frames, new tools and changes in the ways that work gets done (Lutters

and Winter in press). In the process, fields are reconstructed: we move from simple

problems such as mapping ocean currents to the grand challenge of mitigating

global climate change. Grand challenges often require the efforts of distributed

interdisciplinary global teams including participants from India, South Korea, China

and other countries that have recently developed world-class capabilities (Friedman

2005). In short, grand challenges require resources and participation from academic,

practitioner, and policy-oriented communities, as well as cooperation among groups

with differing perspectives over long periods of time.

This section discusses an early example of the grand challenge phenomenon: the

high-energy physics community that built on the Internet to create the World Wide

Web between 1989 and 1991 at Cousiel European pour la Reserche Nuclaire

(CERN), the European Organization for Nuclear Research (Naughton 2000).

CERN’s experiments take place using huge instruments located in deep pits along a

27-km circular tunnel about 100 m underground. By the mid-1990s most of

CERN’s work was focused on the creation of the world’s most powerful particle

accelerator, the Large Hadron Collider (LHC), which accelerates elementary

particles to near the speed of light. In 1993, the US Congress had cancelled funding

for the Superconducting Super-Collider then under construction in Texas, and

CERN became the world’s only option for such an accelerator. With the LHC,

physicists could collide particles at speeds high enough to mimic the conditions of

the big bang when the universe was created. The resulting data would aid the search

for the theoretically postulated Higgs Boson, supersymmetry, and other important

questions in the grand challenge of reconciling quantum physics and relativity. The

LHC would cost about 10 billion USD (including significant NSF funding), so

The rise of cyberinfrastructure and grand challenges for eCommerce 283

123

sharing was essential. The largest of the LHC experiments, the ATLAS (http://

atlas.ch/), has nearly 3,000 principal investigators worldwide. New tools and new

ways of working were necessary.

A CERN researcher, Tim Berners-Lee, needed an easy way to distribute many

documents to thousands of researchers around the world. He and Mark Andreeson, a

young computer scientist at NCSA (NCSA’s supercomputing resources were being

used to process CERN-generated data), developed a hypertext markup language for

documents and an application called Mosaic that could ‘‘read’’ documents created in

that language. Html and Mosaic quickly became important parts of the high energy

physics collaboration toolkit, but were also recognized more broadly as important

innovations in communication. Mosaic was turned into a commercial product called

Netscape by Andreeson and Eric Bina at NCSA, and released to the public in 1993

(Reid 1997). Netscape was the first ‘‘Web Browser’’ that became the heart of

business-to-customer (B2C) ecommerce. Browsers are now available from

commercial providers (Microsoft’s Internet Explorer, Apple’s Safari, Google’s

Chrome) and from open source consortia (Mozilla’s Firefox). Big Science provided

the venue for the rise of the Web, but the Web itself grew up mainly around other

forms of enterprise, many of them commercial.

Big Science has pushed the frontier of ICT dramatically both with respect to the

underlying technologies and to ways in which work can change through their use.

The Large Hadron Collider, when fully operational, will generate up to a petabyte (a

quadrillion bytes) of data per second. Most of these data will be stripped out

immediately and discarded as uninteresting, but about 100 petabytes per year must

be distributed globally to thousands of researchers working on the LHC

experiments. Such data transfers require ‘‘audit trails’’ to allow reanalysis and

reproducibility of important results. Where computational power was once the main

concern (as per Ken Wilson’s crusade for supercomputers), management of data is

now looming as a major challenge. This shift is subtle but vital for understanding

the larger role of cyberinfrastructure in changing work because it is a shift from

‘‘inputs’’ to ‘‘outputs.’’

Ironically, the fantastic drop in the cost of digital storage exacerbates this

challenge because it removes previously powerful, natural impediments to going

‘‘big’’ with data. Continuing with high-energy physics, Traweek (1988) provides a

detailed account of the resource politics affecting physicists’ efforts to secure access

to the major inputs—the large instruments without which contemporary high-energy

physics is impossible. Gaining access to inputs is in service of the real goal, namely

‘‘outputs’’ in the form of findings and discoveries. Data are the vital first steps

toward those payoffs. Not surprisingly, over time high-energy physics has evolved

ways of handling data that affect collaboration and competition in the scientific

‘‘marketplace of ideas.’’

Cyberinfrastructure empowers Big Science to handle very large data sets, but that

is dependent on the evolution of the right social conventions. Gaps in existing

knowledge about the social conventions of handling data are becoming clear. For

example, the law does not permit copyrighting of data, but certain data sets can be copyrighted. Ambiguity around the fundamental question of what can be ‘‘owned’’

makes it difficult to adjust as old presumptions about de facto ownership by

284 S. J. Winter

123

investigators give way to increasing calls for investigators to make their data

available to other investigators. And this signals a broader concern about the nature

of the ‘‘payoff’’ from the work of Big Science, namely credit for publication.

Traditionally, Big Science followed a complicated but well-understood path from

collection and analysis of data to working papers to discussion in conferences and

colloquia, to publication in archival journals. Powerful but often invisible and tacit

social conventions governed this path: attribution of provenance (e.g., authorship

and order of authorship), the character and quality of peer review (e.g., open vs.

single vs. double-blind reviewing), assignment of intellectual property rights (e.g.,

assignment of rights to commercial publishers who sell the product back to the

researchers), and so on. By the early 1990s high-energy physicists were playing

with that traditional path, using the Internet to begin ‘‘pre-print’’ publication of

important papers, a practice since embraced by other fields of science and

scholarship (Harnad 1995). It is no longer clear when, exactly, a work is

‘‘published.’’

Beyond this, there is a question of credit when a work is published. When cost makes it necessary that only one set of data can be collected, how do you share it

among all those who are interested? CERN’s ATLAS experiment with nearly 3,000

investigators illustrates this: papers published from CERN research can have

hundreds of co-authors, with the list of authors longer than the article itself.

University promotion and tenure committees might have difficulty evaluating the

contribution of a single author within that list, so the high-energy physics

community has had to develop internal social conventions for assigning credit that

are both fair and not self-debasing (e.g., not assigning more credit than there is to

assign).

Finally, how in the era of Big Science can reproducibility of results be ensured?

This is especially problematic with the case where a given kind of work can only be done at a single location such as CERN. For one thing, it is not clear exactly how

ownership of intellectual property works in a project costing billions of taxpayer

dollars from more than a dozen countries. But even if that is clarified, it remains to

determine what must be shared to ensure reproducibility. Must data be accompanied

by the software, models, and simulations required to reproduce the work? At what

stage should data be released: raw, cleaned, or analyzed? Who pays the added costs

for later release? And who is responsible for long-term curation of the data, an

especially important issue given the cost of reproducing the data? Answering such

questions requires re-organization of research work at many levels.

Big Science increasingly requires the co-evolution of physical, social, organi-

zational, economic and legal factors. NSF is not alone in this challenge, but as the

major supporter of open science, and the federal government’s largest supporter of

cyberinfrastructure, NSF faces particularly acute challenges in this regard. One of

the challenges comes from resource politics. NSF has long supported single-

investigator science, but in the past 30 years it has expanded to support projects

with large teams of investigators. Some of these projects involve large, established

centers such as the National Center for Atmospheric Research (NCAR), the

Southern California Earthquake Center (SCEC) and the supercomputer centers

already mentioned. Some members of the Big Science community feel the ‘‘right

The rise of cyberinfrastructure and grand challenges for eCommerce 285

123

kinds’’ of research are being shortchanged to do the ‘‘wrong kind’’ of research. NSF

also must learn to manage different kinds of research programs, each with its own

complexity: concentrated and shared infrastructure such as that at CERN or the

National High Magnetic Field Laboratory (MagLab); networked but geographically-

distributed infrastructure such as the George E. Brown Network for Earthquake

Engineering Simulation (NEES); finely distributed sensor networks such as the

recently launched National Ecological Observatory Network (NEON); and ‘‘citizen

science’’ projects such as Zooniverse in astronomy and eBird in ornithology. These

challenges go beyond the problem of incorporating computation into Big Science

(Wolfram 2002; Atkins 2007; NSF 2010).

The work of Big Science evolves through practice. Cyberinfrastructure has

enabled new ways of working, and as seen in the birth of the World Wide Web at

CERN, that enablement sometimes coincides with emerging needs in the scientific

community. The needs of the scientific community, in turn, can reflect the needs of

the society at large. The explosive commercial growth of the Web in the late 1990s

and early 2000s was foreshadowed by the smaller but nonetheless impressive

growth of the Web in Big Science and the academy generally a few years earlier. In

some ways, Big Science is a leading indicator of challenges facing the world of

eCommerce. And, as discussed in the next section, eCommerce offers important

opportunities for the realm of Big Science.

4 Lessons and grand challenges for eCommerce

There is no question that the Internet has transformed commerce. The examples are

legion, from the rise of ‘‘e-tailing’’ to phenomena such as Google (which grew out

of the NSF-funded Digital Libraries Initiative), FaceBook, YouTube, Twitter and so

on. In some cases it is not clear what these phenomena mean. Even Google’s

founders did not imagine when they created the company that it would earn most of

its revenues from advertising, but the Google model profoundly changed the world

of advertising ushering in an era of clickstream revenue and data analytics

(Hofacker and Murphy 2000). Every venture capitalist, if not every investor, wants

to know what the ‘‘next big thing’’ will be. The experience of Big Science cannot

‘‘pick winners’’ in this race, or even point the way, but it can be used to see where

the trends are headed. Clearly, Big Science, eCommerce and government will

continue to shape Web activities and to react to advances in ICT. Though surface

features differ, all face analogous issues arising from similarly disruptive

transformations, many of which are enabled by new technologies.

The Google phenomenon illustrates one of these trends, a transformation

catalyzed by the aggregation and management of large-scale data resources. The

beauty of Google is its simplicity: it collects data from the users of the Web and

sells those data, in different forms, to users of the Web. The magic of Google lies in

the sophistication of its infrastructure, both with respect to collecting and

redistributing data, and operating at very large scale. That sophistication raises

concerns, as well, including those of personal privacy. The question of what Google

‘‘knows’’ about individuals and what it is willing to do with such information has

286 S. J. Winter

123

been at the heart of many criticisms of the company and its practices. The company

famously touts its motto, ‘‘Do no evil,’’ but given the company’s inevitable

transgressions in the service of its transformations, that motto can be seen as cynical

just as easily as it can be seen as heartfelt. The issue is not whether Google or its

leadership are pro-privacy or anti-privacy, but whether our concepts of privacy

grounded in an earlier era are up to the challenges of a new era where technology

enables something like Google to exist.

The challenges of privacy in the Google era are similar to those facing Big

Science. New analytic techniques and new sources of data (especially those arising

from ubiquitous mobile computing), are transforming the human sciences enabling

such new research areas as societally-informed climate models, global-scale

epidemiological models, and understanding human networks. Text can be extracted,

combined and analyzed automatically from traditionally public materials (e.g.,

published articles, speeches, and government reports), but also from tweets,

personal blogs and email messages. Commercial and social data of scientific interest

have long been captured as credit and debit card transactions, but geographic

location information is also available from ubiquitous computing devices such as

cell phones, tollbooths using transponders, and RFID product tracking. Health care

research may be transformed by access to more complete databases linking person-

level genomics, proteomics, and digital medical records with data on hospital

admittance, purchases, and location.

These changes bring both scientific opportunities and challenges, including

privacy concerns similar to those facing Google. Advances in cyberinfrastructure

can link existing information sources (databases, published literature) with

previously private data about individuals, transactions, and social networks. These

can be integrated with data from ubiquitous computing devices, distributed sensor

webs, and other sources in ways that could revolutionize the human sciences.

Exploiting such scientific resources without infringing on personal privacy requires

careful attention to the sensitivity of person-level data and development of new

techniques to allow individual identities to remain hidden during data collection and

analysis. These techniques may be able to mitigate some of the privacy issues facing

Google as well.

Wikipedia represents another web-based phenomenon with important implica-

tions for Big Science. The fifth most popular website, Wikipedia is accessed by over

400 million people per month (about one-third of the Internet connected world) and

provides information on topics of general interest. Building on the scientific

encyclopaedic tradition of aggregating summary reference documents on a wide

variety of topics, Wikipedia has transformed societal expectations regarding the

availability of standardized, structured overview documents. One of these expec-

tations is that identical terms should be disambiguated. Disambiguation is a crucial

service for improving information search efficiency and effectiveness, alerting

readers to various meanings of a term and directing them to information on the

target subject. Wikipedia’s disambiguation service is quite extensive. For example,

if one enters the term ‘‘Portland’’ into Wikipedia, the resulting disambiguation page

lists over 75 different meanings with links to relevant pages for Portland, Oregon,

Portland, Maine (and New Portland, Maine, which is noted to be ‘‘not near Portland,

The rise of cyberinfrastructure and grand challenges for eCommerce 287

123

Maine’’), Portland Cement, the Battle of Portland, Portland Orange (the color

emitted by ‘‘don’t walk’’ pedestrian crossing signals in North America), and

Portland Sheep, among others.

However, Wikipedia is not exhaustive. It provides information of general interest

and is not a substitute for specialized collections of knowledge on subjects such as

science. Multiple efforts (such as the Web of Science) are underway to aggregate

scientific documents into an archive that can be searched by subject, journal, or

author. Subject searches are aided by keywords, journals are uniquely identified by

their ISSN, but author identity has proven much more difficult to determine. There

may be multiple scientists with the same name and journals often do not list middle

initials, or may only list first initials making it hard to distinguish between Mike

Smith and Mark Smith or between Mike James Smith and Mike John Smith.

Wikipedia’s disambiguation feature is of limited value for scientific research since

its database includes only a small subset of scientists. The Wikipedia disambig-

uation page for ‘‘Mike Smith’’ lists only 4 of the many researchers with that name.

Extensive experience with the Wikipedia disambiguation page as a communication

genre is increasing pressure for the creation of a researcher identification scheme

with unique scientist identifiers that can be used in searchable scientific archives.

However, such a scheme could also be misused and some scientists are concerned

that such a database could threaten the security of those engaged in controversial

research.

Privacy and disambiguation are but two of many difficult challenges one can

point to in the realm of eCommerce, Big Science, and government. Interestingly,

blogging and citizen science represent two co-emergent eCommerce and Big

Science trends. Blogging allows individuals to easily generate information and post

it on the web. Others can get notifications of updates, add information, or aggregate

it with other information sources. Citizen science enables individuals not previously

part of the Big Science production system to generate data, post it on the web for

aggregation with other information sources, perform analyses, and make discov-

eries. Cornell’s eBird project allows citizens to report bird sightings by species,

location, and time, creating a large scale database that has been used to better

understand migration patterns, responses to climate and habitat change, and

population dynamics. Both blogging and citizen science face questions about

accuracy, integrity, and control over resulting intellectual property. It is not yet clear

under what formal, social or psychological contracts this work will be done.

Some trends reflect more technical aspects of these transformations. As the US

moves toward eMedicine, the Federal Government is encouraging development and

implementation of interoperable digital health systems to improve medical data

handling. These efforts rely heavily on lessons learned from the digital medical

records systems created by the Federal Government to provide healthcare to active

service military personnel and veterans (a centralized system) and by Petsmart’s

Banfield veterinary medical sevice (a more franchised, distributed model).

eCommerce, Big Science and the federal government are also facing issues of

environmental sustainability and green IT. Large server farms and high performance

computer sites (many of which are supported by the Department of Energy) can

have equally large carbon footprints. Plans for some future supercomputer sites are

288 S. J. Winter

123

expected to require the electrical output equivalent to that of a small nuclear reactor.

Finally, as the gaming segment of eCommerce has grown, chip architecture has

moved to multi-core processors. The supercomputers used for modeling and

simulation-based Big Science are also based on multi-core chips and efforts are

underway to maximize performance through new software, networking and data

architectures and methods. Of course, this commoditization of chips has also

reduced costs and allowed experimentation with multiple configurations that would

not previously been possible.

Some transformations in the social organization of work are disrupting entire

economic sectors with serious implications for eCommerce, Government, and Big

Science. One such transformation is affecting both news journalism and the

scientific promotion and tenure systems. Under the old news production model, a

journalist or editor would develop an idea for a story. The journalist would research

the topic and write the news article. A news outlet would check the article for

accuracy, edit it for clarity, aggregate related stories and publish them. Reader

subscriptions and advertising paid the news outlet for its content and the news outlet

paid the journalist. Libraries further aggregate the published stories into searchable

archives and curated them. Now, search engines such as Google and Yahoo

aggregate and provide searchable content at no charge to the reader and advertising

revenue goes to the aggregators, not to the content providers. This disruption in the

previously established reward system has severely reduced the revenue stream for

news outlets and journalists. Established newspapers and magazines are closing,

journalists are being fired, and the future of the free press (long a cornerstone of US

democracy) is being questioned. However, the business model for aggregators such

as Google and Yahoo depends upon content providers. If news outlets and

journalists disappear, there will be nothing to aggregate and Google is beginning to

invest in existing news outlets to ensure its content stream.

Big Science is facing an analogous problem, with somewhat different surface

features. An earlier era of science confined research to the scientist’s laboratory and

provided an established path from effort to payoff in the form of journal publication,

promotion and tenure. A single scientist developed hypotheses to be tested,

collected data, analyzed the data, wrote-up and submitted an article describing the

results. Scientific conferences and journals used peer review to determine the

accuracy and importance of the research, edited articles for clarity, aggregated and

published them. Academic institutions rewarded the scientist for publications

through promotion and tenure. Journals were compensated for their costs through

advertising and subscriptions from readers and academic libraries, which further

aggregated, archived and curated scientific articles. Authorship was assumed to

represent responsibility for hypothesis development, data collection, data analysis,

and write up of results. Promotion and tenure committees assessed and rewarded

these activities with proportionally reduced rewards for each author on a

collaborative paper.

Big Science has seen a disaggregation of several of the early steps in the research

process and a disruption in the reward structure similar to that faced by news outlets.

By definition, Big Science is too big to be performed by a single investigator and

requires a more elaborate division of labor, but the rewards for scientific activities

The rise of cyberinfrastructure and grand challenges for eCommerce 289

123

have not yet been disaggregated and distributed accordingly. Hypothesis generation,

data collection, analysis and the write-up of results are often performed by different

groups, and currently rewards are distributed unevenly among them. Hypotheses are

developed by domain scientists or by multi-disciplinary teams of domain scientists.

Big Science relies on shared infrastructure (such as distributed sensor nets, shared

equipment like the LHC, or citizen scientist portals), but a new era in which most or

all of the scientists in a field must use a single laboratory or shared dataset turns

conventional work practices upside down.

In Big Science, data collection and analysis becomes a multi-stage process

performed by different sets of scientists, some of whom are responsible for

designing, deploying or maintaining initial sensors, collecting raw data, cleaning

data, early stage data processing, or data distribution. Other computational scientists

may write specialized algorithms and software for data analysis to test hypotheses

developed by the domain scientists, who then write up the research results. In this

new model of Big Science, authorship is often limited to the subset of scientists

involved in the downstream tasks of hypothesis testing, thus under-rewarding those

involved in the early stages of data collection and cleaning. Much like the

journalists who are no longer rewarded by news outlets, but are crucial to the

success of news aggregators like Google, the contributions of scientists who create

and maintain the data and computational infrastructure are not rewarded by

promotion and tenure committees, but are crucial to the success of scientific

research. Some academic communities have addressed this issue by including all of

the thousands of principal investigators involved in an experiment in the list of

authors, resulting in papers with hundreds of authors, and the question of who gets

credit for the work becomes confused as promotion and tenure committees try to

assess the value of 1/300th of a publication. As with the transformation of

journalism, new organizational forms and protocols for scientific research must

evolve to supplement or replace those already in place, and the new order of things

faces serious opposition from incumbent interests.

In some instances, eCommerce, government, and Big Science can work

synergistically, as in the case of Iridium. Iridium is a private company that provides

mobile satellite voice and data communications with service in even the most remote

locations (e.g., across oceans, polar regions) through its network of 66 cross-linked

low earth orbit satellites (Iridium 2010). Most consumers do not require such

complete coverage, but some government entities do. The DoD’s global commu-

nications system relies on an Iridium dedicated gateway. Iridium has recently

announced a new generation of low earth orbiting satellites with space for a

secondary sensor payload that could accommodate scientific instruments. Continual

sensing of the entire Earth’s surface and atmosphere has the potential to improve our

understanding of global climate change and of scientific phenomena in the

magnetosphere and ionosphere with implications for astronomy and satellite

communications systems. Working together, this private company, government

agency, and Big Science may accomplish more than they could achieve on their own.

Though differing in surface features, Big Science, eCommerce and government

are undergoing similar disruptive transformations catalyzed by technological

advances. All three are struggling to address issues and opportunities presented by

290 S. J. Winter

123

the data deluge including privacy, disambiguation, citizen participation, interoper-

ability, environmental sustainability of ICT, adaptation to changing hardware, and

emerging disconnects between the division of labor and the allocation of rewards. At

any given time, one or another will be out in front on a particular issue and each can

learn from the others. All three are creating and reacting to the same underlying

cyberinfrastructure. As in the case of Iridium, Big Science, private companies, and

the government can work together to achieve synergy.

5 Extreme grand challenges for eCommerce

Clearly, the eCommerce revolution is well under way, but faces significant

challenges, many of which are emerging from the gathering data deluge. However,

none of these represents a Grand Challenge. The Extreme Grand Challenges for

ecommerce are shared jointly with government and Big Science, and it is this shared

responsibility that makes them true Grand Challenges.

Government played a crucial role in the birth of the Internet; Big Science

provided the venue for the rise of the Web; but the Web itself grew up mainly

around other forms of enterprise, many of them commercial. Without this interplay

among government, Big Science and commerce, eCommerce could not have arisen

and could not continue. Each played an important role at different stages in the

development of this transformational phenomenon. The Extreme Grand Challenge

is to develop an understanding of the intertwined roles of these three actors in

innovation and commercialization. Neither commerce nor Big Science could have

created the Internet; only the government had the necessary motivation, resources,

and organization. Neither DARPA nor commerce could have prepared the nascent

Internet for commercial uses; only Big Science had the necessary commitment to

open access and open standards. Neither government nor Big Science could have

provided the multitude of services we’ve come to rely upon; only commerce has the

flexibility to create and adjust to market demands. Knowing when in the

developmental trajectory of an innovation each of these sectors can best make a

contribution and how those contributions should be structured is an Extreme Grand

Challenge for eCommerce, government, and Big Science.

eCommerce, Big Science, and government depend upon advancements both in

enabling technologies and in the social conventions of work. It is popular to talk

about the accretion of continuous incremental adjustments that culminate in a

tipping point after which change is swift and inevitable, but identifying the

underlying forces behind those adjustments can be elusive. The Extreme Grand

Challenge is to see beyond the surface features of the issues facing eCommerce, Big

Science, and government and identify their fundamental similarities. True

transformations often unfold over decades in response to underlying disruptive

transformations catalyzed by changes in ICT. Distinguishing between surface

features and underlying transformations requires a much longer time frame and

deeper analysis than is common in IS research. To meet this Extreme Grand

Challenge, we must be willing to adopt what has been called ‘‘the long now’’

encouraging the long view and long-term thinking (Brand 2010).

The rise of cyberinfrastructure and grand challenges for eCommerce 291

123

Finally, The Cold War that ran between 1948 and the early 1990s saw much

innovation in information technology, based in research conducted at the

intersection of military, industrial and academic interests. This paper has shown

that many of the issues facing eCommerce continue to lie at the intersection of

government, industrial and academic interests. In short, we are all in this together

and could benefit from working together to address these common concerns.

However, learning to work together and determining the optimal form for such

collaboration will also be an Extreme Grand Challenge.

Commerce, science and government have changed dramatically since the

beginning of the Cold War in the mid-1950s. Globalization, flattened hierarchies,

and the growth of networked organizations, and doing business ‘‘on-line’’ are the

hallmarks of eCommerce. They are features of Big Science as well. Both grew out

of improved cyberinfrastructure, nurtured in turn by the DoD, NSF, and commercial

interests. This cyberinfrastructure holds great promise, but presents significant

challenges to privacy, data management, sustainability, and existing social and

organizational arrangements. These disruptive transformations are affecting eCom-

merce, Big Science and government. Achieving the Extreme Grand Challenges of

determining when each should take the lead in creating and nurturing new

innovations, adopting the long now to recognize common underlying transforma-

tions, and learning to work together to address shared concerns are important steps

toward an even brighter future for Big Science and eCommerce.

References

Abbate J (1999) Inventing the internet. MIT Press, Cambridge

Aspray W, Williams BO (1994) Arming American scientists: NSF and the provision of scientific

computing facilities for universities, 1950–1973. IEEE Ann Hist Comput 16(4):60–74

Atkins D (2003) Revolutionizing science and engineering through cyberinfrastructure: report of the

National Science Foundation Blue-Ribbon advisory panel on cyberinfrastructure. National Science

Foundation, Arlington, VA, Jan 2003. http://www.nsf.gov/cise/sci/reports/atkins.pdf

Atkins D (2007) Cyberinfrastructure vision for 21st century discovery. National Science Foundation,

Arlington, VA. www.nsf.gov/pubs/2007/nsf0728/nsf0728.pdf

Brand S (2010) About long now. http://en.wikipedia.org/wiki/Long_Now_Foundation

Bush V (1945a) As we may think. The Atlantic magazine, July 1945, pp 101–108. http://www.

theatlantic.com/magazine/archive/1969/12/as-we-may-think/3881/

Bush V (1945b) Science: the endless frontier. A report to the president. Government Printing Office,

Washington, D.C. http://www.nsf.gov/od/lpa/nsf50/vbush1945.htm

Cerf V (1993) How the internet came to be. In: Aboba B (ed) The online user’s encyclopedia. Addison-

Wesley, Reading

Crocker SD (1987) The origins of RFCs. In: Reynolds J, Postel J (eds) The request for comments

reference guide (RFC 1000), August

Edwards PN (1996) The closed world: computers and the politics of discourse in cold war America. MIT

Press, Cambridge

Fisher VA (2006) Physicist shapes modern thought. The Harvard Crimson, June 3

Friedman T (2005) The world is flat. Farrar, Straus, and Giroux, New York

Harnad S (1995) Publicly retrievable FTP archives for esoteric science and scholarship: a subversive

proposal. In: Okerson A, O’Donnell J (eds) Scholarly journals at the crossroads: a subversive

proposal for electronic publishing. Association of Research Libraries, Washington, DC, June 1995.

http://www.arl.org/scomm/subversive/toc.html

292 S. J. Winter

123

Hoare T, Milner R (2005) Grand challenges for computing research. Comp J 48(1):49–52

Hofacker CF, Murphy J (2000) Clickable world wide web banner ads and content sites. J Interact Market

14(1):49–59

Hughes T (1998) Rescuing Prometheus: four monumental projects that changed the modern world. Pantheon Books, Random House

Iridium website (2010) http://www.iridium.com/About/IridiumNEXT.aspx

King JL (2010) Project SAGE, half a century on. Timelines section. ACM Interfaces 17(5):53–55

King JL, Grinter R, Pickering JM (1997) The rise and fall of Netville: institution, infrastructure, and the

Saga of a cybserspace construction boomtown in the great divide. In: Kiesler S (ed) Culture of the

internet. Ehrlbaum Publishers, NJ, pp 3–34

Lambert L, Poole HW, Woodford C, Moschovitis CJP (2005) The internet: a historical encyclopedia.

MTM Publishing, USA

Lutters W, Winter S (in press) Virtual organizations. In: Bainbridge W (ed) Leadership in science and

technology: a reference handbook

Lynch DC (1993) Historical evolution. In: Lynch DC, Rose MT (eds) Internet systems handbook.

Addison-Wesley, Reading

Naughton J (2000) A brief history of the future: the origins of the internet. Orion Books, London

NSF (1950) National Science Foundation Act of 1950, Public Law 81-507 (as amended on-line at

http://www.law.cornell.edu/uscode/42/usc_sup_01_42_10_16.html)

NSF (2010) NSF-ACCI task force on cyber science and engineering. National Science Foundation,

Arlington, VA [url]

Reid RH (1997) Architects of the web: 1, 000 days that built the future of business. Wiley, NY

Traweek S (1988) Beamtimes and lifetimes: the world of high energy physicists. Harvard University

Press, Cambridge

Wolfram S (2002) A new kind of science. Wolfram Media Inc, Champaigh

The rise of cyberinfrastructure and grand challenges for eCommerce 293

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.