Remarks

Dr. Arden L. Bement, Jr. Director National Science Foundation Biography "From Concept to Confluence: Framing our Cyberinfrastructure" SBE/CISE Shared Cyberinfrastructure Workshop Closing Dinner Remarks Airlie Center Warrenton, Virginia March 16, 2005

Good evening, everyone, and thank you for that generous introduction. I'd like to thank the Academy... and let me also thank Drs. Ruzena Bajscy and Fran Berman for inviting me to speak with you tonight. I am reminded that an attribute of an NSF Director is the vision to produce what will happen in the future and the ability to explain why it didn't happen.

I hope you have enjoyed this workshop and that it has yielded a fruitful exchange of ideas. In reviewing the list of breakout sessions, I was struck by the obvious synergy between the computer sciences and the social and behavioral sciences. The session on malevolent uses of cyberinfrastructure, for example, demonstrates how social, behavioral and computer scientists can work together to address privacy and security issues that increasingly affect everyone. As technology pervades daily life, the security of personal data becomes critical. This is especially true as cyber-attacks, such as phishing ¹, become more sophisticated and widespread. The ChoicePoint ID theft scandal, for example, is estimated to affect nearly 140,000 people ². The other sessions are equally relevant, tackling such thorny issues as the impact of cyberinfrastructure on the economy. The progress you've made this week will reap benefits far into the future. I want to spend a few minutes this evening describing that future.

I've entitled my remarks, "From Concept to Confluence: Framing our Cyberinfrastructure," because it describes the tasks ahead of us. As you've no doubt discovered, the concept of cyberinfrastructure, or CI, means different things to different people. To a materials engineer like me, cyberinfrastructure might involve a high-speed computational grid for complex structure-properties-process modeling. To an astronomer, a robust CI may mean remote access to telescopes around the world. To an economist, cyberinfrastruture probably includes more accessible software tools for the analysis of complex data. Every field has a concept of what CI should be and what features it should have. Our next step is to assemble these different concepts into a cohesive whole, generating a confluence of ideas, talent and effort focused on the next generation of cyberinfrastructure.

My vision for that cyberinfrastructure is simple but perhaps a bit idealistic. It will ultimately join the ranks of the electrical grid, the interstate highway system and other traditional infrastructures. A sufficiently advanced CI will simply work, and users won't care how.

With today's electrical grid, I can turn on my CD player, my neighbor can run the air conditioner, and a teenager in Maryland can fire up a video game. We are all using the electrical grid for different things, to perform tasks based on our individual interests, without affecting one another. I'm not worried that my neighbor's air conditioner uses more electricity than my CD player; the power grid adjusts to the load automatically. Likewise, my neighbor and I can use different appliances to meet our individual needs, and as long as the appliances conform to certain electrical standards, they will work reliably.

That adaptive functionality is the ultimate model for cyberinfrastructure: researchers having easy access to the computing, communication and information resources they need, while pursuing different avenues of interest using different tools. Cyberinfrastructure will become standardized, but will also enable the development of dedicated networks and "science portals"; that is, customized, community-specific interfaces that provide data display and analysis tailored for a particular community or field. Middleware will operate in the background, providing security, data validation and other functions, but without involving or concerning the end user. Like the electrical grid, cyberinfrastructure will simply do its job. However, a mature cyberinfrastructure adds new dimensions of utility that surpass analogies to the electrical grid.

Ideally, a capable cyberinfrastrucure will allow the knowledge generated by researchers, whether from cognition studies or astrophysics modeling, to be cataloged and indexed so that it can be searched and shared. I understand this is particularly important to the social and behavioral sciences. Your research data takes myriad forms, from statistics to survey results to video clips. Imagine being able to mine that rich collection of data to unearth hidden trends or discover unforeseen relationships. While producing new knowledge, this type of data mining contributes to national priorities, like the job creation and wealth generation that underlie economic growth.

As many of you know, hiring, training and retaining employees is one of the biggest expenses to any business. Consider the benefit of a "best practices" database for hiring new staff. A small business owner searches the database for companies in the same industry, of about the same size, and finds video excerpts or other data that describe successful hiring techniques. Using these "best practices" helps the owner hire the most qualified people the first time, simultaneously creating jobs and saving money for future investment.

This scenario also illustrates an important aspect of this workshop--exploring the impact of cyberinfrastructure on society. This forum brings together scientists who specialize in human behavior with scientists who create technology that people must learn to use. As one program officer described it, this is "where the silicon meets the carbon." The ubiquitous, utilitarian cyberinfrastructure I've described will forever change our society and cultural environment. This sweeping change confirms the idea that we always return to, that humans are not only an important component, but are, perhaps, the key component, of any successful cyberinfrastructure. In fact, the Blue-Ribbon Advisory Panel report, better known as "the Atkins report," explicitly identifies personnel--people--as an enabling component of cyberinfrastructure³.

Frankly, we forget the "human factor" at times. It is, ironically, human nature to do so. However, we need to recognize how easy and potentially disastrous it is to overlook the role that people play in developing and operating a cyberinfrastructure. The tsunami disaster a few months ago provides two examples of how people can either defeat or surpass the best technology.

Many people are questioning why there wasn't more advanced warning prior to the tsunami. As you may know, this questioning has escalated into legal action. Observers theorize that the underlying earthquake was promptly detected. In an era when e-mail can traverse the globe in seconds, a warning could have been quickly dispatched. The New Zealand Herald ⁴ reports that scientists at the Pacific Tsunami Warning Center in Hawaii could not use an advanced warning system to alert the affected countries because the governments of those countries had not agreed on a procedure for sharing warning information. In other words, poor human organization hindered the effective implementation of cyberinfrastructure.

Conversely, a 10-year-old girl saved many lives without using technology. The February 2005 newsletter of the Association of American Geographers ⁵ recounts the story of Tilly Smith, from Surrey, England. Tilly was vacationing with her family in Thailand. While playing on the beach, she saw the waters suddenly recede. Tilly learned in school that this phenomenon was a precursor to a tsunami. She alerted her parents, and ultimately, the beach was evacuated. The tsunami struck minutes later, but there were no casualties on that particular beach. Tilly's story demonstrates the importance of education, whether it pertains to geography or to cyberinstructure. As we create the next generation of cyberinfrastructure, we need to teach people how to adopt and implement it effectively, and provide appropriate incentives for doing so. The finest technology will be, at best, useless, and at worst, dangerous, without a comprehensive educational component and a strong incentive structure.

NSF's CI-TEAM initiative will address some of these education issues. The program is designed to equip the science and engineering workforce to create, augment and fully utilize cyberinfrastructure over the long term. CI-TEAM also anticipates that advanced cyberinfrastructure will make state-of-the-art science and engineering research more accessible across geographic, institutional and economic boundaries. To that end, the first solicitation seeks demonstration projects that can be replicated and scaled to a national level. Proposals should also focus on building partnerships among diverse organizations, such as K-12 and undergraduate institutions, industry and the non-profit sector.

CI-TEAM meshes nicely with our Next-Generation Cyberinfrastructure Tools program. This project aims to develop tools that can analyze data gathered from individual subjects and organizations - two areas of special interest to social and behavioral scientists. Proposals specializing in the analysis of organizational data will develop tools that integrate heterogeneous forms of data into a unified, searchable database. Proposals geared toward the analysis of individual data will create tools that enable automatic collection, annotation and analysis of disparate data sets. All of the proposals must include techniques for preserving confidentiality of research subjects, as well as incentives to share information--again, two serious concerns of social and behavioral scientists.

Both CI-TEAM and Next-Generation Cyberinfrastructure Tools, along with our other investments, will contribute toward the cyberinfrastructure vision I shared earlier. However, investment is not enough. The cyberinfrastructure I foresee will require new thinking, some risk-taking and unprecedented collaboration across the disciplines, as this workshop demonstrates. Senior management at NSF realized that a different approach was needed to effectively guide and cultivate those collaborations.

Earlier this month, I established the Cyberinfrastructure Initial Implementation Working Group, to define NSF's role in developing the cyberinfrastructure enterprise. The group has representation from all of NSF's research directorates, plus the Offices of Polar Programs and International Science and Engineering, and is chaired by Priscilla Nelson on my staff. The group will analyze NSF's current CI-related portfolio, looking for commonalities, gaps and new opportunities. It will also evaluate the cyberinfrastructure work conducted by others in the public, private and academic sectors, and develop a strategy for internal and external collaborations. Finally, the working group will develop a charter for the formation of a CI Management Panel, which will carry out the long-term planning and management of NSF's cyberinfrastructure investments.

These new management vehicles, combined with continued investments, will increase the momentum generated at this workshop. I want to assure you that this won't be just another event, where you gather to discuss lofty ideas that never reach fruition. We do not want to squander your time or your talent. Your work, and the efforts of your colleagues at other workshops, is of paramount importance to both the Foundation and to the nation. Together, we will assemble the ideas, features and concepts you've developed into a confluent, concerted push toward a ubiquitous, enabling, visionary national cyberinfrastructure. Thank you for your commitment, time and your attention. I'll be happy to take any questions.

¹ Phishing (pronounced fishing) is an attack in which a hacker (also called a phisher) attempts to elicit personal information from a user by sending e-mail that mimics official correspondence from a real institution, such as a bank or credit bureau. The phisher hopes that the recipient will be deceived and will provide personal information that can be used for identity theft and other malicious purposes.
Return to speech.

² "ChoicePoint: More ID theft warnings." CNN Money. Feb. 17, 2005.
Return to speech.

³ Page 5, 3rd paragraph of the Atkins report.
Return to speech.

⁴ Lean, Geoffrey. "Fatal gap in tsunami warning system." New Zealand Herald, Jan. 17, 2005.
Return to speech.

⁵ "10-Year-Old Geography Student Saves Lives." From the Meridian. Association of American Geographers, Feb. 2005.
Return to speech.