2007-08-29: On Networking and Cooperation

From Datafedwiki

Contents 1 Introduction 2 Science 2.1 Diana Rhoten: The Dawn of Networked Science 2.2 Lynn Preston, NSF: Core Researchers as Network Hubs 2.3 Timo Hannay, Nature: Web 2.0 and Science 3 Game theory 4 Business 5 Biological 5.1 MNowak: Cooperation and Evolution, - NY Times Science profile 6 Literature on Cooperation

[edit] Introduction

<math> \mathbf{P(k) \sim k^{- \gamma}}, </math>

[edit] Science

[edit] Diana Rhoten: The Dawn of Networked Science

By DIANA RHOTEN, Chronicle of Higher Eduaction

This summer, Edward Melcarek, a Canadian in his 50s with a master's degree but no Ph.D., solved yet another scientific problem through InnoCentive, an open, online platform that connects world-class scientists with companies to collaborate on complex scientific challenges. InnoCentive works by broadcasting companies' problems to a network of scientists, who receive cash prizes of $10,000 to $100,000 for submitting winning solutions. The money is incentive enough to people like Melcarek, who works on the problems in his spare time, but a drop in the bucket for a company, compared with the costs of hiring researchers to meet all its rapidly evolving scientific needs.

Founded by the pharmaceutical company Eli Lilly in 2001, InnoCentive now has over 120,000 scientists and engineers from more than 170 countries interested in solving problems. It represents the latest stage in a fundamental transformation in scientific research. Over the past 50 years, the old model of Big Science ­ a term coined by the physicist Alvin Weinberg ­ has gradually given way to what's commonly called Team Science. Now an even newer model is emerging: what I call Networked Science.

Whether through Web-based ventures like InnoCentive or cyberinfrastructures like TeraGrid, the world's largest supercomputing grid, research on countless big-ticket scientific issues ­ including bioinformatics, cosmology, climate change, and nanotechnology ­ is increasingly being conducted by broad networks of scientists from different disciplines and distant locations.

That emerging phase of science is made possible by advanced technologies in computing and networking, and it usually relies on different aspects of openness: open systems, open-source software, open participation. Though the shift is still in its early stages, it has important implications for scientists and engineers, and also for the universities that house them and the governments that support them.

During World War II, the Manhattan Project put an end to the age of bench-top science. No longer were scientific investigations only the exclusive province of lone researchers, toiling in their labs over petri dishes and Bunsen burners. Now, alongside bench-top science, we had Big Science, with its array of highly ambitious projects: the first space vehicles, particle accelerators, nuclear reactors, and the like.

But Big Science wasn't big merely because of the scale of its ambitions. It was big because of its very nature ­ top-down, hierarchical, vertical. It depended on dedicated facilities and expensive equipment. It involved astronomical financial outlays and was oriented toward projects that focused on one or two large-scale, long-term research programs heavily tilted toward experimentation and instrumentation.

Because of their scale and organization, the programs that marked the era of Big Science radically altered the culture of scientific practice. Programs like the Hubble Space Telescope or the International Thermonuclear Experimental Reactor tended to recruit hundreds of individuals ­ mainly from the physical sciences and engineering ­ and employ them in house, full time, and for life. The arrival of Big Science transformed the scientist from an independent, curiosity-driven researcher into a member of a tightly managed research enterprise. Instead of having one or two authors for a scientific publication, multiple authors became the norm.

Even in its heyday, Big Science was controversial because of the high costs and huge bureaucracies that it spawned. The resources ­ technical, financial, organizational ­ that it required put much cutting-edge research out of the reach of many institutions, including universities that could not afford or even gain access to the necessary apparatus. Influence over the course of science became increasingly centralized in national agencies and international consortia.

But although protests against Big Science projects sponsored by the military erupted on some American campuses as early as the 1960s, it wasn't until the 1990s that Big Science came under serious political fire. At that point, the government faced dire budgetary constraints, and politicians' demands for science to be conducted on a more economical scale grew louder, while scientists urged greater decentralization.

At about the same time, the scale and complexity of research problems outside the physical sciences and engineering ­ particularly in the life sciences ­ were reaching new heights. As scientists turned their attention to mapping the human genome and assessing global warming, they began to call for a different form of scientific collaboration ­ a new model that would be more bottom-up, horizontal, and multidisciplinary.

Enter Team Science. Whereas Big Science was shaped by the general properties of instrumentation, Team Science, in its best form, is more tailored to the parameters of the specific investigation. And whereas Big Science revolved around practitioners wedded to a particular institution or facility, Team Science is often centered on researchers whose main ties are to the given intellectual challenge. Thus, although Team Science projects are often anchored in a large cooperative research center, they tend to draw talent from multiple organizations, with researchers typically participating in a project while also fulfilling responsibilities to their home departments and institutions.

A prime example of Team Science is the Human Genome Project, whose specialized researchers came from different disciplines and institutions to work on divergent but complementary aspects of a grand challenge. The key assumption of Team Science is that project-based teams can do science smarter, better, and faster than can permanently structured groups from one discipline or organization. In the case of the Human Genome Project, that assumption proved correct, with the team of biologists, computer scientists, engineers, and physicists from more than 20 laboratories in six countries completing the sequence ahead of schedule and under budget.

Even as the Human Genome Project demonstrated the efficiency of Team Science, however, it also foreshadowed the potential of the next stage: Networked Science. For all that the project gained from its flexible and horizontal organizational model, it benefited even more from the effects of Moore's Law, which postulates the exponential, clockworklike rise in computing power. Indeed, according to the scientists who led the project, had it not been for mind-boggling advances in data crunching and networking, they could not possibly have finished even on schedule.

Digital technologies, of course, are not new to science. Computer networking was first developed as a tool for scientists and engineers; e-mail and file transfers have long supported their collaboration. And as early as 1989, William A. Wulf, then head of the Directorate for Computer and Information Science and Engineering at the National Science Foundation, imagined a "collaboratory," or ­ in his words ­ a "'center without walls,' in which the nation's researchers can perform their research without regard to geographical location."

At the turn of the 21st century, that notion has become a reality and Networked Science a possibility. Vastly more sophisticated and powerful capabilities ­ such as high-performance computing, remote access to scientific instruments, shared databases, and astonishingly advanced simulation and visualization technology ­ allow dispersed networks of researchers to exchange data and pool computational resources across both time and space.

On the one hand, Networked Science is about the rise of federally financed collaboratories like the Biomedical Informatics Research Network, which take advantage of cyberinfrastructure developments to distribute work among different groups of researchers. But on the other hand, it is also about the emergence of organizations like InnoCentive, which use the Web to enlist individuals in problem-solving efforts on a more ad hoc basis.

In either case, the rise of the virtual and the return of the individual are what most set the new way of doing science apart from its predecessors. Big Science was centralized and geographically concentrated, and Team Science is decentralized and geographically distributed. But both are organized around brick-and-mortar institutions, and both are essentially closed loops, with participation limited to and defined by programs or teams.

Networked Science is a more loosely coupled and open system. In it, geographic location, institutional affiliation, and even professional reputation are becoming less relevant, while technological connection, personal motivation, and informal interaction are increasingly important.

The history of science has shown that innovative solutions to difficult problems can arise when knowledge from one discipline is applied to another field. Likewise, much has been written about the role of serendipity, such as the chance encounter in the hallway between researchers apparently without much in common that yields a revolutionary breakthrough. Team Science recognizes the importance of serendipity and tries to institutionalize it. But Networked Science takes that idea a step further, using cyberinfrastructure to create a virtual hallway in which the doorways ­ wide enough to accommodate all the scientists who want to pass through ­ lead to labs and offices containing every discipline under the sun. By providing that space, unachievable in the physical world, being virtual can actually surpass being there.

What scientists do in the virtual hallway may be collaborative (working together toward a common end) or simply cooperative (sharing resources for individual ends). Networked Science gives researchers more choice, control, and ownership of their work than they would have in the Team Science model. In short, Networked Science brings us from bench-top science to laptop science.

Networked Science represents a major structural and cultural redesign of how we produce knowledge. Its full potential won't be realized without adaptation on the part of all the players ­ or without some risk. As scientists increasingly work without disciplinary or organizational boundaries, at ever faster speeds and with exponentially broader networks, they will need to learn new strategies for identifying problems, allocating time, and directing their research. The openness of Networked Science also raises questions of authority and issues of security.

For universities the challenge will be to encourage the transition to Networked Science even if it might weaken the loyalties of scientists to them. Institutions will need to establish clear, standardized intellectual-property arrangements that allow researchers to easily take full advantage of new partnerships. In Team Science, a university needed to assess a scholar's contribution to a paper with multiple authors; with Networked Science, the question will be how to assess the contributions to and values of items like problem solutions, database compilations, and computational simulations, not just scholarly publications.

And the government, as the only entity with sufficient resources and the necessary dedication to the public good, will need to lead the way in building the high-tech infrastructure that undergirds Networked Science. At the same time, however, the nature of the work conducted along the virtual hallway must be shaped by the researchers and the problems they investigate, not by the government ­ lest we find ourselves returning to Big Science.

In the 21st century, innovation will be indispensable to the wealth and health of nations. Yet with science and technology racing ahead at mind-bending speed, it's impossible to know precisely where innovation is likeliest to occur, what sorts of projects will be necessary, or which scientists should be involved. Tomorrow's discoveries will depend less on our capacity to manage the biggest accelerator, the largest research center, or the fastest computer, and more on our ability to create fluid, responsive networks of scientists and engineers.

Diana Rhoten is founder and director of the knowledge-institutions program at the Social Science Research Council. http://chronicle.com Section: The Chronicle Review Volume 54, Issue 2, Page B12

[edit] Lynn Preston, NSF: Core Researchers as Network Hubs

While all these advances required interdisciplinary teams, we need to look beneath them to understand the realities of just who gets involved in interdisciplinary research at what stages in their careers. Diana Rhoten from the Social Science Research Council let a study of communication in centers. None were ERCs but her study reveals some interesting patterns that impact all centers, gaps across the disciplines still exist. I chose this slide because it illustrates the interconnect among different levels of a research team. You will note the strong collaboration in the center and left of the slide. That’s led by the key leaders of the center and a few key graduate students. They fan out in strong collaboration with a team of obviously interactive graduate students. However, look where most of the professors and associate professors are...out at the edges and I can’t find the assistant professor. This represents the reality of what we are dealing with because of the culture in academe and its reward system.

[edit] Timo Hannay, Nature: Web 2.0 and Science

Web 2.0 is a potent buzzword that provokes enthusiasm and cynicism in roughly equal measures, but no one can seriously doubt its influence. Its manifestations include:

• “The web as a platform”
• The Long Tail (e.g., Amazon)
• Trust systems and emergent data (e.g., eBay)
• AJAX (e.g., Google Maps)
• Tagging (e.g., del.icio.us)
• Peer-to-peer technologies (e.g., Skype)
• Open APIs and ‘mashups’ (e.g., Flickr)
• “Data as the new ‘Intel Inside’” (e.g. data from MapQuest)
• Software as a service (e.g., Salesforce.com)
• Architectures of participation (e.g., Wikipedia)

Wither the scientific web?. Over the last 10 years or so, much of the discussion about the impact of the web on science – particularly among publishers – has been about the way in which it will change scientific journals. The web’s major impact will be on the way that science itself is practiced.

The barriers to full-scale adoption are not only (or even mainly) technical, but rather social and psychological. This makes the timings almost impossible to predict, but the long-term trends are already unmistakable: greater specialization in research, more immediate and open information-sharing, a reduction in the size of the ‘minimum publishable unit,’ productivity measures that look beyond journal publication records, a blurring of the boundaries between journals and databases, reinventions of the roles of publishers and editors, greater use of audio and video, more virtual meetings. And most important of all, arising from this gradual but inevitable embracement of technology, an increase in rate at which new discoveries are made and exploited for our benefit and that of the world we inhabit.

[edit] Game theory

[edit] Business

[edit] Biological

[edit] MNowak: Cooperation and Evolution, - NY Times Science profile

Nowak has argued that cooperation is one of the three basic principles of evolution. The other two are mutation and selection. On their own, mutation and selection can transform a species, giving rise to new traits like limbs and eyes. But cooperation is essential for life to evolve to a new level of organization. Single-celled protozoa had to cooperate to give rise to the first multicellular animals. Humans had to cooperate for complex societies to emerge.

While cooperation may be central to evolution, however, it poses questions that are not easy to answer. How can competing individuals start to cooperate for the greater good? And how do they continue to cooperate in the face of exploitation? To answer these questions, Dr. Nowak plays games.

Nowak developed the neighborhood model, for human cooperation. “I’m much more likely to interact with my friends, and they’re much more likely to interact with their friends, so it’s more like a network.”

Nowak identified the conditions when cooperationit can arise: B/C>K. That is, cooperation will emerge if the benefit-to-cost (B/C) ratio of cooperation is greater than the average number of neighbors (K).

Another boost for cooperation comes from reputations. When we decide whether to cooperate, we don’t just rely on our past experiences with that particular person. People can gain reputations that precede them. If reputations spread quickly enough, they could increase the chances of cooperation taking hold. Players were less likely to be fooled by defectors and more likely to benefit from cooperation.

“You help because you know it gives you a reputation of a helpful person, who will be helped,” Dr. Nowak said. “You also look at others and help them according to whether they have helped.”

[edit] Literature on Cooperation