by Lauren Crowley
Within the next decade computers will operate at speeds of over one thousand trillion operations per second. Such machines are the first steps in a scientific revolution in how we use machines to help us think, a revolution that will have a profound effect not only on the sciences, but on society in general.
This imminent revolution in supercomputing was the theme of a three-day multidisciplinary conference, "Supercomputing and the Human Endeavor," hosted by the Woodrow Wilson Center and the Los Alamos National Laboratory June 13 - 15. This conference brought together leaders from disciplines ranging from religion to neuroscience, as well as government and business leaders to discuss the scientific, social, and ethical implications of the supercomputing revolution: how human beings will use tremendous computing power to improve our understanding of complex phenomena such as weather, economic trends, and brain function.
Both the opportunities and potential dangers of supercomputing were examined with a special focus on public policy implications. In his opening remarks, former member of Congress and Wilson Center Director Lee Hamilton noted the importance of the policy relevance of the conference: "In today's fast-paced political environment, few policymakers have the time to consider complex long-term issues such as the future of supercomputing. At the Wilson Center, one of our goals is to help policymakers think ahead by identifying crucial challenges of the future and bringing the relevant experts to Washington. I am pleased that several U.S. policymakers will be speaking at the conference." A detailed agenda listing all speakers and topics is available. A full transcript of the conference will be available here soon.
Putting Supercomputing in Context
Dr. Stephen Younger, Senior Associate Laboratory Director for National Security at Los Alamos and one of those who conceived of this conference, set the context for the dialogue. According to Younger, "Supercomputing and its potential for large scale simulation involves nothing less than changing the way we use machines to help us think." Several supercomputers exist today that operate at speeds exceeding three trillion operations per second, faster than several thousand of the fastest desktop computers. And further development has been breathtaking. For example, by the end of the decade, speeds will exceed one thousand trillion operations per second. Such speeds will enable us to interact with computers differently and to use them for tasks that we cannot yet even comprehend.
Until recently, computers have been used to perform numerical calculations. Within the past twenty years, computers have taken on the additional role of simulating reality, with applications ranging from entertainment to the training of jet fighter pilots. Soon, supercomputers will be able to solve problems that touch upon the most basic issues of humanity.
At one thousand trillion operations per second, we will be able to construct a reasonably accurate model of at least a higher animal brain. Large scale simulation of brain function will offer an unprecedented insight into the origin of mental illness. Having the ability to perfectly simulate a human brain though raises some probing questions: Will understanding the operation of the human brain make us less human? What are the religious implications of understanding the detailed processes of life? Will these new computers offer the opportunity to create a fundamentally new type of intelligent life form?
Among the benefits of supercomputing Younger listed a greater precision in psychotherapy, the ability to design airplanes, model car crashes, and understand complex observations from telescopes.
Social and Ethical Implications
These benefits are not without a cost. Although most of the conference presenters were from different fields, a common thread throughout the presentations was the need to address the ethical, legal, and social implications of supercomputing. Pulitzer-prize winning author and historian Richard Rhodes warned of the "revenge of unintended consequences" that often results from great scientific breakthroughs. In her luncheon address, Dr. Bernadine Healy, President of the American Red Cross, cautioned, "We must enter this new technological revolution with a strict, disciplined, well thought out sense of the ethical, legal, and social implications. The greater the impact on human beings, the greater the imperative."
Thankfully, we already have a guide for how to accomplish this task in the Human Genome Project (HGP), another revolutionary breakthrough in science and technolgy. From the beginning of HGP, 5 percent of the budget was dedicated to ELSI, the ethical, legal, and social implications. We must do the same for supercomputing.
Speaking from a theological perspective, Bishop Ronald Haines claimed that technology is already creating ethical challenges faster than we can identify morally justifiable solutions and that this will only accelerate.
Broadening the Dialogue
"Science must benefit humanity and must be driven by human consciousness and wisdom," Healy said. According to both Healy and Raines, social responsibility and transparency—including the general public in the dialogue—is crucial in this process. But supercomputing is not currently in the public lexicon. "It will be a challenge for policymakers to educate the public on this scientific revolution," said Healy.
Despite the Potential for Bad, We Must Move Boldy Forward
Speaking from the perspective of the policymaker, Senator Jon Kyl (R-AZ) stressed the need for the United States to be the leader in supercomputing technology, for the sake of our national security. "We better be able to use what technology and science offers because our enemies will, regardless of the moral implications." Kyl noted that it is precisely because we as a society have not shied away from technology in medicine that the U.S. has made such great advances. According to Kyl, we must do the same with supercomputing. The United States must have preeminence in this field if we wish to determine whether the technology will be used for good or evil. David McCurdy, offering his perspective as a former Congressman and current industry leader (McCurdy is the President of the Electronics Industry Alliance) observed that whereas government used to drive technological innovation, today industry is taking the lead and is in large part responsible for our networked society.
Using Supercomputers in Modeling
In his presentation on modeling the physical universe, Dr. Larry Smarr, director of the California Institute for Telecommunications and Information Technology pointed out that computers enable scientists to do advanced computations and visualize phenomena that help us understand the universe. Computers do the intricate math, help us see three-dimensional projections of two-dimensional outer space, map the birth and death of stars, chart statistics, and project galactic happenings on a huge scale. Such visuals allow scientists to go beyond the capability of telescopes and create and compare simulations of millions of galaxies.
To make a predictive science of biology, we need to be able to model a living planet. Right now humans are manipulating the environment, but we do not know the consequences of our actions. According to Dr. P. Bryan Heidorn of the Graduate School of Library and Information Science, we cannot achieve predictability, and thus control of our environment, until we understand all organisms on our planet. Supercomputers can help us both quantify and characterize all living organisms through autoclassification and pattern recognition, mobile computing and telecommunications, and lastly through very large databases to store all the details on our ecosytem. Understanding our ecosystem will enable us to better control it and keep it productive.
Supercomputing and Health
In his presentation titled "Utilization of Supercomputer Technology to Address Human Needs," Dr. John Mendelsohn, president of the University of Texas, M.D. Anderson Cancer Center described how supercomputing will improve health through the development of new knowledge about the molecular processes that produce life and disturb it. Current technologies are enabling cancer researchers to create new therapies which target particular molecules and molecular processes, detect correlations that will lead to valid predictions of disease presence, and select the most appropriate targeted therapy for not only a particular disease, but for a specific genetic makeup.
Dr. Andrew Von Eschenbach, president-elect of the American Cancer Society spoke too of how new technologies are enabling cancer researchers to develop a new paradigm of cancer treatment—from that of "search and destroy" to a "target and control" strategy. Supercomputing can enable us to process the data that describes the many interactions involved with cancer development and spread. From this data, we can then develop targeted therapeutic strategies. According to Von Eschenbach, "In this new paradigm we can begin to think of cancer as vulnerable, not as overwhelming."
Supercomputing Cannot Trump Humanity
While supercomputing and artificial intelligence demonstrates almost mind-boggling capacity for innovation and scientific progress, certain areas of our lives will always remain controlled by human intelligence, as they should. Former U.S. Secretary of Defense James Schlesinger and Librarian of Congress James Billington each provided examples of this fact. Schlesinger noted that in the military realm, genius, instinct, and luck, not artificial intelligence, would continue to determine the outcome of battles. Billington cautioned against neglecting more traditional modes of information exchange, from books to oral histories, as well as the content of what is being transmitted over our fiber optic lines.
Bernadine Healy noted that while visiting a hospital ravaged by the recent floods in Texas, it was humans that safely evacuated 500 patients with no lives lost, during an electrical outage. "Technology" she concluded, "cannot make moral decisions."
The significance of the supercomputing revolution and its diverse implications call for a multidisciplinary approach. The conference revealed that no one field has all the answers and that we are stronger and more effective when we work together. The conference also pointed out that it is almost as important to ask the right questions as it is to find the right answers. Some of the most probing questions discussed here were: What is the appropriate government investment in supercomputing? How do we validate a computation to make policy on it and what does it mean to depend on a simulation to drive policy?
No matter what the perspective of the speaker—policymaker, bishop, scientist, historian—all those who spoke at the conference made reference to the great moral responsibility we have in insuring that supercomputing is used for good, not evil. As Steve Younger concluded, "Let them say of us: 'They took up the challenge bravely, and they made a better world.' "
Pioneering Conference Offers New Insights on Supercomputing and Its Impact on the Human Endeavor
- Jun 13, 2001
by Lauren Crowley