U.S. computing leadership under threat, says House science chair
U.S. Rep. Lamar Smith, the chair of the House Science, Space and Technology committee, believes the U.S. is losing its leadership position in computing. That may sound good for funding, but Smith's solution is controversial.
Source: Patrick Thibodeau
Smith wants to shift National Science Foundation (NSF) R&D funding from efforts he believes may be "frivolous" or "low risk," to "biology, physics, computer science and engineering."
What Smith is outlining is a change in research priorities, such as moving research funding from anthropology to engineering. But complex systems require insights from human behavior, and this is expertise outside of computer science. Tech firms are hiring people with skills at interpreting behavior, including anthropologists.
America's "pre-eminence in several fields is slipping," argues Smith, citing supercomputing in particular, in an essay he released this week. "Last year China launched the fastest supercomputer in the world, five times faster than any supercomputer in the United States," he said.
The Texas Republican isn't proposing increasing federal R&D spending, arguing that the U.S. is now spending more than any other country. What Smith didn't note is that U.S. R&D investment has been declining for years as percentage of GDP, and that China is expected to surpass the U.S. in the next few years.
Smith questioned why the NSF, which has about $7 billion to spend, funded, for instance, a project "to study textile-making in Iceland during the Viking era." This research, which has received about $840,000 to date, is more complex than Smith makes it out to be.
The textile research project has a number of goals, including offering "a window into how social, cultural and economic systems will adapt to contemporary climate shifts and assist in preparing us for potential changes," according to the NSF grant.
The principal researcher on this project, Michele Hayeur Smith , an anthropologist and research associate at the Haffenreffer Museum of Anthropology at Brown University, said in an email that "my research examines the roles of women in the economies of the European nations of the North Atlantic -- Norway, Scotland, Iceland, Denmark and its territories. I am focusing on textiles for this work because these islands' main export products for nearly 800 years were the cloth these women made. By examining their roles, through the products they made and the imports that competed with them, I am documenting how Europeans colonized and settled the north and how imports, in a globalized economy, undermined women’s roles and these societies' economies."
She continued, "The textiles were the main economic product of the successful Viking societies that colonized in the North Atlantic islands and led to the earliest European contact with North America, around AD 1000. My work, therefore, is tied directly to the national narrative of America's exploration and discovery," said Smith.
This NSF-funded research is also helping train people who are increasingly important to high-tech companies, said Ed Liebow, executive director of the American Anthropological Association.
The skills and research methods used by anthropologists are, for instance, helping to develop self-driving systems. If a person is leaning into a street, is he just scratching something or preparing to cross? A human driver can make a reliable judgment but teaching a driverless system to understand the intent is an entirely different matter, said Liebow.
The hallmark of anthropological training is "participant observation," which means immersing yourself as the observer and learning the unwritten rules to understand context, Liebow said. These are important in many areas, including cybersecurity, privacy and user interface development.
Intel's Genevieve Bell, an anthropologist and Intel fellow, may be one of the best-known people working in the tech field, and her work helps bring understanding to the relationship people have with computing.
The NSF is spending about $7 billion to support science research, and an estimated $995 million will be spent on computer science and engineering this year. The NSF funding helps support a number of academic supercomputers, but the funding for the largest systems comes from the U.S. Department of Energy.
The "NSF is enormously important to the field," said Liebow. Its research improves the skills and research capabilities of people in the field, he said. He believes demand for people with those skills is on the rise.