REDMOND, Wash. The U.S. government needs to spend more money on long-term research and multi-disciplinary education and open its doors wider for international researchers. That was the conclusion of a panel of experts convened at a Microsoft Research event here Monday (July 16).
Those conclusions will be spelled out in the report of a government task force to be published this summer. "The big take away is we need to think more audaciously and stimulate the government to support that thinking," said Dan Reed, a panelist and member of the group that drafted the report.
The report from the President's Council of Advisors on Science and Technology (PCAST) reviews the programs of some 14 government agencies that spend about $3.1 billion on research each year. It is the first independent update on government technology spending since 1999.
The document weighs the global competitiveness of the US technology industry and education. One of its conclusions is "the research fuse is too short and incremental and we need to be more audacious," said Reed, director of the Renaissance Computing Institute sponsored by three universities.
"Today's feedback loop is fairly conservative, and tight funding leads it to be even more conservative which leads to incremental-ism," he added.
Craig Mundie, chief research and strategy officer at Microsoft and another panel member agreed.
"One of the things we don't fully appreciate is that there is a fundamental change in computer architecture coming," said Mundie referring to the move to multi-core processing. "Ultimately we will need machines the likes of which we don't understand today and we don't think there is enough radical work on this in many computer science departments," he said.
Mundie singled out the Defense Advanced Research Projects Agency (DARPA) in particular for focusing too much on short-term projects rather than the big technology breakthroughs it once drove.
"If DARPA doesn't go back to where it once was we will have to look elsewhere because we are not hitting the right balance," Mundie said. "We don't have enough people doing the core computer science work," he added.
The PCAST report will call for the government to simplify visa procedures for international graduate students looking to work in the US. It will also ask for more fellowship money for those students, said Reed.
Mundie said Microsoft recently opened a research office in Vancouver, British Columbia, in part to overcome problems hiring talented overseas researchers in the US. "Over the last few years it has become increasingly difficult to get the people we want to hire to be able to stay in the country," said Mundie in an interview with EE Times.
Reed and other panelists also called for more focus on multi-disciplinary research.
"I really do believe computational thinking will be fundamental like reading, writing and arithmetic," said Jeannette Wing, a panelist and an assistant director for a branch of the National Science Foundation
The NSF will invest $52 million in a program that promotes such multi-disciplinary work, Wing said. "We have to spread this to K-12 education because if we wait until students are undergrads it's too late," she added.
Ed Lazowska, a computer science professor at the University of Washington and moderator of the panel said the NSF program could grow to $250 million within five years.