Gates At Berkeley: Thoughts On Research, Overseas

During a conversation with Dean Richard Newton before an audience of Berkeley students, Gates said computer science and biotechnology hold the greatest promise of increasing people's quality of life in coming decades, but that the United States needs to "rededicate itself to the research university system" by putting more money into schools.

Gates said he's disappointed that corporate research spending in the United States is lower than it has been historically, and that the failure of Bell Labs and Xerox to turn their research into successful products set a "bad precedent" for future investment.

U.S. companies need to look overseas for innovation, too, he said. If China and India become rich, it won't be bad for the United States, because research isn't a "zero-sum game," Gates said. "China and India are the big change agents in the years ahead." Being able to tap advances from those "risky new areas" would also benefit the United States.

And it's not just the cost of PCs that's limiting technology adoption in India and other developing countries--it's the cost of Internet connectivity. Microsoft Research is working on a networking project that could one day connect villages in India wirelessly, he said.

In computer science, the problems of the next 10 to 20 years will be more interesting than those of the past 20, Gates predicted. The software industry's "holy grails" include writing the "perfect program" that's defect-free, understanding speech, and imbuing computers with artificial intelligence. PCs need to be easier to use and capable of feats such as recording a person's memories using sound, graphics, and text. "If we look at the personal computer today, it's a glass half full," he said.

Other problems near the top of Gates' list include recognizing people in photos, understanding the contents of documents for better searches, and building reliable software. Modern programming languages such as Microsoft's C# and Sun Microsystems' Java are "a little better" than mainframe-era Fortran, he said, "but not much."

Gates advocated some types of open-source software licenses as good teaching tools--he reminisced about rooting through a Dumpster with Microsoft co-founder Paul Allen to find discarded printouts of the operating system code for Digital Equipment's PDP-10 minicomputer in order to better find bugs in the code. "When I was first using computers, the source code was very attractive," he said. "There are many cases where having the source code out there is a fantastic thing."

In the spirit of being at UC Berkeley, Gates said Microsoft advocates open-source licenses in the style of the one that governs Berkeley Systems Distribution Unix, which arose from the university, has helped computer-science students for decades, and led to the launch of Sun Microsystems. Open-source licenses that let users create new commercial works, build companies around them, and create jobs and pay taxes benefit the economy, he said. Others, such as the GNU General Public License that governs Linux, "just create more free software." Microsoft believes those should be used narrowly.

When asked which Microsoft products he's proudest of, Gates said he has a personal interest in seeing Tablet PCs become more widespread and Microsoft's Xbox video-game console include new interactive features, such as the ability to assemble online audiences of spectators for games.

Gates also discussed his philanthropy, noting that he donates money to try to correct "market failure"--if governments or companies invest heavily in an area, his wealth won't make as much of a difference. For example, Gates said he donates money to try to cure malaria and other third-world diseases, but not first-world ailments such as cardiovascular disease. According to Gates, he once believed that he'd wait until he was retired from Microsoft and in his 60s to become a philanthropist, and thought making money and giving it away at the same time would be problematic. "I thought it would mess up my discipline," he said. He noted that hasn't happened.

This story courtesy of InformationWeek.