OCTOBER 8, 2009
Basic Research Loses Some Allure
By DON CLARK and CHRISTOPHER RHOADS
WSJ
This year's three Nobel laureates in physics carried out their groundbreaking work while working at corporate research labs decades ago. The conditions that led to their breakthroughs are harder to come by today. Big companies now tend to spend less on such basic research, investing instead on projects likely to pay off quickly. Much of the action in long-term research has shifted to universities, often working in collaboration with government agencies and companies.
Industry executives and researchers differ about how well the new research model will ultimately serve high-tech industry. In semiconductors, for example, university researchers that receive funding from companies and the U.S. government have assumed much of the responsibility for refining alternatives for the silicon now used to make chips. But communications-industry experts are less optimistic about matching the contribution of Bell Labs, the legendary research organization that was part of AT&T Corp. before the telecom monopoly was broken up in 1984. Two of Tuesday's winners, Willard Boyle and George Smith, did their award-winning work there in the 1960s.
"Bell Labs was the leader in research not just for the U.S., but for the world," said Adam Drobot, chief technology officer of Telcordia Technologies Inc., a telecom-equipment maker in Piscataway, N.J., that inherited a large portion of its 400-person research staff from Bell Labs. "We lost that -- and in no way are we making up for it." Bell Labs could afford to employ hundreds of researchers because anything it developed in the field of communications was used only by its parent AT&T, because there were no U.S. competitors. Now, competition is fierce in communications. Most companies aren't willing to risk as much on long-term developments that might benefit other companies.
(snip)
Companies that still conduct basic research include Microsoft Corp., International Business Machines Corp., Xerox Corp. and Hewlett-Packard Co. H-P's research arm, for example, has been doing in work in areas such as nanotechnology and computer algorithms that could more efficiently sift the Internet for huge amounts of public data to predict business trends. Prith Banerjee, director of HPLabs, said about two-thirds of the labs' projects are relatively short-term initiatives that are intended to produce results within about five years. The remaining third is basic research that may not pan out for a decade or more.
(snip)
IBM's scientists last won a Nobel Prize in 1987 for work on high-temperature superconductors. A company spokesman said that some IBM researchers could still get Nobels for work they did decades ago on memory chips and moving atoms one by one. This week, IBM disclosed some new work it is doing using semiconductors to develop a cheap way to map individual human's DNA. Even IBM has changed its research focus. Last year, 70% of its patents were for software and services, reflecting its shift away from computers and semiconductors, where pure science plays a bigger role. A spokesman said that the percentage of hardware-related patents was much higher 10 years ago.
(snip)
http://online.wsj.com/article/SB125487196131869225.html (subscription, maybe)
Printed in The Wall Street Journal, page A8