Twelve years ago, my fellow Physics Today editor Toni Feder and I visited Bell Labs in Murray Hill, New Jersey. Whether Bell Labs had passed its peak by then, I couldn't tell, but it was certainly doing innovative science under its new parent, Lucent Technologies.
During the day-long visit, Toni and I met two of pioneers of quantum cascade lasers, Claire Gmachl and Federico Capasso, who demonstrated how one of their tiny devices could light a match. Tony Tyson showed us the latest images of gravitationally lensed galaxies that he'd obtained using his CCD cameras.
We also met Arun Netrvali, the Bell Labs president. Netravali is one of the leaders in digital technology. He invented, among other things, the digital compression algorithm that underlies high-definition television. Unlike the other researchers we met, he's a theorist.
Despite the obvious practicality of Netravali's work, the notion that an industrial research lab should employ theorists might seem strange. In fact, theorists have a long history—and, I hope, a long future—in industrial research.
Some industrial theorists, such as Claude Shannon, Rolf Landauer, and Charles H. Bennett, lay down frameworks for experimenters to exploit. Shannon invented information theory at Bell Labs. Landauer and Bennett, both IBMers, extended information theory into the quantum realm.
Other industrial theorists, such as John Bardeen and George Hockham, help guide the research of their experimentalist colleagues. Bardeen made crucial contributions to the invention of the transistor at Bell Labs. Hockham explained the opacity measurements on optical fibers that his coworker Charles Kao made at Standard Telecommunication Laboratories.
Although industrial labs around the world are shrinking, theorists are still employed by them. Hewlett Packard, for example, runs a social computing laboratory at its research campus in Palo Alto, California. The lab's director, Bernardo Huberman, used to work in condensed-matter physics.
Indeed, the ranks of industrial theorists, past and present, are exalted enough that newly graduated theorists should give thought to working for a company, not a university. Last week IBM posted a job ad for theoretical physicist to work in superconducting qubits.
When Toni and I met Netravali, he told us why he liked working at Bell Labs. "It's a problem-rich environment," he said.
