Computer Science Department celebrates 40 years of changing the world

By David Orenstein

Through ground-breaking research, teaching and often entrepreneurship, Stanford computer science faculty, students and staff have had an impact on technology that is broad, deep and unique. The department's optimistic and ambitious nature was readily apparent among the more than 400 faculty and alumni who gathered to celebrate its 40th anniversary at the Arrillaga Alumni Center on March 21.

"Our department culture was to hire people we thought would change the world," said Ed Feigenbaum, the Kumagai Professor in the School of Engineering, Emeritus, paraphrasing an explanation university President John Hennessy offered last year for why the department has been so successful. Both men are former chairs of the department. "For over 40 years our faculty and our graduates changed the world of computer science and continue to do so today."

Although entrepreneurship is not a criterion for success in the department, its prevalence among the faculty and students reflects the desire of Stanford computer scientists to make that world-changing mark, added Bill Dally, the department's current chair and holder of the Willard R. and Inez Kerr Bell Professorship. "It is a delicate balance because you don't want academic values to be compromised by corporate ones," he said in an interview with a reporter at the event. "A university is about creating knowledge and a company is about creating value. I think Stanford has found an appropriate balance. What we really want to do is have an impact."

Indeed, Stanford computer scientists, often in collaboration with electrical engineers, musicians, biologists and others from Stanford and elsewhere, have achieved historic advances in a wide variety of areas.

Take the Internet, for example. In 1969, the Stanford Artificial Intelligence Lab (SAIL), founded by Professor Emeritus John McCarthy, became one of the first nodes on the ARPAnet, the precursor of the modern Internet. Only five years later, then Professor Vinton Cerf helped develop the Transmission Control Protocol, networking software that still governs some of the Internet's essential workings.

In 1984, Computer Science Department facilities director Leonard Bosack and economics alumna Sandy Lerner started Cisco Systems to commercialize Internet router hardware technology developed at Stanford. A decade later, Jerry Yang and David Filo, electrical engineering graduate students in the Computer Systems Lab, started Yahoo! Then, in 1998, computer science graduate students Larry Page and Sergey Brin founded Google. Both companies now compete to help millions of Internet users make sense and make use of the vast contents of the World Wide Web. (The first U.S. website, by the way, was at the Stanford Linear Accelerator Center.)

The department also has produced academic and commercial innovations in other areas of computing, such as in technology's intersection with the arts. In 1967, for example, music doctoral student John Chowning, working at SAIL, developed a synthesizer he later licensed to Yamaha. Fast forward to the nineties, when Professor Marc Levoy developed the technology to scan statues such as Michelangelo's David so precisely that they could be digitally preserved in exquisite 3-D detail. In 2004, Professor Pat Hanrahan shared a technical Academy Award (his second) for research on simulating how light scatters in translucent materials.

In robotics, achievements at SAIL span from the 1969 development of a programmable robot arm by mechanical engineering graduate student Victor Scheinman to the 2005 victory of Stanley the autonomous car, which was the first to cross a rugged 132-mile desert course without any human intervention.

In systems, the department's impact has included a role in the beginnings of legendary companies such as Sun Microsystems (started by several students including Andy Bechtolsheim and with key software development from computer science Professor Emeritus Vaughan Pratt), Silicon Graphics (started by former Professor Jim Clark) and MIPS (co-founded by Hennessy).

In medicine, Stanford computer scientists also have made advances. In 1967, for example, Feigenbaum, then research associate Bruce Buchanan, chemistry Professor Carl Djerassi and genetics Professor Joshua Lederberg (a Noel laureate) demonstrated the DENDRAL project, a so-called "expert system" that helped compute molecular structures from mass spectrogram data. In recent years, Professors Jean-Claude Latombe, the Kumagai Professor in the School of Engineering, and Professor J. Kenneth Salisbury each have made substantial advances in computer-assisted medicine.

The department's history of broad relevance traces back to the vision of its founder, George Forsythe, who launched the department as its first chair on Jan. 1, 1965. Recruited by then Provost Fred Terman and Associate Provost Albert Bowker to help establish computer science as its own discipline, Forsythe had an expansive view of what computing could become. In 1961, he said: "Enough is known already of the diverse applications of computing for us to recognize the birth of a coherent body of technique, which I call computer science. Whether computers are used for engineering design, medical data processing, composing music or other purposes, the structure of computing is much the same."

Four members of Forsythe's founding team, Gene Golub, the Fletcher Jones Professor of Computer Science; William Miller, a professor emeritus and former provost; Feigenbaum; and McCarthy, joined Don Knuth, the Fletcher Jones Professor of Computer Science, Emeritus, and Nils Nilsson, the Kumagai Professor in the School of Engineering, Emeritus, on a panel at the 40th anniversary event, where they shared their recollections of the department's history.

"We all are the core memory of the Computer Science Department," Feigenbaum quipped, referring to an old-fashioned means of making computer memory involving magnetic cores.

When the department launched from within the Department of Mathematics, its expertise was in scientific computing and numerical analysis (Cleve Moler, one of Forsythe's doctoral students, later went on to create Matlab, a leading scientific computing software program). But Forsythe sought to expand the scope of the department to include not only the computation expertise of professors like himself, Miller and Golub, but also the artificial intelligence expertise of professors such as McCarthy and Feigenbaum.

In 1967, on the advice of Knuth, Forsythe hired Bob Floyd, who pioneered the idea that programs could be proven to work. Knuth, known worldwide as the inventor of the TeX and Metafont systems for formatting online text and as the author of the multivolume set The Art of Computer Programming, came aboard in 1969.

In 1985, Nilsson, who at SRI led the development of the first mobile robot, "Shakey," joined the department as chair. That year he led the department's migration from the School of Humanities and Sciences to the School of Engineering, both to establish an undergraduate major in computer science and to engender more collaboration and less competition with the Electrical Engineering Department. Overlaps with computer science had become abundant.

Today well over one-third of the department's students are undergraduates. Much of the well-regarded curriculum they follow, particularly the introductory course Computer Science 106A, which is often the largest course at Stanford, is credited to Professor Eric Roberts, the John and Cynthia Gunn University Fellow in Undergraduate Education.

Within several years of joining the School of Engineering, the department, which had long been split up among many cramped buildings, had plans well under way for a new building. The Gates Building opened in 1995, unifying the department under one roof for the first time.

Today the department's 43 faculty members and 765 students have the challenge—at Stanford it is usually viewed as an opportunity—of making their own mark on the field. Several faculty speakers at the March 21 conference touched on broad themes as they explained their specific research directions.

Among those themes was the overwhelming amount of data piling up in the world. Database expert Jennifer Widom, a professor of computer science and of electrical engineering, suggested that the ratio of data that's been read to data that's been written is approaching zero as time goes on. On the bright side, however, more of that data will be available (if anyone can find the time to read it) via the Internet. By 2016, she added, databases will be able to manage themselves to some extent, even as disparate information remains difficult to integrate.

Artificial intelligence expert Daphne Koller, an associate professor of computer science and 2004 MacArthur "genius grant" recipient, focused on a particular data-intensive set of applications in her remarks. Looking at biological data, she sees an important future for computer scientists to help biologists unlock the mysteries of cells and genomes hidden in mountains of data.

Koller's colleague in artificial intelligence, computer science and electrical engineering Associate Professor Sebastian Thrun, said robotics holds great potential to improve the quality of people's lives. The technology behind Stanley the robot car could help save lives by reducing human error, and it could increase the carrying capacity of existing highways by allowing cars to drive closer together. A robotics project led by computer science Assistant Professor Andrew Ng seeks to unify many subdisciplines of artificial intelligence to create a generally intelligent servant robot to assist the elderly and others in daily tasks.

Many advances will depend on the continued advance of computer processing performance, but Dally cautioned that progress will require a fundamental rethinking of the processor-architecture ideas that engineers have depended on for about 60 years. Old methods are literally starting to burn out as power-hungry processors begin to generate unacceptable levels of heat. Rather than pushing processors to run faster and hotter, engineers may have to build chips with scores of specialized processor "cores" that can carry out their tasks in parallel without being pushed to the brink of a mini-meltdown.

Dan Boneh, an associate professor of computer science and of electrical engineering, struck a sobering theme by pointing out that security threats and software vulnerabilities are on the rise today and that new kinds of computer attacks—on cell phones, electronic voting systems and even cars—likely will emerge in the future. Computer scientists can develop tools to find software bugs, protect user privacy and quarantine infected networks, Boneh said, but it will still be up to people and policymakers to deploy them.

In the day's final panel, alumni and faculty who've started some of the scores of spinoff companies from the Computer Science Department talked about the university's relationship with industry, the startup culture and other factors that affect the uniquely productive climate for technology entrepreneurship in Silicon Valley. The panel was a veritable all-star team of entrepreneurs including Yahoo's Yang and Filo, SGI's Clark (he later helped found Netscape and Healtheon), Sun's Bechtolsheim, Wall Street techno-mogul David Shaw, Associate Professor Mendel Rosenblum (cofounder of software company VMWare) and Mark Horowitz, the Yahoo! Founders Professor in the School of Engineering (co-founder of memory maker Rambus).

Moderator John Markoff of the New York Times, also an adjunct professor in the Communication Department, asked the panelists pointed questions about the Valley's resilience during economic downturns and the effects of government cutbacks on basic research funding. He elicited sometimes sharp criticism from some panelists of recent shifts in federal research funding.

Markoff also asked panel members to describe how they became inspired to entrepreneurship while they were here.

"It was the environment Stanford provided to us," Filo said. "Part of it is being right in the middle of Silicon Valley. It was about having professors who have gone off and started companies. You are surrounded with professors, with students, with all these people who have either done it or are thinking about doing it and that's what rubbed off on us."

The infectious nature of success is a theme that kept coming up during the event, often as an explanation for why the department has sustained its track record over four decades. "It's because of what you might call a virtuous cycle," Nilsson offered. "Because we were excellent, excellent students came to us. Because we had excellent students, it wasn't too hard to get excellent faculty."

David Orenstein is the communications and public relations manager at the Stanford School of Engineering.

Editor Note:

Photos from the event are available on the web at http://www-db.stanford.edu/~hector/photos/2006/40thSel/album.htm.

-30-