Spirit of Entrepreneurship
CHRIS QUIRK
Around 2,500 years ago, the pre-Socratic philosophers began their attempts to understand the phenomena of the world in a new way. Turning away from myth, they breached a passage into wide-open and fertile intellectual terrain, appealing to reason and formulating arguments and ways of asking questions that presaged what we now call science. Plato absorbed and expanded these efforts so convincingly that his writings defined Western philosophy for a couple of millennia — to such a degree that English philosopher Alfred North Whitehead characterized the history of the discipline as “a series of footnotes to Plato.”
Intrepid first arrivals into unexplored scientific territory often have an outsized effect, and the field of computer science is no exception. The early work of the pioneers of computer science at Carnegie Mellon University built the inventions and created the intellectual frameworks that have mapped out the field as it exists today, and produced many of its most important landmarks and milestones along the way.
Besides being in the right place at the right time, a major reason for CMU’s success and influence stems from the fact that early on, the faculty, administration and students created a culture of experimentation that encoded innovation into the DNA of the school. From the start, that culture embraced ingenuity, pragmatism, invention and also a sense of service to the scientific community and society at large.
Whether it was operating systems for Apple or Microsoft, computer vision or self-driving technology or innovations in speech recognition, Martial Hebert, the dean of the School of Computer Science, believes SCS has played a pivotal role in the development of almost every area in the field and produced an outsized number of major figures and innovations.
“Part of the reason we’ve always been at the leading edge of research and education in computer science goes back to the fundamental aspects of creating new things,” said Hebert. “There is a spirit of entrepreneurship that is our academic mission.”
At the Core
As far back as the 1950s, the core of intellectual firepower at CMU created the basis of what would later become the School of Computer Science, acting as a magnet that attracted top talent. Herb Simon, Allen Newell and Alan Perlis were at CMU simultaneously, and all three were later winners of the A.M. Turing Award. Simon, Newell and colleague Arthur Samuels also attended the Dartmouth Summer Research Project on Artificial Intelligence, the 1956 conference that gave birth to the field.
Computer science was so new that it didn’t have a proper name until the mid-1960s. When Raj Reddy, the Moza Bint Nasser University Professor of Computer Science and Robotics, began his Ph.D. program, no one knew what to call it. “I went to Stanford to do computer science in 1963, but they hadn’t created the name yet.” Reddy said CMU was already one of the major poles in the nascent field, and Newell and his colleagues were a big reason. “It was a small department, but in almost all the areas that we are now working in, like AI or machine architectures, we were already doing that work back in the ‘60s.”
Exploring All Avenues of Computer Science
The Wild West quality of computer science at that time meant there were plenty of open vistas to explore and a “do it yourself” component that went with it. “Essentially, many of the things that we take for granted here today, 50 years ago did not exist,” said Reddy. “So if you thought of something, you had to make it yourself.”
Richard Rashid, Former Computer Science Professor and Co-founder of Microsoft Research
Richard Rashid, who went on to co-found Microsoft Research after serving as a professor of computer science for 12 years, recalled that there were virtually no off-the-shelf tools or hardware at their disposal. “We had to develop everything on our own. You were always making your own tools because there really weren’t very many standardized products that you could take advantage of. It was perfect because we were writing the operating systems and the microcode, and in the end, we went from having very minimal facilities to having the best,” he recalled. “For me, it was an ideal environment. I just love programming to the basic metal. The idea that I could animate a piece of hardware with my mind was just intoxicating.”
It was certainly more work; Reddy said the opportunities were endless. “There were so many problems to work on, but everything we touched turned into gold.” The benefits these early innovators created were more broadly enjoyed because of the extra efforts they made to ensure their inventions were compatible with existing systems or could be adapted for other uses. “We built a wide variety of software tools, for example, that allowed us to do new things, but we added versioning and revision tools so other universities and outside organizations could manage them in different ways,” said Rashid. “People were willing to do the extra work necessary so the ideas and technology that we created could actually be used by others. There was very much a sense that we were building something for the real world, not just for the academic value of it.”
The Proof Is in the Code
Beyond the VIP roster of faculty and researchers who have etched their names into the history of computer science, empirical proof abounds of the profound influence of SCS in almost every mobile phone on the planet. Within them you will find code written by Rashid or Avie Tevanian (SCS 1985, 1988) decades ago. “If someone had asked me back in 1983 as I was working on some of this code, ‘Do you realize this is someday going to be running on a cell phone?’ I would have replied, ‘What’s a cell phone?’” said Rashid. “People laugh at me when I tell them that there’s software on their Apple watch that I wrote back in the 1980s.”
Rashid and Tevanian created the Mach operating system, parts of which help to run computers and devices to this day. Tevanian moved on to NeXT Inc. after earning his degree, and employed Mach as the foundation for his programming work there. In addition, SCS veterans founded or co-founded some of the tech companies that have become household names: Adobe Inc., the seminal desktop publishing (and more) company; Sun Microsystems, the early information systems giant; and of late Duolingo, the ubiquitous online language tutor, with an estimated 116 million users last year. (See sidebar for more.)
SCS still regularly spins off one or two dozen companies each year, proof that the spirit of innovation is alive and well. “There was a sense back then that we were creating technologies that were going to have huge impacts. That we were really pushing the envelope at that particular point in time. And the department was very practical in its ideas about how to create things that other people were going to use, Mach being an example,” Rashid said.
The list of spinoffs and major tech companies founded by CMU faculty and students would fill a large volume with more than 1,600 entries. Here is a selection of some notable ones, past and present.
Charles Geschke (SCS 1972) co-founded Adobe Systems.
Vinod Khosla (ENG 1978) and Andy Bechtolsheim (ENG 1976) co-founded Sun Microsystems in 1982.
Jim Levy (GSIA 1966, IM 1966) co-founded and served as CEO of Activision video game publisher.
Avie Tevanian (SCS 1985, 1988) designed the NeXTSTEP operating system, the precursor to Apple’s macOS, and he served as vice president of engineering for NeXT Inc.
Professor Alfred Spector, Jeffrey Eppinger (SCS 1988), Michael L. Kazar (SCS 1984) and Dean Thompson (ENG 1991) co-founded Transarc, which commercialized the Andrew File System. AFS anticipated and laid the foundation for cloud computing.
Professor Richard Rashid co-founded Microsoft Research.
Former faculty member Luis von Ahn (SCS 2005) founded reCAPTCHA, co-founded and served as CEO of Duolingo and was named a MacArthur fellow in 2006.
Professor Nesra Yannier founded NoRILLA, a new educational app that engages young students with feedback from a virtual gorilla.
Spirit of Autonomy
Besides having a transformational brain trust in the right place at the right time, the early faculty and administration in the department (and since its founding in 1988 as the School of Computer Science) have cultivated a sense of freedom and autonomy — supported through policy. “The school was built on the idea of academic entrepreneurship,” said Hebert. “One example is that we have had a number of faculty to whom we have granted partial leave to create their companies. We’ve been very flexible in ways like that.”
”The departments we have now, like the Robotics Institute, or the Machine Learning Department, and even the degrees we offer, were also created a bit like creating startups,” said Hebert. “Like entrepreneurship, for the most part they aren’t based on top-down decisions. It’s individual faculty coming together to create new things, just like creating new companies.”
“You empower people to do things instead of asking permission all the time.”
— Raj Reddy, Moza Bint Nasser University Professor of Computer Science and Robotics
Reddy cites the reasonable person principle — the notion that if a reasonable person would do something you are proposing, you have the license to do it — as central to the SCS philosophy. “It’s an empowerment principle, and one of the keys to the growth of the school. You empower people to do things instead of asking for permission all the time. It has been our fundamental motto in the whole
of the School of Computer Science from the beginning to now,” Reddy said. “When I was the dean of the school, I said, don’t come to me with any problems. You’re there to solve the problems, so solve them! To their credit, they did.”
Rashid recalled the sense of openness and collaboration from his time at SCS. “I was given a huge amount of freedom, and at the same time, I got incredible mentorship. People like Allen Newell and Raj Reddy were just tremendous mentors to me and gave me lots of great advice,” said Rashid. “Not just advice about being a professor, but also about how to deal with other people, and how to deal with the world. They themselves were tremendous creators, and when I went to start Microsoft Research, my model was the Computer Science Department at Carnegie Mellon. It worked tremendously well.”
Between the variety of disciplines within SCS and the university broadly, there are frequent collaborations across silos to find new ways to solve problems and develop products. “What’s unique here at CMU is that it’s fundamentally part of the culture, so we have all these opportunities to create new things, because there’s a deep understanding of what the challenges are on both sides, and a dedication to meet them,” said Hebert.
Creating Value
With university budgets increasingly under siege, it’s easy to lose sight of the value academic research creates. The 2025 global market for information technology, comprised of everything from artificial intelligence to wearables, is estimated at $5.7 trillion. The market caps of top tech firms Alphabet, Apple and Microsoft are all north of $2 trillion each, and a 2023 study by the Information Technology and Innovation Foundation found that the technology sector drove one-third of the nation’s economic growth from 2012 to 2022. Work done at CMU and other schools built the foundation for that value and helped create the products that continue to drive it.
Economic analysis bears out the evidence. Citing a comprehensive study that nondefense R&D government spending increased the nation’s economic output by $2 for every $1 invested, Glenn Hubbard, chair of the U.S. Council of Economic Advisers under President George W. Bush, wrote in the Financial Times that “Government-funded R&D is vitally important for innovation and productivity
growth. The case is clear.”
Moreover, a 2020 paper by Lawrence Summers, now at the Harvard Kennedy School of Government, and Benjamin F. Jones, a professor at Northwestern Universty’s Kellogg School of Management, found the broader social return on R&D investment dollars into universities and other public and private entities to be more than 13 to one.
“We’ve always been at the leading edge of research and education in computer science.”
— Martial Hebert
The Persistence of Influence
“We’ve always been at the leading edge of research and education in computer science, and when you talk about our influence on the industry, it’s very deep,” said Hebert.
“There’s a creative structure here, but we also remove barriers and make sure things can be done without obstacles and impediments,” Hebert said. “It’s a feature of our environment, of our culture, and it is why we produce people who will lead new industries.” ■
