How DARPA Invented The Internet

How DARPA Invented The Internet

DARPA is an agency you’ve never heard of that funds programs you’ve only ever dreamed of. In this week’s edited excerpt, Michael Belfiore uncovers the history of this enigmatic government entity and reveals the incredible technologies it develops.

A former NASA program manager named Bob Taylor rose through the ranks of IPTO program managers to become the office’s third director in 1966. He later described the time-sharing setup at his disposal. “In my office in the Pentagon, I had one terminal that connected to a time-sharing system at UC Berkeley. I had one that connected to a time-sharing system at the System Development Corporation in Santa Monica. There was another terminal that connected to the Rand Corporation.” Taylor could interact long distance with any one of those computers through the remote terminals in his office.

He himself saw the limitations of interacting at a distance with a computer, however, and the year he took over IPTO, he got the agency to spring for a brand new SDS 940, the first computer designed expressly for time sharing, for Engelbart’s group at SRI. Even before then, Engelbart and his team had been at work building monitors (most computer output in those days came through Teletypes and punch cards), input devices and software to allow users to interact with computers as never before. But the new computer, along with ongoing support from ARPA, NASA and the Air Force Office of Scientific Research, allowed them to rocket to a whole new level, culminating in a system they called simply the oN-Line System, or NLS. It was the precursor of virtually every aspect of what we know today as the personal computer.

By December 1968, Engelbart and his team at SRI were ready to reveal the fruits of their labour to the rest of the world for the first time. The occasion was the Fall Joint Computer Conference at the San Francisco Civic Auditorium, and the place was packed with, by one account, 2000 to 3000 computer researchers. The event would go down in computer lore as the Mother of All Demos, and it would forever change the direction of computer hardware and software development – though many of the concepts Engelbart introduced would take decades to come into their own.

In a staid, matter-of-fact delivery that made it all seem so inevitable, Engelbart proceeded to demonstrate for the first time, from his seat at a workstation, whose display was projected overhead for all to see (itself no easy feat – he’d had to borrow the only projector on the West Coast capable of projecting a video image from NASA), word processing, video chat (with colleagues at their own workstations back at SRI 60km down the coast), real-time collaborative document editing, hypertext links that connected key words to relevant supplemental material, and, among all the rest, the computer mouse. “I don’t know why we call it a mouse,” Engelbart monotoned. “Sometimes I apologise. It started that way, and we never did change it.” To say that the demo was a tour de force was an understatement. In a world of punch cards, Teletypes and batch processing, and in which computers were anything but personal, the debut of the concepts behind today’s personal computers blew the minds of Engelbart’s audience. At the conclusion of the demo, they erupted to their feet in a standing ovation. “I looked up,” Engelbart later recalled with wonder, “and everyone was standing, cheering like crazy.” One conference attendee later described the demo as hands down the most astounding event he’d ever witnessed. Another topped that by saying that it not only altered his concept for what computers could do for people, but also changed his life.

Surely missed by most in the audience at the demonstration amid all the astonishing innovations was a brief mention of a new project, “this ARPA computer network-experimental network that’s going to come into being in its first form in about a year and end up sometime later with some 20 experimental computers in a network.” Little did Engelbart’s audiences know that that experimental computer network would turn out to be just as world changing as anything they had just witnessed.

ADVANCES IN POWER generation may well join past DARPA projects such as the ARPANET in forever changing the way we live and work. But we must not take these and other marvels for granted. They have to be conceived and developed by dedicated managers, scientists, engineers and other visionaries. Visionaries such as those working for the Defence Advanced Research Projects Agency. One program manager described DARPA to me as a national treasure, and I tend to agree. However, I believe it will stay that way only if, along with giving the agency its due, we also practice the benign neglect that it requires to thrive.

DARPA was conceived and has operated, first and foremost, as an agent of change for the United States armed services, and I believe it should stay that way. Its mission to equip our nation’s war fighters with a technological edge over their adversaries gives DARPA a razor-sharp focus it would otherwise lack, and its emphasis on quickly moving projects from concept to working prototype while making as efficient use of funds as possible should be a model for research and development throughout the Department of defence.

Even more than that, DARPA should also be a model for research and development throughout the federal government. It has proven again and again that true innovation need not depend on massive expenditures and armies of bureaucrats. Operating on one half of 1 per cent of the US defence budget, with a staff housed in a single office building of modest size, DARPA has fostered some of the most useful technological innovations of all time. Key to its success – in addition to its minimal bureaucracy – has been the term limits for its managers and its low overhead. The term limits ensure that the people who do its most important work care more about fulfilling the agency’s mission than protecting their jobs. The decision made at the outset that the agency shouldn’t maintain its own laboratories but instead farm out work to other organisations has allowed it to consistently stay ahead of the technological curve by quickly developing new capabilities and letting others go as the need arises. Managers of any new government agency seeking to capitalise on DARPA’s recipe for success – for instance, the Advanced Research Projects Agency – Energy, or ARPA–E, funded under the Obama administration in 2009 – would do well to remember these essential ingredients.

Here’s to the next 50 years.

Michael Belfiore has been a full-time writer since 1995. He has covered the launch of the first privately built spaceship for the New York Post and Reuters as a freelance journalist and has written about space flight and advanced technology for Popular Science, New Scientist,, Air & Space, Financial Times and others. In addition to his work as a journalist, Michael provides marketing and public relations writing services with his wife, Wendy Kagan. They live in Woodstock, New York, with their daughters Amelie and Jade.

The Department of Mad Scientists: How DARPA Is Remaking Our World, from the Internet to Artificial Limbs is available from Amazon.