But in order to understand that arc, you first have to understand who he was as a kid.
“I was a pretty good student,” Lynch says today. “I got mostly As. Except in deportment and conduct.”
Lynch’s statement underplays his substantial intellect (he once dreamed how to calculate the interior space of a Chinese lantern–and got the math right). And his conduct problems don’t reflect tension with peers — he just didn’t think authorities were always right. He liked to do things his way, or at least try. It’s a theme you often hear from technology leaders and luminaries, and one that likely helped secure Lynch a position in the 2019 Internet Hall of Fame inductee class for playing a key role in driving global adoption of TCP/IP protocols and fueling the Internet’s commercial success.
As you get to know Dan Lynch, what emerges is the rare and right individual who could grok complexity, grasp possibility, and navigate the human personalities necessary to build a dream. The more you learn about Dan Lynch before he was a leader in the development of the Internet, the better you can understand how the Internet developed as it did.
Lynch grew up in Southern California. His father had seriously considered the Jesuit priesthood, but instead became the enforcer of Hollywood’s Hays Code which placed so-called moral limits on what the movies could depict. His mother had a masters in chemistry. As he notes, Lynch was a successful student, even when he skipped three months of tenth grade to tend to a dying uncle whose brilliant military career he admired.
Lynch soon began a military career of his own, starting in the Air Force ROTC while attending Loyola. His mathematics grad program at UCLA taught the “pretty good student” something new: He wasn’t the smartest guy in the room. The best guidance he’s received throughout his life, Lynch says, has been “about tamping my ego down.” In the presence of his fellow students, he learned to appreciate those bigger intellects and figure out how to get them to work together.
Masters degree in hand, Lynch was assigned to Holloman Air Force base in New Mexico. The Cold War’s demand for both missile defense and global communications were structuring components of his future work, giving him experience in the real-time computing skills that presaged the Internet’s needs. But Lynch didn’t want to make a career of the military. “We had figured out that the offense has it all over the defense. I didn’t want to spend the rest of my life playing cat and mouse. By then, I had become a computer guy and I wanted to work for a computer manufacturer.” However, his first interviews failed. Instead, he got a job with Columbia University on a highly classified radar program that kept him at a secret location in Southern New Mexico. Perhaps that was best because that’s where he fell in love with programming.
“I remember when I threw up my first program, writing a calibration command for an optical instrument. I noticed this kid across the room and I could tell he was having fun. I asked him what he was doing and he said, ‘Debugging. Here, figure out how to program.’ And he handed me a little tiny booklet, which I have on the wall now. I wrote a program and when it ran the very first time I thought, ‘Holy cow, that was pure damn luck.’”
With friends beginning to build major computer businesses, Lynch knew he was going to move to either the San Francisco Bay Area or Boston, at the time the two main centers of computing. He took a job at Lockheed Martin in Silicon Valley alongside a friend from the Air Force where the Lynch ‘independent streak’ put him in conflict with the slow, bureaucratic gears of big business. He knew he needed to find another perch.
“I got a phone call from a guy I met in New Mexico from Stanford’s Artificial Intelligence Center.” The Stanford Research Institute (SRI), the laboratory that had created the very first autonomous robot, was looking for someone to run their computer center. The team was made up of about 50 people in Palo Alto and five other locations around the globe, including London and at Carnegie Mellon in Pittsburgh, and the University of Utah in Salt Lake City. This fit felt right. In the present day, Lynch quips that he told his interviewers, “‘You go ahead and interview anybody you want, but I’m your guy’–I didn’t quite say it like that, but I knew and they knew.”
The SRI computer center supported language research, robotics, and vision applications. “I knew nothing about AI,” admits Lynch. “You know what AI is? It’s what programmers with an IQ of over 180 do.”
Turns out, he fell in love with one of these geniuses and married her. Falling in love with brilliant complexity is a consistent Lynch theme. On the first day at SRI, a colleague showed Lynch the computer he’d be working on. “He logs into the computer and from there he switches over to one in Boston. And then he goes to one in London. Then to Utah, then back to one down the hall from us! I fell in love with the ARPANET, hour one.”
A few years later, Lynch led the operation to unite computers like those and thousands of others from the early-ARPANET transfer layer protocol NCP to the more flexible and powerful TCP/IP, which he helped develop for Bolt Beranek and Newman. This work came to a climax on January 1, 1983, in what is now known as Flag Day.
He was at the University of California Information Sciences Institute (ISI) in Los Angeles by then, working to bring together global communications, a big mandate with no budget. “I was like Tom Sawyer,” he recalls of the interpersonal skills necessary to manage the mass coordination of the transition to TCP/IP. “I got people to do things by inveigling them or embarrassing them — whatever it takes.”
The moment of switchover was midnight on New Year’s Eve, as Lynch puts it, “… when everyone is drunk.” Despite the party connotation, by dint of Lynch’s coordination and the global commitment, the change happened seamlessly, an uneventful technical birth at a moment of champagne and frolic. Flag Day now marks the start of the modern Internet.
With that project complete, Lynch, like other insiders, realized that the world of computing was changing. Having run a computer center for ten years and launching a historic project, he was itching to be a part of other big work. Friends were starting landmark companies like Cisco and Sun MicroSystems, but Lynch’s earlier attempts to join the computing big leagues weren’t commercial successes, and he’s frank about that. After a few failed starts, he had an idea that took him back to the ARPANET team.
Lynch told his former colleagues, “You’ve made it so the world can have this code but you haven’t told them how to use it. Let me put together a workshop and invite all the implementation teams from around the country. I’ll take the ten guys from your research world, who invented all this stuff, and have them explain what works and what doesn’t.” The first INTEROP conference was held in Monterey in August 1986 with Vint Cerf as the keynote speaker. Lynch charged participants a hundred bucks, “… to pay for donuts.” Audience response to the conference? We want more of this.
The conference grew by 100 percent year after year, training thousands on the operation of Internet-enabled equipment. When Lynch was ready to move on, there was a buyer eager to take over. The $25 million sale price gave Lynch an opportunity to pursue some business-building dreams.
“One of my children told me in the early 90s, ‘Dad, this Internet stuff is great but I want to be able to sell cartoons for a nickel by the million.’ I knew the Internet would fail if we couldn’t move money on it,” says Lynch. He brought together Bill Melton, who had founded Verifone, Steve Crocker, and Bruce Wilson to launch CyberCash, taking the company public in 1996. But the ego-tamping Lynch reflects, “We had great technology, but lousy marketing.” CyberCash was liquidated (Lynch’s deportment in its demise should earn him A’s–he and his partners made sure to shut the company down without damaging others’ reputations or careers), with some tech assets ending up on the backend of PayPal, where the CyberCash name can still be seen in lines of code.
That ghostly flash of code is not unlike the impression Dan Lynch has left on the commercial Internet. Whether it’s the free flow of data that we all take for granted and which is facilitated by his work on TCP/IP or the thousands of individual technologists and other facilitators who were informed and educated at Interop conferences, Lynch’s rare and right presence is always there.