Around the building people joked that with their robot arm they were the Father, the Son, and the Holy Ghost.
Several floors below were the engineers building Multics—people like Tom Van Vleck, Dave Clark, John Donovan, and Fernando Corbató, whose earlier work helped invent modern time-sharing.
The whole environment was guided by ARPA’s Dr. J. C. R. Licklider. Around the lab he was simply called “Lick.”
On a couple of occasions the young and somewhat irreverent hacker that I was got called into his office after wandering off to explore some idea that seemed more interesting than whatever assignment I had been given.
I expected a reprimand.
Instead, he leaned back, listened for a moment, and said something I’ve never forgotten.
“Good work, son. That’s exactly what I would have done.”
It was an extraordinary place to learn.
Some of us who stayed all night eventually wandered out at dawn for dinner at the Fox and Tishmann Diner in Kendall Square. The wonderful thing about the F&T was that it didn’t matter what time you arrived. If a group of hungry programmers and mathematicians showed up at seven in the morning wanting meat loaf, roast turkey, or fried fish, that’s what they served.
We sat shoulder to shoulder with Nobel Prize-winning scientists who had also been up all night chasing ideas.
At the time, none of us thought we were building anything grand.
We were just experimenting with computers.
A paging trick here. A language feature there. A strange little program that played with patterns in text. Each of us worked on some tiny corner of the machine.
We saw the problems up close, in the small. The machine in front of us. The code we were trying to make behave.
Most of the time it felt like tinkering.
But over the years, those small ideas accumulated. One improvement made another possible. One experiment suggested the next. Slowly, the machines grew more capable, and the things people tried to do with them grew more ambitious.
None of that felt inevitable at the time.
It just felt like curiosity.
Only much later did I begin to realize that some of those small experiments—especially the ones involving patterns in text—were pointing toward something much larger.
In the next article, I’ll come back to those early pattern tricks and show how they eventually evolved into the systems people today call artificial intelligence.