05 Brewster Kahle talks about people at Thinking Machines


Let’s talk about some of the people at Thinking Machines. So Danny Hillis was, in my world, he was my mentor. For the Connection Machine Model 1, every transistor in that thing was there because of Danny. I played a role of helping make it come true. I was one step closer to reality from Danny. I was a certain level of filter that took some of these ideas– some of them are good, some of them are bad– and I would sort of filter some of them down, and try to make them–the next level work. And there are people that would work all the way through. And I don’t think of It as down, and he didn’t think of It as down. It’s over. In the sense that there was ideas, there’s architecture, there’s chip design, there’s board design; there was packaging, and how do you go and make this? How do you communicate it? How do you do the software for this thing, which is really unknown! He had his hand in all of these things, but there were people that were just unbelievable. How do you go and make a company out of a bunch of bonzos from MIT, and make that come about, and communicate such that you… If you’re gonna sell a machine for five million dollars–
–I think it was only three. –For three million dollars.
–Cheap! –You have to go and get approval through a lot of people that had no idea what you’re talking about. So who are some of the people involved? There’s Danny and Sheryl, they were the founders. There’s Howard Resnikoff, who was one of the early shakers. I loved his line of: “To err is *human.*” And he’d put little stars around the “human.” So it’s a snarky sort of, “yeah, maybe we’re gonna get over this.” –What was Howard’s role in the company? –In the founding of the company, he worked on trying to build up the community and outreach to smart people and bring them together. Sheryl had this great idea of going and having it be in this mansion in the middle of Waltham [Paine Mansion]. –Which was wonderful!
–Oh it was wonderful! It was such a great setting to go and think great thoughts. And she sort of talked them into allowing us to use this old historic mansion as a place. And so the idea of bringing together Richard Feynman, Carl Feynman’s father, Stephen Wolfram came for a time, and they were just all of these different points of view. A lot of them were woven together by Howard Resnikoff, and then he left. Your coming just brought a level of artistic and thoughtful design, not only to the machine, but to the people and the organization
building the machine. –What does that mean?
–Well okay, that you went and spent time going and making the design, the T-Shirt, that actually is kind of iconic out of what Thinking Machines became, as part of–
–[gesturing to his shirt] The T-Shirt. –Yes, the T-Shirt, right? It’s on Richard Feynman, on billboards, on the Apple ads. That was a thoughtful thing towards having us weave together. Because we weren’t going to come up with this as just the brainchild of Danny Hillis, and it’s springing fully-formed. It had a large number of people pouring their all into the different aspects to it. Carl Feynman was involved in the project
for as long as I was…Cliff Lasser, these were the software guys… –What was Carl Feynman’s role?
–Carl Feynman Is one of those amazingly brilliant people, that if you give them just the right problem, they will come out with an answer that you’ll have never thought of. He’s not the person to build the bridge, he’s the person to go and say, “We’ve got a real problem and it looks like it’s gonna fall down because of this type of joint,” and he’ll rethink it. So: an example: we were stuck. The Connection Machine… We were designing this computer chip. I think it was the Connection Machine Model 2. It might have been the Model 1 chip. We needed a router to go and direct all the messages around, so the equivalent of the Internet, within the computer. –There was a router in the 1.
–There’s a router in the 1 and we had… The basic cell for making the decisions was too many gates. And it was too big. And I went off to California to work with the– and we tried to cram it down and we couldn’t get it. It was like the whole project was going to stop. So we gave it to Carl Feynman and he took it with his father. Danny and I worked on it, I got It down to here; then Danny, he’s brilliant, got It down to 16 gates, something like that. And we thought we were done, that was all I could do. But it still wasn’t enough. And Carl Feynman and Richard Feynman went off and… he sent a postcard. The design came back on a postcard! And it was down at 11 and a half gates. –Oh my god, you don’t have that postcard, do you?
–No, hopefully it’s part of the Internet Archive’s… excuse me, the Thinking Machines archives.
–Is there a Thinking Machines archive? –I don’t know, I hope so.
But this postcard came back and it was hand drawn. And if you look at it, it requires deeply understanding how chips are laid out at a very low-level, to build it back up into what it is that was done. So he saved the Connection Machine. So he’s a brilliant man. There’s the guys that tried to make the software work– Guy Steele. Cliff Lasser– and made it so that other people could
figure out how to program this thing. Jim Salem. It was a cluster of
some of the smartest people. And Thinking Machines, I think by having a high dream, was able to bring people together, to work on, not a slavish thing– we weren’t working for the next quarter’s profits– it was, “Let’s build something great. Let’s set the goal high. Let’s make something beautiful.” Who works on a computer to make them beautiful? Nah, you make it so they can ship!

Leave a Reply

Your email address will not be published. Required fields are marked *