One-on-one with Jerry Brody

Q: So how did you get hooked up with this project?

Jerry Brody: Actually, I was in between projects -- back in '89, late '89 -- and they brought in Murray Campbell and F.H. (Feng-hsiung Hsu) to IBM. F.H. had worked on this machine, the Deep Thought machine, and they said "How would you like to work on this project?" And I said, "Why not? Sounds great." And we've been together ever since. It's a good team. We work well together.

Q: Did you ever think then that this is where you'd be today?

J.B.: No, because as an engineer, most of the stuff I work on, only pieces of it get into products or some of the stuff becomes mostly internal type tools. Nothing I work on ever really is an actual product, so it's very exciting. Never thought I'd be famous.

Q: Are you enjoying the fame? You're certainly getting a lot of notoriety for this.

J.B.: It's wonderful, it's absolutely wonderful. I never expected this kind of press and everything else that's going on.

Q: How many interviews have you done over the past couple months?

J.B.: Only a couple. I don't get into that too much, because the programmers are the ones that really have done most of the hard work on it.

Q: Can you describe your role to us more specifically?

J.B.: Support Engineer. We de-bug cards, design some of the stuff, some of the special stuff that we use. I started building another machine that we were going to use at one time, but it became better to use the SP so we put that design aside. The SP really makes it very strong.

Q: At the heart of this event, obviously, is a chess match between Kasparov and Deep Blue, but there's certainly a lot more at stake than just the match itself. What would you say is the deeper meaning going on behind this event?

J.B.: Well, it shows off the SP machine for what it can really do. And by using accelerator chips, there are all kinds of possibilities to help man-kind. The way I look at it, the SP can be used in the medical profession and in chemistry, or anything.

Q: So you look at it almost as a product test for Deep Blue?

J.B.: Absolutely. That's what it really is. It's a good test, a good scientific test.

Q: There really is only one person in the world who can test Deep Blue, and that's Garry Kasparov.

J.B.: He's the best. He's the best chess player anyhow, and I guess chess is a very good tool to use because it requires a lot of number crunching.

Q: What makes the Deep Blue machine so special?

J.B.: Deep Blue was designed to play chess. I don't know if it was really designed originally to play chess to show how powerful this could be, but I think that using accelerators in any machine does improve its performance. I've worked on some accelerators before, but nothing like this.

Q: Last year in Philadelphia, it probably was a pretty exciting event for the team. The event has received so much media attention. Personally, how was the event for you last year? Exciting? Nerve wracking?

J.B.: It was exciting. It was nerve-wracking when we lost a few games, but no machine is going to be able to beat him [Kasparov] every time. There's always that element that you can't look at, but even the best chess players lose games and even the best chess-playing computer can lose a game.

Q: We know that this year's model has been routinely beating Deep Blue from last year. How much better of a player is this year's model?

J.B.: I don't know. It's hard to tell until you actually play the big guy. It's better because the software's improved, there are a lot of improvements this year. Deep Blue is more prepared. There's more tuning going on, so it's a better machine…it plays better.

Q: Many people like to say that this is man vs. machine, but actually, it's man vs. machine that was created by other men.

J.B.: It's man vs. many, many, many people.

Q: What kinds of things did you learn from last year that you were able to build into Deep Blue this year?

J.B.: That's a software issue, and I'm not a software person, so…

Q: How did you feel before last year's game in terms of how prepared you were compared to how you feel this year? Did you feel confident last year before the game?

J.B.: Well, everything's an unknown, especially the first time we're out playing someone like Kasparov. There's a lot of apprehension, and when we won, of course we were very pleased. I feel more confident this time that we'll do better than we did last time.

Q: What is it about the new architecture and the new Power 2 Superchip that makes it so much more powerful than the chips used last year?

J.B.: The machine runs twice as fast, so it can crunch the numbers more quickly. Usually, according to the chess curve, the faster the machine, the better the play. Of course, the software's important too, you can search deeper and make better moves.