Stork: What did you think about it?
Hoane: I absolutely did not understand the obelisk. I think I was too young to understand the social context. Everything seemed natural and reasonable except the statements being made by the obelisk.

Stork: Did the technology seem reasonable?
Hoane: I was a kid who read science fiction. Everything seems reasonable when you're sixteen and you read science fiction all the time. You want to believe everything

Stork: So tell us about the part you play in the Deep Blue project.
Hoane: I'm a software engineer. I'm responsible for the whole system working. I wrote the run time software. The part of Deep Blue that's a chess program

Stork: Can you quantify how big the full system is from a software perspective, in terms of lines of code etc.?
Hoane: Half a million lines of code. Hardware is capable of probably a trillion operations per second if you had to do it in a software equivalent.

Stork: In the1997 match you sat across from Kasparov, what was it like?
Hoane: It was scary. This is a man who has never lost a chess match and he was losing this one. He looked at me and I only had to nod defeat. (If I didn't imagine it). But I think at some point he realized the sort of thing he used against other computers wasn't working. You know it was just over. And so then it's like well what's going to happen. I mean he's going to be really angry, really upset you know, I'm sure he found it really hard to accept.

Stork: What about after the game, what was it like being there with his reaction?
Hoane: Well he was upset. He had the form to shake my hand. There was nothing personal there. He was all involved in his own...misery, defeat. It was very unpleasant for him and that I was acutely aware of.

Stork: Can you recall the last moments of that match?
Hoane: The scene in the last game was that we believe that Kasparov had tried an open line that had previously made his practice computer opponents self-destruct, meaning they got foiled because they didn't understand the king safety issue. They wouldn't know how to prosecute the attack properly and they would do the wrong thing and basically self-destruct. I believe Deep Blue did not make the mistake that he was inviting. And so from that point on the game was over. We could all see it coming for some time before he actually resigned.

Stork: How did you feel after the loss of the first match?
Hoane: I felt like we had built an engine that could win a match against the world champion. Having had Joel Benjamin in for almost a year, practicing and finding the weaknesses and correcting them, we came to the first match with this faith that we could win. Faith because chess relies on a lot of luck. We knew we weren't going to get blown out of the water. We knew we had a credible opponent for the world champion. And you know I would state that Deep Blue is probably not better than the world champion. It's just that he's close to Deep Blue. Our job's done; we have created a machine that can play at the world champion's level. Of course with more years and more effort we could make a better machine that would play better than the world champion and dominate him but we had faith going into the match that we had created such a machine.

Stork: Is the problem of computer chess solved?
Hoane: Chess hasn't been solved by a computer. We didn't create a machine that learned chess. We created a machine that could operationally play chess based on the concepts we put in. The humans involved understand chess concepts and managed to articulate them to the machine. And the machine operated on those concepts so Éit's a mechanistic sort of way. A bigger challenge would be to design a machine, give it the rules of chess, talk to Gerry Tesauro, talk to the gentleman who programmed the backgammon program. What he did was start with the rules of backgammon and built a machine that played world class backgammon almost without touching it. You know, to do that in chess is seemingly inconceivable at this point, so many specific rules and ideas that it would have to learn. Maybe if you had some device like Hal that was a complex as a human and you trained it for twenty years that would be a great accomplishment sure.

Stork: What went on behind the scenes at the Kasparov match, technically?
Hoane: Technically, the thing we identified as one of Mr. Kasparov's strengths is adaptability. You know the first match he lost the first game and it was crushing. He reconfigured himself and used other strengths that he has a chess player, that the machine couldn't adapt to. So we went in with the strategy that we could be adaptive ourselves. Between games we could make some changes and we did with the idea of not letting him get our number like he did the first time. Make little changes and as the match progressed, we fixed various things.

There was a particular move in game 5. The grandmaster, Joel Benjamin, was there and he said, "O my god that is a terrible move." And he was considering it and thinking maybe it wasn't so bad. Then Miguel Illescas came up and said it's a strong move and that out of the top 100 people in the world only 2 would make this move, and then only after a few drinks. And it was maybe one of the strongest moves of the whole match.

Stork: What about the raw mechanics of making the system fly?
Hoane: It was a huge effort. Essentially the Deep Blue chip that played that match was the sixth or seventh revision of the hardware over the history of the machine.