The Rematch the Computer Won

1. The earlier match

In 1996, IBM’s Deep Blue won a game against Garry Kasparov but lost the match. The world champion had demonstrated that one game was not enough. The match put a specialized computer against a world champion. It tested a particular chess system rather than every form of intelligence.

The computer’s next move began in the laboratory.

2. A clear rule for a difficult contest

Chess provided a defined board, legal moves and a result that spectators could recognize. Inside those clear rules lay an enormous range of choices. The machine's achievement depended on handling that particular task effectively, which made the match an important test without turning it into an examination of every kind of thought.

The rules were simple to state and difficult to master.

3. The upgrade

Engineers improved the system and consulted chess specialists. Deep Blue evaluated enormous numbers of positions using specialised computation. Searching possible moves required both computing power and chess evaluation. The machine did not need to think in exactly the way a person did.

Its preparation involved processors as well as opening theory.

4. A new meeting

Kasparov and Deep Blue met again in New York in May 1997. The six-game rematch drew intense attention beyond the chess world. Each game affected the position of the overall match. A single impressive move was not the same as winning the whole contest.

Every move was carrying an extra argument about machines.

5. The sixth game

After wins and draws, the final game decided the match. Deep Blue won, finishing with a 3½–2½ score. The result depended on a sequence of actual games. The human player could not undo a mistake by beginning the match again.

The rematch had reversed the earlier result.

6. A game belongs inside a match

A striking move matters because it changes the actual position and the decisions available afterward. The contest could not be judged from one impressive-looking moment alone. Its outcome developed through successive games, where mistakes and recoveries affected a result neither participant could settle by simply replaying a favorable demonstration.

The next position could still argue back.

7. The boundary

The result demonstrated a specialised system’s extraordinary chess capability. It did not establish that the computer possessed human understanding across every kind of task. The achievement was significant within a tightly defined task. It does not establish that the system understood human life outside the chessboard.

The machine had won the match, not every philosophical debate.

Sources

• IBM — Deep Blue history