Garry Kasparov vs. Deep Blue: The Rise of Machines!

Hello everyone, Today I'm going to tell you about an event that led to a major technological advancement in chess history. In 1997, GM Garry Kasparov was preparing to play a big rematch against a bot called Deep Blue, developed by IBM to defeat him, which could calculate 200 million positions per second. Everything was ready for the match, and it began. It was the first match, and Kasparov thought, “Computers can't understand strategy. They can't beat me for long.” This mindset led to his defeat, and he conceded on the 19th move. These matches weren't just a single game; overall, Kasparov lost 3.5 to 2.5. Deep Blue's victory over Kasparov marked a significant advancement in technology history. The Deep Blue bot is also known as the ancestor of today's bots. 

If you'd like, let's analyze one game played against you.

This match was one that Kasparov won, and the sole reason for his victory was that he played the Caro-Kahn Opening. Deep Blue, expecting a different opening from his opponent, made mistakes when faced with the Caro-Kahn because it was an opening it did not know, and resigned on the 19th move. 

After this match, Kasparov lost the remaining matches due to his opening mistakes. Deep Blue won that tournament, but how did it win? Could a bot win a match with the capabilities available at that time? You may have questions like these in mind, so let's look for some theory to answer them.

What I'm About to Tell You Garry Kasparov – “Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins” (2017)
→ Kasparov's own account of all the claims, behind the scenes of the match. Taken from his book.

In his post-match comments, Kasparov stated that he made the h5 move in a game he played against Deep Blue, describing it as highly creative and something only a human could do, interpreting it as a human move. IBM, however, claims that this is not the case and that it was entirely the bot playing. Both sides may be correct, but here we will only speculate and theorize. This move could have been caused by a bug in the AI. Years later, the engineers who created Deep Blue said, “Deep Blue sometimes chose a safe move at random in difficult positions.” This could be true because the h5 move played could have been a safe move chosen in a difficult position, but it could be interpreted as interesting because it was a move that won the game. It is possible that the bot made it by chance.

As additional information: Matches were held in 1996 and 1997, one with human assistance and the other by developing the bot's software and hardware to defeat the bot. These two matches are considered bot-human defeats.

What are your thoughts on this topic? Don't forget to share them in the comments. See you soon.