I recently imported some of the Deep Blue v Kasparov games to have a look at and saw that DB did not make any inaccuracies/mistakes/blunders - like the AIs of today. What has changed in the last 17 or so years between then and now? Are AIs better than DB, and if so, in what ways?
Software-wise modern AIs, as you call them, are an order of magnitude better: better quiescence search, smarter extensions, better evaluation, etc. But DB would still be a very strong program even today thanks to its extremely fast hardware. However, the latest version of Stockfish on the fastest commercially available PC would wipe the floor with DB, believe it or not.
i heard DB wasnt actually a software but was in fact a computer whose sole purpose was to play chess, i.e DB was a piece of hardware
dunno if this is true or not though
dunno if this is true or not though
Yes it is... but that is what I mean - is there much of a difference between DB performance and engine performance, and what are those differences?
Today's engines are far better at eliminating useless moves, they becoming a little smarter and don't need as much processing power while DB could calculate 200 million positions a second a top program today could probably have less than 1% of that processing power and still beat Deep Blue.
So DB's "brute force" approach was not as good as modern "intelligent" software (though to be fair, I think an engine from back then, if there was such a thing, would be rubbish compared to modern ones)
I don't know how advanced DB's AI was in comparison to chess software of that time. But yeah, DB's AI is primitive compared to today's engines.
This topic has been archived and can no longer be replied to.