It may be the buzzword of the current tech revolution, but AI has been part of our lives for quite some time—especially if you're a fan of video games.
AI has powered pretty much any single-player experience since the dawn of gaming, and in the realm of chess programs, it has been challenging human players for decades.
Recent developments in AI have understandably focused on chess, given the game's rich history and prestige. For example, the European Chess Union signed a strategic partnership over the weekend with smart home robotic brand SenseRobot in a collaboration which "represents a significant step toward modernizing chess education and development."
One person who isn't impressed by all of this is original Xbox creator Seamus Blackley, who took to social media to pour cold water over the power of modern chess programs:
No, I’m not impressed when your fancy chess AI crushes me. I’ve been being beat by shit chess code since the early 80s. The ZX80 could beat me for half a ten billion trillionth the resources and energy.
Blackley certainly has a point—I'm sure we can all vividly remember having our arses handed to us by games like Battle Chess or Chessmaster, both of which were running on platforms which are significantly less powerful than the smartphones we carry around in our pockets today.
The debate about AI's suitability as a chess opponent hasn't been fully settled yet, either; according to a research team at Palisade Research, several leading AI models resort to cheating when faced with a superior opponent.
Nefarious tactics include the AI model running its own copy of its opponent (the open-source chess engine Stockfish) to learn its moves and even overwriting the chessboard to gain the advantage.
Chessmaster on the Game Boy never did that.
[source bsky.app]
Comments 4
Based Blackley.
The thing about "AI" chess models is that they are already obsolete. The IBM supercomputer Deep Blue beat a grandmaster in 1997. Modern dedicated chess engines can calculate tens of thousands of possible moves in seconds and are effectively unbeatable by humans.
@JJtheTexan It's also worth knowing that "Deep Blue" isn't considered a particularly good Chess AI by today's standards. I'm sure I remember testing out one called Fritz or something many years ago that's claim to fame was it could beat Deep Blue while running on commodity hardware.
@Kushan I was a teenager when Deep Blue beat Gary Kasparov and I remember it being huge international news. It's astonishing that almost 30 years later we have more powerful chess engines running on hardware that costs a few hundred dollars
Leave A Comment
Hold on there, you need to login to post a comment...