This confuses Meow; why is this a surprise? A LLM is not a game player, it's a prediction engine trained on (illegally sourced) text. Why would it be any good at chess to begin with? Why does anyone think an LLM would be good at anything?
That was my first thought. Even if it had been trained only on chess games, there are surely far more bad plays than good. Since LLMs are about what probably comes next, that would mean most move predictions would be bad moves.
I love that the AI is giving off
tech-bro-who-paid-someone-else-to-level-up-his-videogame-account-and-gets-exposed-when-he-has-any-one-actually-watching-him-play-before-he-just-rage-quits-and-does-ketamine-instead energy with all of its excuses.
Comments
That very old very limited Atari system on the contrary has algorithms specifically designed for chess.
tech-bro-who-paid-someone-else-to-level-up-his-videogame-account-and-gets-exposed-when-he-has-any-one-actually-watching-him-play-before-he-just-rage-quits-and-does-ketamine-instead energy with all of its excuses.
The world isn't ready for a middle-aged, fist bumping hippie screeching
A TAR I