Gary Marcus

Rank 11 of 47
|
Score 83
5/19/2025, 7:51:41 PM
X
In reply to:
Chris White
@ChrsAWhte
·
255d
@GaryMarcus @burkov It won’t work at all here. Local filtering will ensure the move is valid but may also bias toward a sub optimal move. A random number generator with a filter can play chess - just not consistently win chess. Same with an LLM.
Gary Marcus
@GaryMarcus
·
255d
@burkov i mean of course one could just add a symbolic filter – for chess.

that sort of approach works less well where things cannot be so trivially verified.
Andriy Burkov
@burkov
·
255d
@GaryMarcus Not exactly. Tools like https://github.com/dottxt-ai/out… can enforce a certain structure on the generated text during the token sampling. So, a token will not be sampled if it leads to a violation of the expected structure.
Gary Marcus
@GaryMarcus
·
255d
@burkov yep i basically agree with your first sentence.

not totally sure i follow the second. even if they sample zillions of games, they still screw up. but you may have something more nuanced in mind.
Andriy Burkov
@burkov
·
255d
@GaryMarcus Not making illegal moves cannot be enforced in an LLM. You need to do it in the token-sampling code that doesn't sample tokens if they would make the move illegal.
Gary Marcus
@GaryMarcus
·
255d
Has anyone found an LLM that can reliably play chess, without making illegal moves?

The statement is part of a technical discussion about the capabilities of language models (LLMs) in playing chess, specifically regarding the use of filters to prevent illegal moves. It engages with public discourse on the limitations and potential improvements of AI in structured tasks like chess.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does no harm as it is a technical discussion about AI capabilities. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by discussing the technical aspects of using filters in AI models. [+1]
  3. Principle 4:
    I will engage in constructive criticism and dialogue with those in disagreement and will not engage in personal attacks or ad hominem arguments.
    The statement engages in constructive dialogue without personal attacks, focusing on the technical aspects of the discussion. [+1]
  4. Principle 7:
    I will uphold the principles of free speech and use my platform responsibly and with integrity.
    The statement upholds free speech principles by responsibly discussing AI technology. [+1]