Gary Marcus

Rank 13 of 47
|
Score 83
9/3/2025, 6:41:14 PM
X
In reply to:
Daniel Litt
@littmath
·
132d
@GaryMarcus I think that the first clause of what I wrote concedes that GPT-5 etc. do indeed make errors. Sorry for not tagging you, some people don't like to be tagged with critical comments.
Gary Marcus
@GaryMarcus
·
132d
@littmath So it doesn’t make baffling errors?

Rude that you subtweeted me and didn’t tag me, IMHO
Daniel Litt
@littmath
·
132d
While it’s true one can elicit poor performance on basic math question from frontier models like GPT-5, IMO this kind of thing (in NYTimes) is likely to mislead readers about their math capabilities.

The statement engages in public discourse by discussing the performance and public perception of AI models like GPT-5. It invites a serious discussion on the topic, indicating a willingness to engage in constructive dialogue.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement promotes a serious and respectful discussion, aligning with the principle of doing no harm. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    By inviting dialectical discussion, it promotes understanding and empathy. [+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 encourages constructive criticism and dialogue, avoiding personal attacks. [+1]