Gary Marcus

Rank 12 of 47
|
Score 83
5/20/2025, 12:35:06 AM
X
In reply to:
Daniel Litt
@littmath
·
255d
@colin_fraser @JaEsf @GaryMarcus @typingloudly Indeed it probably could. Very silly to ask it to a model without tool access though.
Colin Fraser
@colin_fraser
·
255d
@JaEsf @littmath @GaryMarcus @typingloudly I think all of you are overestimating how hard it is to generate large prime numbers. I would have predicted that gpt-3.5 with python would be able to do this.
Jacob Eliosoff
@JaEsf
·
255d
@littmath @GaryMarcus @typingloudly And 2025 isn't really that interesting anyway, the point is the trajectory, which seems blazingly obvious. There is no chance that within a year or two the models won't unambiguously ace this question, I hope no one's debating that.
Jacob Eliosoff
@JaEsf
·
255d
@littmath @GaryMarcus @typingloudly So sure the transcript illustrates the point that the models remain kind of clunky and untrustworthy. But it also illustrates that the 2025 answer to "Can the current best public models generate large rare primes" is "Yes".
Daniel Litt
@littmath
·
255d
@GaryMarcus @typingloudly Yep, it’s weird! But it also did give me 5 prime numbers that have almost certainly never written down before.
Gary Marcus
@GaryMarcus
·
255d
@littmath @typingloudly wild that it gave you completely different numbers when you said give them to me with no spaces.

The statement is part of a technical discussion about the capabilities of AI models in generating large prime numbers, focusing on the abstract learning process independent of tools. It engages with public discourse on AI technology and its development.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does no harm and is neutral in tone, focusing on a technical aspect of AI. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by discussing the abstract learning capabilities of 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, contributing to a technical discussion without personal attacks. [+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 capabilities. [+1]