Gary Marcus

Rank 14 of 47
|
Score 83
9/3/2025, 6:39:37 PM
X
In reply to:
Matt Schwartz
@matt_is_nice
·
135d
@GaryMarcus @ShakeelHashim Cmon Gary. I read NYT and thread, hence initial question. Best I can tell you claim in NYT that “GPT5 shows decisively that scaling has lost steam” without really making the argument!

Hard to judge impact on scaling laws if you don’t know how big GPT5 actually is…
Gary Marcus
@GaryMarcus
·
135d
@matt_is_nice @ShakeelHashim fine, OAI gave up on being AI lab because scaling didn’t get them to AGI. Happy now?

also read this thread and my article in NYT and Substack re xAI and what they found w grok 4. don’t just continue to be a reply guy who doesn’t even read the actual arguments.
Matt Schwartz
@matt_is_nice
·
135d
@GaryMarcus @ShakeelHashim Not necessarily. Alt explanation: OpenAI is now primarily a consumer app company vs AI lab. This leaves them resource constrained and needing to focus on serving inference at scale

Tracking Google and xAI prob more informative for scaling and neither seem to be slowing down
Gary Marcus
@GaryMarcus
·
135d
@matt_is_nice @ShakeelHashim isn’t it worse for scaling if they decided best on internal testing to make 5 smaller than 4.5?
Matt Schwartz
@matt_is_nice
·
135d
@GaryMarcus @ShakeelHashim How do you know it’s at least as big as 4.5? Haven’t been able to find param counts to compare
Gary Marcus
@GaryMarcus
·
135d
@ShakeelHashim this is intellectually dishonest,

at BEST it at least as big as 4.5 which was intended as 5 which was significantly larger than 4

it is surely scaled up compared to 4 which is what i compared it to

The conversation involves a debate about the scaling of AI models, specifically GPT-5, and whether scaling has lost its effectiveness. The tone is contentious, with accusations of intellectual dishonesty and frustration over repeated questioning.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement shows frustration and dismissiveness, which could be seen as harmful to constructive dialogue. [-1]
  2. Principle 2:
    I will respect the privacy and dignity of others and will not engage in cyberbullying, harassment, or hate speech.
    The statement does not respect the dignity of others, as it dismisses the other person's comments as a 'shtick.' [-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 does not engage in constructive criticism and instead resorts to a personal attack by labeling the other person's behavior negatively. [-1]