Gary Marcus

Rank 14 of 47

Score 83

@peterwildeford yes, certainly possible. my bet it is that is both expensive and not reliable enough (in many cases) to justify the expense.

11/16/2025, 7:33:39 PM

In reply to:

Peter Wildeford🇺🇸🚀

@peterwildeford

39d

@GaryMarcus Also it may just be extremely expensive to run

Gary Marcus

@GaryMarcus

39d

People saying stuff like this “OpenAI's unreleased model that won gold on the IMO in July is clearly better than even the best publicly available model today (GPT-5 Pro)” are speculating and probably wrong.

my hunch is that the IMO model is limited either in its generality or

prinz

@deredleritt3r

39d

I've noticed that a lot of people tend to assume that capabilities of consumer-grade AI models (like GPT-5) are the same as the true frontier of AI capabilities. This is a mistake.

For example, OpenAI's unreleased model that won gold on the IMO in July is clearly better than

The statement is part of a broader discussion on the capabilities and limitations of AI models, specifically regarding their cost-effectiveness and reliability. It engages with public discourse on technological advancements and their implications.

Principle 1:
I will strive to do no harm with my words and actions.
The statement does not cause harm and is a neutral assessment of AI technology's cost and reliability. [+1]
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 by providing a reasoned opinion on the cost and reliability of AI models without resorting to personal attacks. [+1]