Gary Marcus

Rank 13 of 47

Score 115

@davidmanheim no, i laid out my cards (what i think the gaps are) in 2001 and have stuck to them .

and yes i don’t think those gaps can be filled without a solid neurosymbolic system

read my essay on o3/grok 4 vindicating neurosymbolic approaches

8/31/2025, 1:39:03 PM

In reply to:

David Manheim

@davidmanheim

@GaryMarcus As in the past, I think you're playing a "failures of the gaps" game; every time there's an advance, you note that it's imperfect, but when there's a failure, or a limitation of a method, you say it's fundamental and can only be fixed with neuro-symbolic AI.

Gary Marcus

@GaryMarcus

Like I have been saying since 2001, embeddings aren’t enough.

Sooner or later. you are going to need true compositionality.

In the end neurosymbolic AI will win, and win big.

Rohan Paul

@rohanpaul_ai

BRILLIANT @GoogleDeepMind research.

Even the best embeddings cannot represent all possible query-document combinations, which means some answers are mathematically impossible to recover.

Reveals a sharp truth, embedding models can only capture so many pairings, and beyond that,

The statement is part of a technical discussion on the limitations of embedding models and the potential of neurosymbolic AI. It presents a viewpoint on the necessity of neurosymbolic systems to address gaps in current AI models, referencing past work and an essay for further reading.

Principle 1:
I will strive to do no harm with my words and actions.
The statement does not cause harm and contributes to a technical debate, aligning with the principle of doing no harm. [+1]
Principle 3:
I will use my words and actions to promote understanding, empathy, and compassion.
It promotes understanding by referencing an essay for deeper insights into neurosymbolic approaches, aligning with the principle of promoting understanding. [+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 criticism by addressing perceived gaps in current AI models without personal attacks, aligning with the principle of constructive dialogue. [+1]