Gary Marcus

Rank 11 of 47
|
Score 106
6/8/2025, 1:46:20 PM
X
In reply to:
𝞍 Shin Megami Boson 𝞍
@shinboson
·
41d
@GaryMarcus @rao2z Also I flat out don’t believe that a typical “bright teenager” can write out the correct sequence of 255 moves for the 8 disk tower of Hanoi problem from their head without transcription errors. I don’t believe the paper even claims this fwiw.
𝞍 Shin Megami Boson 𝞍
@shinboson
·
41d
@GaryMarcus @rao2z Gary I did implement it. See the link. I implemented it 48 minutes after you said it couldn’t be done.
Gary Marcus
@GaryMarcus
·
41d
@shinboson It’s not a trivial test; it’s failure modes are not in fact same as people (e.g. a bright teenager can easily learn to do 8 discs), and if you can trviially implement with a scratchpad, you should do so and demonstrate it. you should also read the related work by @rao2z that i
𝞍 Shin Megami Boson 𝞍
@shinboson
·
41d
@GaryMarcus 1) what this paper evaluates is very specifically whether LLMs are able to perfectly *execute* a formal algorithm without any sort of scratchpad or tooling like symbolic reasoning

2) this paper determines that LLMs are bad at this, which is a trivial result and also so are we
Gary Marcus
@GaryMarcus
·
42d
See you later, LLM’s?

Weekend long read on how the new Apple paper spells trouble, though not quite the end:

[The link you need to type is in bottom left of the image, because Elon]

The statement is part of a technical discussion about the capabilities of large language models (LLMs) and their comparison to human problem-solving abilities. The tone is skeptical and challenges the claim about a teenager's ability to solve a complex problem without errors. The intent is to question the validity of a previous assertion in the conversation.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm and is part of a technical debate, maintaining a neutral tone. [+1]
  2. 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 questioning a claim, without resorting to personal attacks. [+1]