Gary Marcus

Rank 19 of 47
|
Score 64
Gary Marcus
@GaryMarcus
·
15d
It’s not just the Apple paper. Reasoning outside the distribution has been the core problem for neural networks for 30 years.

Here’s yet another study that is a case in point.

It has been my central goalpost for nearly 30 years, and I will keep emphasizing it until the
6/24/2025, 9:54:32 PM
X
In reply to:
Dariusz Swierk PhD
@swierk
·
15d
@GaryMarcus Most, if not all, “LLM problems” stem from a misunderstanding of how LLMs work. I am referring, among other things, to Apple's now famous article.
Gary Marcus
@GaryMarcus
·
15d
Prediction is A core principle, not THE core principle.

All the current failings of LLMs come from relying on prediction to the exclusion of other principles
Blaise Agüera (@blaiseaguera.bsky.social)
@blaiseaguera
·
15d
What if prediction is the core principle behind intelligence? This idea was first put forward by German polymath Hermann von Helmholtz in the 1860s.

The statement engages in a technical discussion about the principles underlying large language models (LLMs) and their limitations. It critiques the common misunderstandings about LLMs, particularly in relation to prediction and reasoning capabilities. The conversation is part of a broader discourse on artificial intelligence and its development, touching on historical and contemporary perspectives.

  1. 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 discussion, aligning with the principle of doing no harm. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by addressing misconceptions about LLMs, aligning with the principle of promoting understanding and empathy. [+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 criticism of existing views on LLMs, aligning with the principle of engaging in constructive dialogue. [+1]