Gary Marcus

Rank 13 of 47
|
Score 115
7/22/2025, 3:25:45 AM
X
In reply to:
Boaz Barak
@boazbaraktcs
·
54d
@GaryMarcus The benchmark thing is a strawman, I don't really care about GPQA. The fact that you can ask o3 to go search up the numbers, draw the images, explain its work, shows how much of a step change it is from 4o. (And models are getting better all the time- ChatGPT Agent just came
Gary Marcus
@GaryMarcus
·
54d
@boazbaraktcs 1. I plotted some progress, not none. So you’re post is strawman that doesn’t correctly capture what I plotted. (Of course what I plotted was not to be taken literally. Sorry if that was not clear.)

2. We all know there has been a LOT of focus on benchmarks and virtually no
Boaz Barak
@boazbaraktcs
·
54d
Yes we really have made no progress
Gary Marcus
@GaryMarcus
·
54d
GPT-5: Last Year’s Dreams vs This Year’s Reality

The statement is part of a public discourse on the progress and challenges of deep learning models. It addresses a perceived misrepresentation of the speaker's views and reiterates concerns about the limitations of deep learning.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement aims to clarify a misunderstanding, which aligns with striving to do no harm by correcting potential misinformation. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    By addressing the misrepresentation, the statement promotes understanding and seeks to clarify the speaker's original point. [+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 by focusing on the content of the argument rather than attacking the individual. [+1]