Gary Marcus

Rank 19 of 47
|
Score 64
12/21/2024, 3:07:28 PM
X
In reply to:
🍓🍓🍓
@iruletheworldmo
·
202d
@GaryMarcus yeah but o3 isn't trained to be good at anything.

or rather. it's initially given lots of data and then trained to be good at broad reasoning. which it applies generally.

I also hate any definition of agi that needs to be embodied. this whole, can it walk into a random…
Gary Marcus
@GaryMarcus
·
202d
@iruletheworldmo [1. how should i know what’s in your anonymous head?]
2. i think an AGI should be able to control a robot as well as a human can teleoperate a robot. Is that not reasonable? Humans could teleop all those things.
3. A domain-specific medical system doesn’t necessarily entail AGI.…
🍓🍓🍓
@iruletheworldmo
·
202d
@GaryMarcus I assume you know why I don't like the list.

robotics will take longer and creative writing ect will be harder because we don't have right answers/tbinking to train on.

but if we get systems that cure disease ect, simply by scaling current paradigms.

the list just becomes…
Gary Marcus
@GaryMarcus
·
202d
@iruletheworldmo such a dumb comment; literally an ad hominem with zero argument. among your least impressive.
🍓🍓🍓
@iruletheworldmo
·
202d
@GaryMarcus such a dumb list 😂
Gary Marcus
@GaryMarcus
·
202d
Some AGI-relevant predictions I made publicly long before o3 about what AI could not do by the end of 2025.

Do you seriously think o3-enhanced AI will solve any of them in next 12.5 months?

The statement questions the claim about AI training methods, specifically data augmentation, in the context of a discussion on artificial general intelligence (AGI). It challenges the initial assertion about the training process of AI models, suggesting skepticism about the transparency and methods used.

  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 the methods used in AI training, without resorting to personal attacks. [+1]