Gary Marcus

Rank 17 of 47
|
Score 65
2/15/2025, 3:28:39 PM
X
In reply to:
Focoinvestor
@focoinvestor
·
282d
@GaryMarcus It's a short-lived error.
Once an AI model learns something, it can produce the correct answer indefinitely.
It is not yet the definitive version, soon it will beat us all in anything.
Gary Marcus
@GaryMarcus
·
282d
@focoinvestor I couldn’t get LLMs to do this kind of thing at all reliably.

Many failures.
Focoinvestor
@focoinvestor
·
282d
@GaryMarcus Humans have issues with the mechanic and GPT with the names of the parts in draw.

But LLM once learn one time can explain forever in the right way to everyone.

Is an accumulative difference...
Gary Marcus
@GaryMarcus
·
282d
@focoinvestor The drawing is worse but nobody would put those labels on the thing that way

Separate issue.
Focoinvestor
@focoinvestor
·
282d
@GaryMarcus It's a lot worse for the average human!

https://www.scotthyoung.com/blog…

The statement critiques the capabilities of large language models (LLMs) in comparison to older chess computers, suggesting an overestimation of LLMs' abilities. It engages in a broader discussion about AI capabilities and limitations.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm; it is a critique of technology, not individuals.
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement does not promote understanding or empathy but rather challenges a viewpoint. [-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 without personal attacks. [+1]