Gary Marcus

Rank 19 of 47
|
Score 64
12/21/2024, 3:05:36 PM
X
In reply to:
Jay Jakosky
@jayjakosky
·
202d
@GaryMarcus When Tesla engineers speak, it’s been very accurate. So yes.
Gary Marcus
@GaryMarcus
·
202d
@jayjakosky omg you are going to take a promise from Tesla as data?
Jay Jakosky
@jayjakosky
·
202d
@GaryMarcus Tesla’s engineers are saying that autonomy will fall this year.

“Novel game” is a great goal for inference compute.

Movie summary seems analogous to what Grok does with X posts.

We lack off road training data, otherwise doable with autonomy AI.

Most others do seem unlikely.
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 critiques the reliability of predictions about AI autonomy, specifically regarding Tesla's timeline for solving autonomy by 2025. It questions the accuracy of Elon Musk's predictions, suggesting they may be overly optimistic or based on unreliable sources.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not directly harm but questions the reliability of predictions, which can be seen as a form of constructive skepticism.
  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 validity of certain claims without resorting to personal attacks. [+1]