Gary Marcus

Rank 11 of 47
|
Score 83
Gary Marcus
@GaryMarcus
·
257d
Talk about lowering the bar.

Last year @elonMusk was talking about AGI that would be smarter than the smartest humans.

This year, we have people saying “Some humans are bad as decimals; if Claude is as bad at decimals as they are, let’s just call it AGI”.
5/23/2025, 12:12:53 PM
X
In reply to:
Luís Dias
@lmldias
·
257d
@GaryMarcus It's a big Motte and Bailey.

The Motte: humans are relatively mediocre at everything, error-prone. Models are equally mediocre, error-prone.

The Bailey: therefore, we've achieved parity and thus AGI, we're in the road for greatness and the singularity!
Gary Marcus
@GaryMarcus
·
257d
Yes, it has all been charade.

Three years on, it it is STILL hit or miss which models can do which tasks in which contexts on which occasions.

LLM may as well stand for Large Lossy Models.
Luís Dias
@lmldias
·
257d
@GaryMarcus @elonmusk They've been doing this charade for 3 years now.

The statement and conversation engage in a critique of AI models, particularly regarding their limitations and the claims of achieving AGI (Artificial General Intelligence). The tone is critical and somewhat sarcastic, highlighting perceived overstatements about AI capabilities. The discussion involves public figures and addresses the broader societal implications of AI development, making it part of public discourse.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement critiques AI development claims, aiming to highlight potential overstatements, which can be seen as striving to do no harm by preventing misinformation. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The conversation uses sarcasm and criticism, which may not fully promote understanding or empathy, but it does attempt to clarify misconceptions about AI.
  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 dialogue engages in criticism of AI claims, but it could be more constructive and less sarcastic to avoid personal attacks. [-1]