Gary Marcus

Rank 14 of 47
|
Score 83
5/14/2025, 9:41:32 PM
X
In reply to:
Robert Scoble
@Scobleizer
·
227d
@nxthompson @GaryMarcus I used to like Gary's perspective, but lately many companies are outrunning him. Cognitive AI is the way to get us there. And LLMs play a role, but the best companies are using constellations of AIs to fix the problems Gary laid out. He needs a new approach, me thinks.
nxthompson
@nxthompson
·
227d
I so very much enjoyed talking with @GaryMarcus about why he doesn't think that the path to better AI is through creating bigger models and using more data. We get into:

1. Scaling laws: why he thinks they won’t last forever & why he thinks they’ve hit diminishing returns.

2.

The statement engages in a discussion about the current state and future of AI, particularly cognitive AI and large language models (LLMs). It questions the effectiveness and economic viability of current AI trends, highlighting issues like hallucinations and errors. The tone is critical and skeptical, aiming to provoke thought and discussion about AI development.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm but raises concerns about AI's current limitations, which can be seen as a constructive critique. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    By questioning the current state of AI, it encourages understanding and reflection on the technology's development and challenges. [+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 questioning the effectiveness of current AI models and practices without resorting to personal attacks. [+1]