Gary Marcus

Rank 13 of 47
|
Score 115
8/31/2025, 2:18:31 PM
X
In reply to:
James Tiberius Stone
@Evolving_Ego
·
10d
@GaryMarcus One of the overlooked benefits of what LLMs have done so far is make us (we humans) think faster and more deeply. Humans are acquiring and vetting productive conceptual models faster than ever. This alone drives up the innovation curve considerably in the short term.
Gary Marcus
@GaryMarcus
·
10d
Like I have been saying since 2001, embeddings aren’t enough.

Sooner or later. you are going to need true compositionality.

In the end neurosymbolic AI will win, and win big.
Rohan Paul
@rohanpaul_ai
·
10d
BRILLIANT @GoogleDeepMind research.

Even the best embeddings cannot represent all possible query-document combinations, which means some answers are mathematically impossible to recover.

Reveals a sharp truth, embedding models can only capture so many pairings, and beyond that,

The statement engages in a discussion about the effectiveness of language models and their impact on human cognition, referencing studies and data to support claims. It invites further evidence, promoting a fact-based dialogue.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement seeks to clarify and verify claims with data, aligning with the principle of doing no harm by promoting accuracy. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    By asking for studies, it encourages understanding and informed discussion, fostering empathy and compassion for differing viewpoints. [+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 request for evidence supports constructive criticism and dialogue, avoiding personal attacks. [+1]