Gary Marcus

Rank 19 of 47
|
Score 64
9/30/2024, 5:14:08 PM
X
In reply to:
James
@jmac_ai
·
285d
@GaryMarcus My experience is it's useful for doing API search for well documented libraries or writing self-contained code that mostly involves calling the right APIs.

It is awful for "real" programming and I can absolutely believe it's a net drag for people who try to use it for that.
Gary Marcus
@GaryMarcus
·
285d
What if literally the biggest putative benefit of LLMs—coding—turned out to be a mirage?

New study hints that this might be the case, perhaps (I speculate) all fueled by a confirmation bias in which people notice immediate benefits (e.g., code written faster) but discount the…
Gary Marcus
@GaryMarcus
·
285d
“Devs gaining little (if anything) from coding assistants” @CIOonline

The statement 'That’s what I would expect, as well' is a brief agreement with the preceding comment about the limited utility of AI coding assistants for 'real' programming. The conversation involves a discussion on the effectiveness of AI tools in software development, which is a public issue related to technology and productivity.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does no harm and is a simple agreement, thus adhering to the principle of striving to do no harm. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement does not directly promote understanding, empathy, or compassion, but it does contribute to a dialogue about the effectiveness of AI tools, which can be seen as promoting understanding in a technical context. [+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 dialogue by agreeing with a point made in the discussion without resorting to personal attacks or ad hominem arguments. [+1]