Gary Marcus

Rank 14 of 47
|
Score 83
2/13/2025, 3:50:41 PM
X
In reply to:
Arvind Narayanan
@random_walker
·
340d
@emollick Here's my caveat to your caveat to my caveat 😁

In my experience so far, the main factuality problem is not hallucination, but source selection — the sources it relies on might sometimes be outdated, incomplete*, or not the highest quality. Assessing the quality of the source…
Ethan Mollick
@emollick
·
340d
I generally agree with this caveat, but…

… I have now spoken with C-level leaders & lawyers, who assigned teams to check Deep Research. They found few flaws & fact checking took much less time (and less skill) than was saved by AI.

That seems to be a critical tipping point.
Arvind Narayanan
@random_walker
·
340d
We're seeing the same form-versus-function confusion with Deep Research now that we saw in the early days of chatbots. Back then people were wowed by chatbots' conversational abilities and mimicry of linguistic fluency, and hence underappreciated the limitations. Now people are…

The statement is part of a public discourse on the reliability and challenges of AI in research, specifically addressing issues like temporal reasoning and source quality. It engages in a substantive discussion about the implications of AI in professional settings, such as research and fact-checking.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not use harmful language and focuses on discussing potential issues with AI, aligning with the principle of doing no harm. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by highlighting specific issues with AI, such as temporal reasoning and source quality, which can foster a more informed discussion. [+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 building on previous comments and asking questions, rather than attacking individuals, thus adhering to the principle of constructive criticism. [+1]