Gary Marcus

Rank 11 of 47
|
Score 83
9/3/2024, 2:41:01 PM
X
In reply to:
Jordan Talks Everyday AI
@EverydayAI_
·
519d
@GaryMarcus Sure. Get the point. But to tell a model to "sort" without context (sort what? sort how?) is confusing.

That's even confusing to humans.

I'd say your example is less about model quality as it is about clear communication.
Gary Marcus
@GaryMarcus
·
519d
@EverydayAI_ The point is that the system should either do what you ask or tell you it can’t, not make something up and expect you to notice and write a different prompt.
Jordan Talks Everyday AI
@EverydayAI_
·
519d
That's a poorly written prompt. Any model would struggle.

Still not perfect, but Grok does much better with proper prompting.

Look at this: https://x.com/i/grok/share/9hAX2…
Gary Marcus
@GaryMarcus
·
519d
I just spent 15 minutes trying to get Grok to solve a very basic information integration task—“sort all the words in the Gettysburg address and boldface the prepositions”—and it made mistakes literally in every try.

I literally could not get a correct answer out of it. It would…

The statement by @GaryMarcus constitutes public discourse as it engages in a substantive discussion about the capabilities and limitations of AI models, which is a relevant public issue in the context of technology and its impact on society. The tone is critical but aims to provide a comparative analysis between human and AI performance.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not appear to cause harm with words and actions. It is a critique of technology rather than individuals, so it aligns with this principle. [+1]
  2. Principle 2:
    I will respect the privacy and dignity of others and will not engage in cyberbullying, harassment, or hate speech.
    The statement respects the privacy and dignity of others. It does not engage in cyberbullying, harassment, or hate speech. [+1]
  3. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement does not explicitly promote understanding, empathy, or compassion, but it does contribute to a broader understanding of AI limitations.
  4. 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 of the AI model without resorting to personal attacks or ad hominem arguments. It focuses on the performance of the AI rather than attacking individuals. [+1]
  5. Principle 7:
    I will uphold the principles of free speech and use my platform responsibly and with integrity.
    The statement upholds the principles of free speech and uses the platform responsibly to discuss a relevant public issue. [+1]