Gary Marcus

Rank 13 of 47
|
Score 83
4/28/2025, 4:07:49 PM
X
In reply to:
Sicarius
@soumilrathi
·
262d
@GaryMarcus Don't you think personality misalignments can happen with AGI as well?

Definitely not the lying, but I'd assume you could end up making an AGI more syncophant-y by mistake.

To me, that's what AI alignment is all about
Gary Marcus
@GaryMarcus
·
262d
If we actually knew how to build AGI this sort of stuff wouldn’t happen.
Sam Altman
@sama
·
262d
the last couple of GPT-4o updates have made the personality too sycophant-y and annoying (even though there are some very good parts of it), and we are working on fixes asap, some today and some this week.

at some point will share our learnings from this, it's been interesting.

The statement engages in a discussion about the potential for AGI (Artificial General Intelligence) to follow constraints reliably, in response to concerns about personality misalignments in AI systems. This is a substantive engagement with public issues related to AI development and alignment.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm and is a neutral contribution to the discussion, focusing on the potential for AGI to follow constraints. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by addressing the technical aspect of AGI alignment, contributing to 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 constructively with the previous comments, focusing on the technical aspects rather than personal attacks. [+1]
  4. Principle 7:
    I will uphold the principles of free speech and use my platform responsibly and with integrity.
    The statement upholds free speech principles by responsibly contributing to the discourse on AGI alignment. [+1]