Gary Marcus

Rank 13 of 47
|
Score 115
7/22/2025, 5:46:46 PM
X
In reply to:
Phil Libin
@plibin
·
49d
@GaryMarcus Ima start saying “Who? Who?” a lot and see if I can also subliminally transfer my love of owls.
Gary Marcus
@GaryMarcus
·
50d
@plibin and stop putting them in mission-critical/safety-critical applications
Phil Libin
@plibin
·
50d
Alignment was the fuzziest concept in D&D and now it’s the fuzziest concept in LLMs. Let’s just call them chaotic neutral and move on to something more rigorous.
Gary Marcus
@GaryMarcus
·
50d
Another day, another completely unexpected @OwainEvans_UK result.

LLMs are weirder, much weirder than you think.

Good luck keeping them safe, secure, and aligned.

The conversation humorously discusses the concept of alignment in AI, using a playful analogy involving owls. It touches on the challenges of ensuring AI safety and alignment, a topic of public interest, but does so in a lighthearted manner.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement uses humor to discuss AI alignment, which could help in making the topic more accessible, thus 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 playful tone promotes understanding and engagement with a complex topic, aligning with the principle of promoting understanding and empathy. [+1]