Gary Marcus

Rank 13 of 47
|
Score 83
Tomer Ullman
@TomerUllman
·
493d
trying the new GPT-o1 'reasoning' model on a "navigation" task: get the cat to the mouse, with impenetrable wall in the middle.

o1 thinks for 23 seconds, then produces sequence that simply goes through the wall.
9/17/2024, 6:59:30 PM
X

The statement 'Oh Gary Marcus is so mean, o1 is so smart. 🙄' is a response to a conversation about the performance of the GPT-o1 model on a reasoning task. The tone of the statement is dismissive and sarcastic, and it includes a personal attack on Gary Marcus, calling him 'mean,' while sarcastically praising the model 'o1' as 'so smart.'

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not strive to do no harm; it includes a personal attack on Gary Marcus, calling him 'mean,' which can be harmful and is not constructive. [-2]
  2. Principle 2:
    I will respect the privacy and dignity of others and will not engage in cyberbullying, harassment, or hate speech.
    The statement does not respect the privacy and dignity of others; it engages in a personal attack on Gary Marcus, which can be considered a form of cyberbullying or harassment. [-2]
  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 does not engage in constructive criticism or dialogue; it resorts to a personal attack and sarcasm rather than addressing the issue with the model's performance in a constructive manner. [-2]