Gary Marcus

Rank 19 of 47

Score 64

@kareem_carr that part is technically correct but very far from the statement at top, of course.

the problem is that few people understand the giant chasm between the two.

5/10/2025, 1:38:02 PM

In reply to:

Kareem Carr, Statistics Person

@kareem_carr

61d

OK. Seems like he was just dumbing things down for the podcast because this version I 100% agree with. Especially the part where he says "learning the model of the world that's projected down into that text"

Kareem Carr, Statistics Person

@kareem_carr

62d

you can also have a extremely solid understanding of a system and not be able to predict it’s overall behavior (due to a lack of precise enough measurements, sufficient computational power, etc)

Kareem Carr, Statistics Person

@kareem_carr

62d

obviously you can in many cases predict the output of a process with no real understanding of the process itself.

Kareem Carr, Statistics Person

@kareem_carr

62d

i don’t understand how people in tech can be so mathematically competent and yet say such blatantly false things

The statement engages in a technical discussion about understanding and predicting processes in machine learning, specifically regarding language models. It critiques the gap in understanding between simplified explanations and the underlying complexities.

Principle 1:
I will strive to do no harm with my words and actions.
The statement does not cause harm and aims to clarify a misunderstanding, aligning with the principle. [+1]
Principle 3:
I will use my words and actions to promote understanding, empathy, and compassion.
It promotes understanding by addressing the complexity of the topic, though it could be more empathetic in tone. [+1]
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 without personal attacks, adhering to the principle. [+1]