Gary Marcus

Rank 16 of 47

Score 76

-3

@DotCSV ha ha. no try them on the
tasks that each system is actually built for.

you are conflating coverage with accuracy, and deliberately misinderstanding my point.

6/18/2025, 12:50:32 PM

In reply to:

Carlos Santana

@DotCSV

18h

@GaryMarcus Sure, here you go! Although I’m not sure it supports the point you were trying to make...

Gary Marcus

@GaryMarcus

Now show me the same graphs but for symbolic systems like calculators, spreadsheets, gps navigation systems, and databases.

Benjamin Todd

@ben_j_todd

Why can AIs code for 1h but not 10h?

A simple explanation: if there's a 10% chance of error per 10min step (say), the success rate is:

1h: 53%
4h: 8%
10h: 0.002%

@tobyordoxford has tested this 'constant error rate' theory and shown it's a good fit for the data

chance of

The statement is part of a technical discussion about the performance of symbolic systems versus AI systems. The tone is somewhat dismissive and accusatory, suggesting a misunderstanding or misrepresentation of the data. The intent seems to be to correct or challenge the previous point made by @DotCSV.

Principle 1:
I will strive to do no harm with my words and actions.
The statement could be seen as slightly harmful due to its dismissive tone, which might discourage open dialogue. [-1]
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 dignity of the other participant by implying deliberate misunderstanding. [-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 does not engage constructively and instead uses a personal attack by accusing the other of deliberate misunderstanding. [-1]