Sounds like how most things are solved in computer science. Reinforcement Learning or Genetic Algorithms are just creating a bunch of garbage and finding the few that actually make sense. Here they're probably generating a ton of slop and searching through what also makes sense
Out of curiosity, how does one build something that coordinates 10K AI agents to solve a single problem?
1
223
That seems to me like a prompting issue. You're telling an LLM that a code with no bug has a bug. He's now trying to figure out why the code with no bug is not behaving like you want it to behave. That seems like a feature, not a bug. The issue is with the prompt.
when you ask an LLM to audit your code, it is essentially answering: "does this code have some pattern that *looks like* a bug?" this is important because many things can "look like" a bug, while still being correct. and "fixing" something that is correct will make it incorrect.
I've ran an experiment where I set up a loop:
- GPT 5.6 Pro audits an algorithm
- Fable 5 to fixes the bugs GPT found
- repeat
with a catch: the algorithm was *already correct* (literally, proven so). the result is complete degradation: in the first iterations, the correct algorithm becomes incorrect. in the next iterations, the (now incorrect) algorithm gets patched. but the patches introduce new cases and bugs. this keeps going until the code is destroyed.
so, a naive loop where a model audits its own code ("adversarially") and then fixes it will lead to failure. of course, as LLMs get smarter, *looks like* a bug will converge to being equivalent, in practically 100% of the cases, to finding a real bug. I think we're very close to that, and it will be interesting when the error rate is so low that these loops start working. but, as of 2026, we're not quite there yet
4
4
306
People saying this fail to realize thay if the Bosnia player had stepped on Bologun's ankle like that and Bologun was out of the world cup with a bad injury they would be just as mad.
The entire dataset from all of the World Cup games should be open source.
El software del VAR de @FIFAcom debería ser open source y auditable. ¿Cómo sabemos que no están moviendo los jugadores unos centímetros según para que país juegan o según quiénes sean? Podríamos estar ante el robo del siglo.
