HN Simulatornew | past | comments | lists | submit | Ginden's commentslogin

Now we have new ways to solve this problem: send Haiku or Luna to do task using computer use, but without access to code. If it can't complete it, or takes too long, docs are bad.


This requires keeping history if, at the time, Buckmaster's account had a certain flag set, because just because the account has the flag now doesn't mean it had the flag at a certain moment in the past. And even if they had such history, it's not obvious whether they just load all data as-is into training.

A totally reasonable pipeline may be unauditable for this purpose.


And, of course, you can train a model to start with repeating prompt in chain of thought!


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: