News for hackers. Stuff that matters.
Re:The Unreasonable Effectiveness of an LLM Agent Loop with ... by johnsmith1840 on Thursday May 15, 2025 @05:51PM (#43999689)
Drop in replacement files per update should be done on the heavy test time compute methods.

o1-pro, o1-preview can generate updated full file responses into the 1k LOC range.

It's something about their internal verification methods that make it an actual viable development method.

Re:The Unreasonable Effectiveness of an LLM Agent Loop with ... by nico on Thursday May 15, 2025 @06:32PM (#44000042)
True. Also, the APIs don't care too much about restricting output length, they might actually be more verbose to charge more

It's interesting how the same model being served through different interfaces (chat vs api), can behave differently based on the economic incentives of the providers