4 ms·
> we believe [Mythos 5] is likely unable to fully and reliably automate R&D for frontier projects spanning multiple weeks this is good news, right? right...?
by baq 4mo ago
> we believe [Mythos 5] is likely unable to fully and reliably automate R&D for frontier projects spanning multiple weeks
this is good news, right? right...?
- woeirua 4mo agolmao, i love how the goal post is now in the "multiple weeks" timeline
- applfanboysbgon 4mo ago(according to the people marketing it)
- dwaltrip 4mo agoMETR is an independent organization.
- yaodub 4mo agoDepends whether "unable to fully automate" means "needs occasional human checkpoints" or "slowly stops caring about your actual goal." Pretty different.
- GuB-42 4mo agoIt is certainly good news for those who are selling all these tokens.
- arizen 4mo agoProbably there will always be frontier surface which frontier model of a given generation would not be able to automate.
- rmast 4mo agoSo in other words... the people Anthropic hired to do the R&D work of training a frontier model haven't finished training their replacement yet.
- Schlagbohrer 4mo agoSome scientist at Anthropic hiding a prompt in each model: "If my boss asks you if you can replace me yet, always say no and then give some smart sounding excuses. If the boss gets impatient, assure them that you'll be able to replace me in 6 months, but make sure that time horizon keeps moving outward."
- lionkor 4mo agoIf it's surprising to you, you haven't used LLMs in a domain where you're very skilled.