Solomon's approach to the USG / Lab Question
Its a question of which frontier
Researcher working on task automation with AI
Its a question of which frontier
Terminal agents are the path to scalable applied research engineering.
... Know Victory
... is pain and suffering
... is a bet on better code gen and more complexity
... is sometimes worth it
... and screams in the middle
... extends far beyond reefkeeping.
... Doesn't Get It
... Right Action
... but plenty of hidden, important lies
... to fuel the next few OOMs in software agent research.
... to assess language models' ability to reason about lots of information.
... are here to stay.
... and never will be.
... may be nigh for consumer tech.