Hot take: frontier models haven’t actually become much smarter
I know this is a controversial take, but I would like to hear others’ opinions: LLMs have become much more useful, but I’m not convinced their “raw intelligence” has improved nearly as much as their overall performance suggests.