@eric23332 Probably (although it is a moot point). There are already five open-weights models that exceed Opus 4.1 on autonomous coding tasks, so we are already exposed to tremendous risk per unit time even if no new models are trained.
same thread: 2026721001149522325 2026721004949553444 2026721008216936651 2026721013279474063 2026772151467221160 2026775882161910249 2026778471708688432 2026780379802202206 2026999464016457847 2027000554116022344 2027110409493217307 2027111529313698097 2027112393193468237 2027145232253419576
cited on: claude-opus-4-1
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.