alt Hacker NewsA capability isn't binary. There is a massive difference between can produce an impressive demo and can reliably complete the task without a human babysitting it.
A capability isn't binary. There is a massive difference between can produce an impressive demo and can reliably complete the task without a human babysitting it.
Yup, new SOTA models especially with high/xhigh/max reasoning too often overengineer solutions, good for benchmarks that usually measure task completion, bad for normal development where you don't want 'rewrite in rust and 1k LOC unit tests' style solutions when agent does mundane bug fixes.