News for hackers. Stuff that matters.
Re:November 2026 Claude Fable 5.2 Scores 65 on Artificial An... by baq on Monday November 09, 2026 @05:22PM (#50433124)
fable is a high agency model (some people call it 'big model smell'). if you want something that blindly follows instructions, use sonnet. I want to be challenged by the thing since I expect it to be smarter than me at everything except the little part of human knowledge I'm an actual expert in.
Re:November 2026 Claude Fable 5.2 Scores 65 on Artificial An... by jasomill on Monday November 09, 2026 @05:41PM (#50433320)
Do you also want it to refuse to continue the conversation because it's suspicious of your motives?
Re:November 2026 Claude Fable 5.2 Scores 65 on Artificial An... by baq on Monday November 09, 2026 @05:59PM (#50433542)
not really, no - but I almost exclusively use the biggest models as advisor and reviewer subagents, so if they refuse to their smaller siblings I don't get to see it
Re:November 2026 Claude Fable 5.2 Scores 65 on Artificial An... by xvector on Monday November 09, 2026 @07:05PM (#50434256)
No, but I certainly want it to do that for random internet strangers more than I don't want it to do that for me.