This is utterly daft to say if you actually used the thing for hard problems, something that benchmarks have been known to be unable to capture. It is night and day compared to Opus and every other model out there. It was nice while it lasted.
It's strange how uninformed people are when they are so willing to to make assertions. I used it too and it really felt like a generational shift and not an incremental one.
These threads about Anthropic always seem so astroturfed with some of the loudest and most uninformed people around.
It's really annoying when people seem to assume that their experiences and use-cases are universal, and that other experiences are "psychosis". Fable seems to excel at tasks that are research-level hard. The vast majority of people do not have these problems. FAANG engineers probably don't either. Codex is great, I use it too. It has a major advantage in terms of token limits so you can often brute-force things through. If it is a task that Codex can do with enough tokens, it isn't worth using Fable for it. I used Fable on the tasks that Codex and Opus consistently failed at. Delicate computational geometry tasks, mostly, really darn hard ones. It worked first try, every time, every single one. That's terrifying to me. I don't know what to tell you, I don't care about hyping Anthropic.
