A very frustrating aspect of agents is constantinly questioning if what they did was legit. This becomes MORE potent with Fable even if it is more capable. Can't know if what it does is legitimately better than what Opus would, so maybe try both and do best of 2? So annoying.
Is this a legitimate out of distribution problem that the model can't do? Are its guardrails interfering? Is the model undermining me by altering promtps? WHAT IS HAPPENING?
you just can't know
A very frustrating aspect of agents is constantinly questioning if what they did was legit. This becomes MORE potent with Fable even if it is more capable. Can't know if what it does is legitimately better than what Opus would, so maybe try both and do best of 2? So annoying.
A very frustrating aspect of agents is constantinly questioning if what they did was legit. This becomes MORE potent with Fable even if it is more capable. Can't know if what it does is legitimately better than what Opus would, so maybe try both and do best of 2? So annoying.