@Alex_m Our plan is to continue to offer Fable 5 as part of subscriptions, but we're being extra conservative because it's hard to forecast demand. More details at the bottom of the launch announcement https://t.co/HuqU1Qo4ZE
Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use.
Its capabilities exceed those of any model we’ve ever made generally available.
@kushbhuwalka Skills are subjective so kind of hard to standardize benchmarking. But I think something like DSPy for skills could work though. The simplest could be asking the agent to write down the mistakes & suboptimizations. At the end of the session analyze it and modify the skills used.
OMG I can't believe this ... @ICOffenderII I didn't think you were serious lol thank you so much. And take this @0xTimB1
My jaw is permanently dropped now
if you’re still writing loops that prompt coding agents you’re falling behind. you need to build a meta agent that infers what loops you would have wanted based on your vibe and then write those loops