The beginning of an interesting thread about NLP:
well, all true, but at least they work. like, previous-gen models *also* didn't understand meaning, and *also* considered only form. they were just much worse at this. so much worse that no one could ever imagine that they capture any kind of meaning whatsoever. they didn't work.
— (((ل()(ل() 'yoav))))👾 (@yoavgo) August 29, 2021
The end of that thread:
and the thing that we *can* do to improve result, like futzing around with prompts and small architectural tweaks, i find kinda arbitrary and intellectually dull. but hey, at least these things work.
— (((ل()(ل() 'yoav))))👾 (@yoavgo) August 29, 2021
