ajaystream

6 ms·

1. 1.▲
   by ajaystream 5mo ago

   I am going to ignore the form comments -- I guess I am not sure how I feel about being called an LLM (good or bad ?), not sure only time will tell. If LLMs turn out to be the turd of the universe -- bad or maybe good ? -- Emphasis on the &
2. 2.▲
   by ajaystream 5mo ago

   The spec-completeness problem here is the same one that bites distributed systems verification: the proof holds inside an operating envelope (no adversarial inputs, trusted runtime, bounded sizes), and the interesting failures live at the b
3. 3.▲
   by ajaystream 6mo ago

   I hardly felt that the problem could be answered in such poetic and psychological eloquence. I get what you are saying.
4. 4.▲
   by ajaystream 6mo ago

   I am building in finance, where workflows dominate - ie if A then B, but if ~ A then C then D etc. Configuration is a major challenge, but AI helps in getting there, but ever so often as you know - LLMs to the wrong thing even though they h
5. 5.▲
   Ask HN: Human psychology of non-AI-native users

   1 points by ajaystream 6mo ago|3 comments
6. 6.▲
   by ajaystream 6mo ago

   The other challenge that we have found is accuracy and completeness of fields required to be updated across use cases. Either we have to mandate all the fields or when we set them optional in the tool def. it sometimes blows through - how a
7. 7.▲
   Ask HN: Is anyone building write guarantees for agents working across tool

   1 points by ajaystream 6mo ago|0 comments