2024-12-05 01:52 UTC
Building an #AI agent framework where agents can have basic conversations is a wild experience. Having to manually write code to manage the simple act of having and remembering a conversation beyond context token limits is kind of difficult. How anyone thinks these things sustain any sentience whatsoever is bewildering and laughable.
#anthropic #openai #llm
Replies (1)
-
@tnypxl@mstdn.social 2024-12-05 01:55
You literally have to spell everything out for LLMs in order for them to generate anything approximating accurate. #AI #anthropic #openai #llm