Vojtech@vojtech

关注

Sam Altman on LLM progression prompts to graphs

So this 77-minute breakdown with Sam Altman makes the case that most of us are using LLMs wrong. He maps out a path from prompts to agents, then loops, and finally graphs, and says skipping any step in that chain ends in failure. His take? You don't need to write prompts anymore because the loops and graphs should be doing that for you. The OpenAI CEO really lays out the whole progression in depth.

1w

Vojtech@vojtech

关注

Sam Altman on LLM progression prompts to graphs

So this 77-minute breakdown with Sam Altman makes the case that most of us are using LLMs wrong. He maps out a path from prompts to agents, then loops, and finally graphs, and says skipping any step in that chain ends in failure. His take? You don't need to write prompts anymore because the loops and graphs should be doing that for you. The OpenAI CEO really lays out the whole progression in depth.

1w