Comment by golol

People have played with (multi-) agentic frameworks for LLMs from the very beginning but it seems like only now with powerful reasoning models it is really making a difference.