I think we have been working on the same thing, same vision

[gustofied]

cool, nice project

would love to talk

[gustofied]

I've been reading this great blog post series on multi-agent systems: https://www.arnovich.com/snippets/state_machines_for_multi_agent_part_4.

The global reasoner and Erik propose some solid approaches, and I think we could add another layer on top, perhaps a market-based system as an optimizer to find the best trajectory. Im thinking of injecting economic models as the control loop over the global reasoning paths.

I bit yappy of me to say, but I will be clearer in my message in the future

[andrewhopper]

Good read, thanks. Yes, LLM orchestration is just state machine orchestration.

Love the idea of using economic models for control paths.

Happy to chat.