[21asdffdsa12]

To be honest humans often have no overview over a application either. We navigate up and down the namespace, building the "overview" as we go. Nothing i see what prevents an agent from moving up and down that namespace, writing his assumptions into the codebase and requesting feedback from other agents working on different sets of the file.

[airstrike]

"Nothing" prevents it other than the fact that an agent doesn't really have memory and has a pretty limited context, hallucinates information, mistakes metadata with data, and so on.

The path forward is always one that starts from the assumption that it will go wrong in all those different ways, and then builds from there

[21asdffdsa12]

So one hub architect agent for overview- which generates tokens for the spoke agent and receives architectural problem reports from spoke agents?

[airstrike]

That seems to be the most common "fix" people deploy today, but it's still structurally flawed due to inherent limitations in LLMs.

[21asdffdsa12]

Everything is a compromise- everything is "structurally" flawed. Who is to say- that the brain itself is not something similar, a labyrinthine city, with domain specific shortlived LLMs "excitements" that wake up other short live LLMs? Where not resolving an excitement builds up a frustration counter until a new agent is born whos sole purpose is to solve this problem?

[don_esteban]

A hybrid approach is needed, precise/algorithmic manager, with LLM workers.