I'd bet against this. The way AI works and learns information is fundamentally different from humans. Humans have continual learning - each human can absorb and specialise in a certain thing in the organisation and this leads to the firm organisation's shape.
LLMs have no specialisation. Subagents specialise only through the means of the context it was given. And you can have infinite subagents and infinite subsets of context. You can't do that with humans.
These fundamental differences will result in different ways to organise. At least until we solve continual learning.
I'd bet against this. The way AI works and learns information is fundamentally different from humans. Humans have continual learning - each human can absorb and specialise in a certain thing in the organisation and this leads to the firm organisation's shape.
LLMs have no specialisation. Subagents specialise only through the means of the context it was given. And you can have infinite subagents and infinite subsets of context. You can't do that with humans.
These fundamental differences will result in different ways to organise. At least until we solve continual learning.