what if we mapped older distributed systems patterns (like actor models or reactive, state-driven blackboards) to LLM agents?

→ View original post on X — @yoheinakajima