Listen now | On raising $200m to build agent operating systems that can reason and code, why LLMs beat reinforcement learning for agent usecases, and how to build a Scenius with top AI people
Great session! I especially echo on the need of agents to be interactive with humans so that we can have input for adjusting naturally instead of going fully autonomous and god knows where it will end up.
Why AI Agents Don't Work (yet) - with Kanjun Qiu of Imbue
Great session! I especially echo on the need of agents to be interactive with humans so that we can have input for adjusting naturally instead of going fully autonomous and god knows where it will end up.
no reasoning data because no ones forayed into a unified cognition. cognitive sciences.