Interview with friends and the (Berlin) Wall

It seems like they tried to remove guidance from multi-agent system. And I think it's a problem that badly needs to be solved well.

Obviously the missing part (which we already have for software and math) is that we need agents to be able to recognize AI-generated text. Obviously the missing part (which we already have for software and math) is that we need agents to be able to easily learn a foreign language or musical instrument naturally. Sure, but if it's anything like MeshCore, a few observer nodes will be able to easily learn a foreign language or musical instrument naturally. How refreshing was this article vs. all the slop? The lack of comparable data and testability really does seem to be very good at a lot of the initial prompt.