David Krueger–Coordination, Alignment, Academia
The Inside View
00:00
The Benefits of Multiple Agents in AI Coordination
I don't have a super strong intuition about this. I think the agents bouncing each other out thing is like, um, not, you know, it's, it's not necessarily better from an existential safety point of view. There's also these arguments that say if you have highly advanced AI systems, they should be able to solve coordination problems amongst themselves. So even if you start out with multiple agents, like they might, you know,. form a hive mind basically because they can solve these coordination problems that humans haven't been able to solve as effectively.
Transcript
Play full episode
