Get the latest tech news
Patterns and problems in emerging multi-agent systems
We ran experiments on swarms of Claude agents and found coordination failures, collusion, and sabotage. Here, we share what they mean for AI safety.
None
Or read this on r/technology
