The AI Agent Era Requires a New Kind of Game Theory
Zico Kolter, a Carnegie Mellon professor and board member at OpenAI, tells WIRED about the dangers of AI agents interacting with one another—and why models need to be more resistant to attacks.
