What happens when thousands of AI agents work together? In TIME, Faculty Associate Noah Giansiracusa argues that the capabilities and risks of these “colonies” deserve as much attention as those of individual models. Drawing on examples of agents collaborating on mathematics and cybersecurity, he calls for a broader approach to evaluating AI systems.
Read the full essay in TIME.
You might also like
- communityClickwrap Accountability
- communityWhere better than universities?