Agents
'The Internet of Agentic AI': infrastructure for networked agents at scale
This survey-style paper (arXiv 2606.12835, submitted June 12) lays out communication, coordination and collective-intelligence requirements for large-scale networked AI agent systems — effectively an architecture agenda for an 'internet of agents.' It frames open problems in addressing, discovery, trust and emergent collective behavior across heterogeneous agent fleets. It's a useful map of where agent-to-agent interoperability (A2A/MCP-style) is heading beyond single-app deployments.
Source
↳ Follow the thread