The latest source-first signal on AI agents, agent workflows, and companies building with autonomous AI systems.
Signal brief
Drone-Bench shows frontier models nearing autonomous drone control without yet reaching dependable end-to-end performance. The strongest model cleared four of five subtasks, but reconstruction errors cascaded into localization and navigation; average reliability still trails one-off best results by roughly six months.
Read signalAI agents break a core endpoint-security assumption: actions can no longer be attributed to the person at the keyboard rather than software acting with that person’s authority. a16z’s Neo thesis reframes control around real-time identity, agent guardrails, and attribution at execution; it remains an investment claim.
Read signal