How Third Moment Research runs 10,000 agents for 36+ hours

Third Moment Research runs 10,000 AI agents for over 36 hours using GPT-6 Astra, GPT-6.1 Sol, Codex, and Daybreak for cybersecurity. These agents maintain focus for days and improve with use, enabling sustained autonomous operations in complex tasks. The lab's approach highlights advances in long-horizon agent reliability and self-improvement, with implications for AI-driven security and research applications.