Mammoth May🦣: Sean Li on Application Abstraction, Verifiable Agents, and AI’s Trust Problemhttps://t.co/vPFLta9d7H