@theneurondaily: AI Cant Grade Itself 🤖📝 AI coding agents can generate increasingly complex software, but how do you verify that what they build actually works? Microsoft Research's Alex Lavaee explains why reliable AI coding agents need a foundation of real tests, from static analysis to end-to-end verification, before an AI model evaluates the results. ▶️ Watch our live replay to learn more. #AICoding #CodingAgents #AIAgents #SoftwareEngineering #AI #TheNeuron