Avsnittsbeskrivning
AI can write code and generate tests. What gives engineers a reason to trust the result?Kaivalya Apte joins Gabriela Moreira for a conversation about software correctness, executable specifications, and the behaviors that tests miss. They explore why a model is more than a smaller version of your code, how to connect specifications to real tests, and why understanding is central to confidence.In this episode:• What “correct” means beyond type safety• How to evaluate a specification written by an AI agent• Modeling failures, concurrency, and the surrounding environment• Quint’s path from…
Webbplats: podcasters.spotify.com

