Most developers say AI makes them more productive. Most companies say their teams aren't shipping faster. What's going on?In this special solo episode, Romain unpacks the data behind the "acceleration whiplash," walks through a four-level AI maturity framework built from hundreds of customer conversations, and shares the practical lessons that separate teams getting incremental gains from those achieving 10x outcomes. This is a recording of his AI-DLC presentation, delivered at the AWS Summit in Zurich and iterated based on feedback from startups, enterprises, and digital-native companies across Europe.Key topics covered: • The adoption paradox: individual productivity up, team delivery flat • Why vibe coding is a mirage for production systems • Theory of constraints applied to AI-era software development • The four-level AI maturity framework: Traditional, AI-Assisted, AI-Augmented, and AI-Native • From prompt engineering to context engineering to loop engineering • Cross-functional teams and the evolution of Amazon's two-pizza teams • AI-DLC workshops: mob elaboration, construction bolts, and continuous delivery • Building AI fluency across your organization • The technology stack for AI-native development • Werner Vogels' Renaissance Developer and T-shaped engineersChapters: