Traditional productivity metrics like lines of code are misleading, especially with AI generating large volumes of unverified code. Instead, true engineering performance hinges on developer experience (DevEx), which prioritizes flow state, manageable cognitive load, and rapid feedback loops. AI can enhance productivity by automating routine tasks and aiding prototyping, but it also introduces new challenges—such as increased review burden and disrupted focus—that can hinder deep work. The key is to treat DevEx as a product, using structured frameworks like the upcoming seven-step model in *Frictionless* to identify pain points, run experiments, and measure impact through both developer-centric and business-aligned outcomes. Early wins come from lightweight process improvements and strong leadership support, while long-term gains require dedicated DevEx teams that use well-designed surveys and telemetry to guide change. Ultimately, AI’s value is maximized not in isolation, but when integrated into a holistic strategy that values human judgment, team health, and purposeful innovation.