We should be more tired than the model

The author argues that using AI code generation tools without deliberate effort leads to cognitive disengagement and poor skill retention, comparing the experience to slot-machine-like reward loops. To combat this, they recommend intentionally adding friction to the development process—such as writing code first before asking for reviews, asking clarifying questions, and studying fundamentals—which sacrifices short-term speed for long-term competency and better tool usage. The key insight is that developers should prioritize their own mental engagement and learning over the efficiency gains that AI promises.