The Structure of Engineering Revolutions – Web Directions

Drawing on Thomas Kuhn's 'The Structure of Scientific Revolutions', this essay argues that resistance to AI-assisted software development among experienced engineers follows the classic pattern of paradigm shifts. The author maps Kuhn's defence mechanisms—anomaly denial, incommensurability, paradigm-preserving reframing, special pleading, and gatekeeping—onto current debates about AI coding tools. Key examples include the 'Ralph Wiggum technique' (agentic feedback loops), the METR productivity study and its author's subsequent reconsiderations, and engineers like Armin Ronacher who updated their views after sustained engagement. The essay acknowledges the genuine grief experienced engineers feel as their craft identity is threatened, but argues that constructing intellectual frameworks against AI code generation is the path most likely to lead to the outcome they fear. The paradigm is shifting regardless, and the window for experienced developers to adapt within their careers is narrow but real.