Conceptual integrity and counting lines of code
AI coding agents shift the bottleneck from code production to cognitive capacity, not headcount
“I can churn out code a hundred times faster. I don't have the cognitive capacity to stay on top of 100 times the amount of code.”
Simon Willison argues that AI coding agents make lines-of-code a valid productivity metric again, since 1000+ debugged lines/day is a genuine step-change from the pre-AI ceiling of 50-200. However, he warns that the real new constraint is cognitive capacity — engineers still can't mentally manage 100x more code — and that AI agents erode 'conceptual integrity' by making it trivially easy to bolt on features without architectural discipline, producing what Claire Giordano called the 'Winchester Mystery House' effect. This reframes the AI productivity debate from raw output to software quality and team cognitive load.