A GitClear analysis of 623 million code changes from 2023-2026 reveals that heavy AI tool users increased output by 25% compared to their prior velocity, but code duplication rose 81%. While AI-adopting teams outproduce peers by 4-10x, ROI remains unclear as output gains don't necessarily translate to business value, and code quality concerns are mounting.
A JPMorgan Chase software engineer with 15 years of experience reflects on how AI has fundamentally transformed coding work, making manual programming less economically viable while creating uncertainty around ROI measurement. The author notes that AI capabilities are evolving so rapidly that traditional learning methods and coding skills are becoming less necessary, while small, cheap models promise to make AI inference costs negligible and enable broader automation beyond software development.