AI Engineering the Acceleration Whiplash
11 hours ago
- #AI adoption
- #software development
- #productivity impact
- AI has become the primary author of code in most organizations, with high adoption and acceptance rates.
- Business value is evident through increased developer productivity metrics such as epics completed, task throughput, and PR merge rates.
- Throughput gains come with an asterisk, as code churn has increased by 861%, indicating more code rework or refactoring.
- Production incidents have more than tripled relative to code changes merged, indicating a reliability problem.
- Bugs per developer have risen to 54%, showing a worsening defect rate as AI adoption deepens.
- AI has made starting work easy but finishing difficult, with increased context switching, work restarts, and stalled tasks.
- Senior engineers face increased review burdens, with review times soaring due to AI-generated code that looks correct but has hidden flaws.
- More code is entering production without any review, up 31.3%, likely due to overwhelmed reviewers.
- Strong engineering foundations do not protect against AI's downsides; high-performing teams experience similar deterioration in quality and reliability.
- Organizations should reconsider engineering headcount cuts, as AI increases the work needed to ensure code safety, correctness, and maintainability.