Day Two DevOps D2DO290: AI’s Impact on Developer Productivity Vs. Development Productivity
Dec 17, 2025
Rachel Stephens, Research Director at RedMonk who studies developer-led tech trends, joins to dissect AI’s real effects. She contrasts individual developer productivity with overall development throughput. Conversations cover where AI speeds create bottlenecks, risks of long-running agents, shifts in education and verification, and how to adapt processes to keep humans central.
AI Snips
Chapters
Books
Transcript
Episode notes
AI Boosts Developers But Not Organization Throughput
- AI improves individual developer throughput but writing code has never been the main bottleneck to shipping features.
- DORA research showed AI boosted developer productivity yet initially reduced organizational throughput because downstream processes (tests, CI, release) were overwhelmed.
Code Glut Outpaces Release Capacity
- Generating more code with AI can create a 'glut' that existing pipelines can't digest, causing test/build/operational slowdowns.
- Rachel uses the boa constrictor metaphor: you can swallow lots of code but digestion (release) is still slow.
The Role Shift From Coding To Code Reading
- AI shifts developer work from writing code to reading and understanding AI-produced code, changing skill emphasis.
- Rachel notes vendors previously wanted to return devs to writing code, but AI now writes code and devs become code readers.
