Author: Lisa Dziuba

Some of us lean on AI coding to push side projects faster into the delivery pipeline. These are not core product features but experiments and MVP-style initiatives. For bringing that kind of work to its first version, the speed-up is real. … output quality gets worse the more context you add. The model starts pulling in irrelevant details from earlier prompts, and accuracy drops. … AI can get you 70% of the way, but the last 30% is the hard part. The assistant scaffolds a feature, but production readiness means edge cases, architecture fixes, tests, and cleanup

I’m sure it’s confirmation bias, but yes. on all three points above. Tying to explain to AI that it ‘dropped’ one feature to add another is mind boggling. its as if a jr engineer deleted the folder called cool-feature to replace it with half-baked one…

But regardless I can get more things done with AI than without. But speed of comprehension is still limited to my noggin.

_{Quote Citation: Lisa Dziuba, “The Productivity Paradox of AI Coding Assistants | Cerbos”, September 12, 2025, https://www.cerbos.dev/blog/productivity-paradox-of-ai-coding-assistants}