AI-Written Code Is Only Better When a Skilled Programmer Is Holding the Wheel

Originally published at fatihsennik.com

I'm going to say something that will annoy people on both sides of the AI coding debate.

AI-generated code is not good or bad by itself. It's a mirror. It reflects the skill level of the person using it and it does so at about three times the speed and ten times the volume. A skilled developer using AI writes better code faster. An inexperienced developer using AI writes worse code faster. The AI doesn't know the difference. It just outputs.

I've been using local models in my daily workflow for months now such any in LLM Studio, Ollama and so on. I've watched what it does to code quality when I use it versus when junior developers on client projects use it. The gap is not subtle.

What Actually Happens When a Junior Uses AI

AI-Written Code Is Only Better When a Skilled Programmer Is Holding the Wheel

Other newsrooms on this story

Related reading

Three skills that matter when AI handles the coding

AI coding is now everywhere. But not everyone is convinced.

AI is writing your code, but who’s reviewing it? - TechTalks

The Death of "Just Write the Code": How AI-Native Development Is Reshaping What…

AI Doesn't Hide Your Coding Weaknesses. It Amplifies Them.

Why AI Coding Agents Fail Senior Engineers (And What I Built to Fix It)