AI is Killing Traditional Programming! …Not Quite.
AI is great when you’re unsure. Coding wins when you’re not.
Last night I had an interesting discussion with a professional programmer. I don’t really consider myself in that class, even though I do get paid to write code.
We were talking about AI and software development. His opinion was that AI is mostly useless and actually slows things down, because you still have to verify every line of code anyway. That goes against what many studies are saying, but it got me thinking.
Where does AI actually fit in the development world?
For now, there will always be a place for a developer to check and validate what AI produces. But beyond that, I think there’s a simple way to look at it.
AI is the Windows of Programming. Coding is Linux.
Let me explain.
Windows is less precise. It’s often slower for tasks you do regularly, and you don’t have full control over what’s happening under the hood. But it has a gentle learning curve. Almost anyone can use it, and it usually gets you close enough to what you want.
Linux, on the other hand, is precise. You get exactly what you ask for, and you can tweak it until it behaves exactly how you want. But it takes time to learn and even longer to master.
So in practice:
- Linux is faster if you’re an expert
- Windows is faster if you’re not, or if you’re doing something unfamiliar
That’s Exactly the AI Situation
If you’re a professional developer, working daily in your tools and languages, AI can feel slower and more cumbersome. You already know what to do, and verifying AI output just adds overhead.
But if you’re:
- using a new API
- working in a language you haven’t touched in a while
- or just figuring things out
then AI becomes incredibly useful.
It helps you get moving quickly, even if the result isn’t perfect.
So Where Does That Leave Us?
AI isn’t replacing programming. It’s changing where it’s useful.
- If you know exactly what you’re doing → just code
- If you don’t → AI can get you 80% of the way very quickly
And that’s the real shift.
AI doesn’t replace skill — it amplifies the lack of it, and fills in the gaps.
Final Thought
The debate isn’t whether AI is good or bad for programming.
It’s about knowing when to use it.
Because right now:
- AI is faster when you’re uncertain
- Coding is faster when you’re certain
Windows vs Linux 🙂
Just my take.