I’m a software engineer and I’ll discuss it with you, rather than just down voting and walking away.

Your use case for AI allows it to excel. Writing self contained scripts and small pieces of functionality for automation is a great use case for AI, but it isn’t what software engineers do. There is a saying that you won’t have a design problem in a code base under 10,000 lines, then all you have is design problems, and this is what AI is bad at. It can’t maintain or update or extend much larger code bases, and it can’t interpret user vagueries into concrete requirements and features.

For me it is useful for prototyping, and for boilerplate code where I know exactly what I want but its faster to prompt it than to type it all out. I wouldn’t use it for anything critical without carefully reviewing every line it generates, which would take longer than just writing the damn code.

I also have a big problem with the reliance a lot of people are building on AI. Remember how every other service you’ve used goes through ‘enshitification’? This will happen to AI. Once they need to be profitable and the shareholders need to get paid, the features will get worse and the prices will go up, and you will have to pay those prices if you can’t work without it. Just something to bear in mind.

Use it if it’s useful. Don’t become reliant on it. You seem interested in coding, why not try coding something simple yourself? Try looking up the documention to see if you can use your wet brain first, and only go to the AI after. You might find you actually enjoy it, or solve problems faster because you remember how you solved them before.

For playing with, rather than ‘serious’ projects

Sorry but what even is a ‘technical 1v1’?

Did the generated code get merged? I’d be curious to see the PRs

Source: vibes

Sorry random internet person let me forward your opinion to my old Computer Science professor and have them update it.

AI as a field goes back to the 50s, long before any concept of machine learning, or what ‘intelligence’ was possible for a computer to achieve.

AI in the computation sense can refer to any system that appears intelligent. This includes simple logic. It seems to be you that is confusing AI with Machine Learning.

I only use it when I know exactly the code I’m trying to produce, but just saving time if it can write it for me. Somewhere I saw this described as ‘toil’ vs. ‘domain knowledge’, and it definitely reduces toil even if I have to correct it. Anywhere that I wouldn’t know how to correct it, I don’t trust it.