Comment AI fails the detail problem. (Score 1) 131
When using AI to quickly mock up small chunks of code I find it an accelerator. And I do mean small.
But when code scales up simple systems or API's it falls apart pretty quickly. As the scale of the system grows the requirements grow even faster. GDPR, PII, FIRB, NIST, all start to pile up on as code bases grow. AI lacks the understand of the "business" need. So you get this blob of code out and then you have to spend large amounts of time understanding it so you can re-factor it because the AI engine missed the mark by even a small amount.
In addition I find AI code generation lacks the ability to plan for the future. What potential use cases will come up next week, month, year. And this in flexibility means the code base has to be essentially tossed in the bin. It's even worse when you think about the impacts on you dataset. You effectively can't adapt over time. The code effectively is locked because your data set is your business.
Then there is the problem with conforming to standards. AI code simple sucks at this. You'll have 95% conformance and basic testing often comes out totally successful but those 5% edge cases fall apart.
So I find AI useful for doing the drudge work. Even some of the simple logic work. But I rarely let AI work outside of the context of a single function/method. It can be a decent time saver when I keep it simple.