General Discussion

I have no doubt it can do some sort of “debugging” and will confidently tell you that it did so. I’m questioning the accuracy of it, but, will look more into it. I wonder about bugs driven by user interaction especially - I’m skeptical that AI in its current state could possibly test for those.

AI code may be “bug free” when generated, but, have many bugs when that code is used in a specific context - eg users are feeding data into a function in a different way that the AI code could “test” for. That, or, humans will always find different scenarios to blow your code up.

However, I believe the point is moot because I really doubt AI’s ability to test code to begin with. I’d love to see a breakdown of how effective AI “debugging is” and don’t trust any claim AI companies make about their products.