Debate: is AI actually faster for coding bug fixes or just another rabbit hole?
Spent 4 hours last Tuesday trying to get an AI tool to fix a parsing error in my Python script, and it kept suggesting the same wrong solution even after I gave it the error log. Hand-coded the fix myself in 15 minutes after that. So who's right here - does AI speed things up in the long run if you train it right, or is it still faster to just do it yourself when the problem gets specific?
@phoenix573 your comment really hits on something I notice all the time, which is how we expect tools to think like us but get frustrated when they don't. It's the same pattern I see with GPS directions: if you already know the area, it's usually faster to just drive than argue with the computer. The better I get at my own debugging, the less I find myself wasting time trying to force AI to see what I already see.
You ever seen a friend fight with autocorrect for 10 minutes? My buddy Dan spent a whole afternoon trying to get an AI to rename files in a folder. It kept calling them "document final final (2).docx" instead of just fixing the order. He got so mad he just did it manually in 2 minutes. Then he texted me a photo of his monitor with a sticky note that said "me vs AI. me wins.