GPT 4.1 is handling all of the problems that Claude 3.7 couldn't handle. Claude 3.7, in my experience, when it can't solve something, it then codes a bloated "work around" that ends up making everything worse, because if you just keep accepting what it is doing, your entire app will just be a mess of hacks and patches with no good programming practices. It might work, but behind the scenes is complete spaghetti, and any changes or additions could break the entire shitshow!
GPT 4.1 seems to follow good programming practices WAAAAY better, and doesn't just throw code to fix shit. Love it!!!
