No, brute force things like that can actually get a lot of erros. Latest reasoning models are very good at understanding the architecture, catching bad patterns, understanding what has been a bad design choice/quick fix to get something delivered. Planning a strategy is also very solid, but like i said, you need to watch out and provide senior guidance in case direction is pivoting off.
Like before, with good questions you get good answers. AI is going to superpower skilled peogrammers who are not in denial a LOT