That hasn't been my experience
I use chat GPT iteratively. If I ask it a question, say, about Spring framework, it will almost always come up with the right answer right away. And I'll know it, because it will be an answer I've likely used before and simply forgotten.
Moreover, the answer it gives will be more consistent with the underlying philosophy of the framework itself as well as taking into account its changes over time. Again, it's a subject I know quite well from years and years of experience. It saves me time, and it catches me up with the latest.
Obviously I then check it by checking formal documentation. But Google, while it will take you to the formal documentation, doesn't give you the in situ, use case specific examples that chat GPT does. And Google will also take you to other sites which other people trust and contribute to. But the trust is based on a voting system at best and the contributions are free. So, you end up with things that merely work, but are not well integrated and consistent with the underlying philosophy of the framework. The Google-discovered responses will also frequently have a flavor of human self-interest where the response will try to lead you toward or away from a solution favored by the author, either for commercial reasons or clout
As I said, I don't really think it saves that much time to use chat GPT for getting the coding right, because developers don't spend most of their time coding.