Comment by luxshan_t

In research, we can use AI efficiently to identify gaps and then validate them, but the final idea should be our own. I’ve faced an issue where, when AI suggests something unfamiliar, the research may not become successful because we start relying on it too much. Instead of actually learning and doing research, we become good prompters, and our brains start thinking, "How should I ask Claude to do this?" rather than thinking deeply about the research problem itself.