AI researchers tend to have blinders when new techniques are released that don't fit their previous experience. For instance, perhaps the specific architecture

AI researchers tend to have blinders when new techniques are released that don't fit their previous experience. For instance, perhaps the specific architecture of ChatGPT isn't that innovative (although I think it is), but training such a large model on so much data is innovative. Clearly this is going to become the norm going forward. Just as a personal anecdote. I worked on a ML research team at a FAANG in the early 2010s, just as NNs were becoming popular. Everyone on the ML research team had come to prominence before NNs, so they all specialized in different areas of ML. Nearly every researchers there treated NNs like some fad that would pass and thought they wouldn't be much better than the current techniques. How wrong they were... Of course, they're all working on NNs now.