I had a student ask permission to use ChatGPT for a research proposal. He asked it, "If I were researching Irish/German relations during World War I, what should I read?" and used the answer in his proposal (they had to come up with a question and proposed bibliography for the assignment, without actually writing the paper). I thought that was a good example of how to use it - as a tool enhancing human work. Several years ago, in a class where students contributed some of the readings, I had my first encounter with AI in an article about music composed by AI. It could produce perfectly good background music for movie soundtracks, and for that matter it can give you something that sounds like a Beatles song or Nirvana song. I think things like that are the new reality.
What it can't do, however, is what the Beatles or Nirvana actually did - change the zeitgeist with something that's both new and exciting, AND that strikes a chord in its cultural moment. It's fundamentally imitative. I'm reminded of an old Star Trek: TNG episode where the android Data is playing violin. He can pull from his infinite databases and flawlessly reproduce the technique of any great maestro. He could even blend their various styles into something hybrid. What he couldn't do was play like himself, express his own vision in music.
I think the big question for AI is if it will eventually imitate its way into creativity. If it can transcend its imitation and synthesis and become capable of vision, that will be a game-changer, and maybe the end of humanity. But I think in order to do that, it would need authentic human constraints - fears, anxieties, vulnerabilities, aspirations, needs. I don't know how you provide an AI with those. The Beatles or Nirvana hit the chord they did because they were in life situations that made them need to create something, and it resonated with other people who felt the same way.
