FightTheDragons said: "I'm not exactly sure where Ken is getting the idea that audience response was "very positive." I heard nothing but mixed (at best) or very bad (at worst, and more common) things.
The use of AI for this is... so stupid it's unbelievable. Just read the responses yourself and come up with things to work on rather than having ChatGPT vomit out what it thinks you should do (and will likely be wrong about). This is not ethical use of AI in theater, it's stupid use of AI in theater."
Unless the elites have access to a much higher level of AI than the public, I don't think we're at risk of being replaced yet... a few months ago I was feeling depressed about the state of AI in art, so I decided to see how much of a threat it really was. I logged into ChatGPT and said "AI, write me a full-length, two act, two and a half hour musical about [prompt]." I didn't look at what it created, I just yes-and'd as it asked me "do you want an outline? Do you want the script? Do you want the songs with lyrics and chords? Do you want me to turn that into sheet music? Do you want me to arrange it for a full orchestra?"
I let it knock itself out until it seemed to be done, then looked at what it had made. Thank god, it was not only trash, but incomprehensible and dishonest: most of the things it said it had written completely were barely thumbnails, the full-length songs it claimed to have written were sometimes no more than three lines long, and when I opened the "sheet music" files it was just incomprehensible MIDI static.
Granted, I felt weird using the prompt at all, but I was hugely relieved at how inept and lazy its output was.