I disagree with lots of this. They are making many iffy assumptions about the future AI as if its not going to change, when current AI not even optimized and ar

I disagree with lots of this. They are making many iffy assumptions about the future AI as if its not going to change, when current AI not even optimized and architected for interactive storytelling are already doing a surprisingly good job They are also discounting the value of interactivity. And of course GPT4 or Pygmalion are not going to write a coherent story... just like they aren't going to write coherent code. But they are already decent assistants for filling in blanks. But this is the actual bit that worries me: > Will people watch a certain degree of regurgitated nonsense they’ve seen before? Sure, to a point. Oh, they absolutely will, for a long time. In the short term, I am very worried producers will trade quality for volume even more than they already do. Why gamble on a hit when you can market 200 stinkers for the same cost?