Also, a quick (and brief) follow up: the games industry has spent the last half-century or so investing a ton of time, money and effort into the "golden goose" of procedurally-generated environments and narratives. If they didn't have to hire artists and designers and writers, games would not just be much cheaper to develop, but easier, too.

And in every case, across the board, procedurally-generated content has proved inferior to hand-crafted content.

Even when dealing with very few variables -- a two-dimensional level of a forest or a dungeon or whatever, where the entire thing is built out of tiles in a grid pattern -- these AI-created spaces have a distinctly artificial quality, are almost universally inferior to deliberately-crafted spaces, and are often seriously compromised (EG impassable terrain, poor player guidance, etc.).

These spaces are conceptually very simple, and only really need to satisfy two needs: to be realistic approximations of the intended environment, and the be interesting spaces for players to navigate.

Literature, meanwhile, is exponentially more complex and needs to serve many, many more goals.

And there's a helluva lot more money in the video games industry than in the printing industry -- around 80 billion USD to around 350 billion USD. If the one industry cannot use AI to accomplish a far simpler goal with much more time and much more money, what are the odds the other will be able to surpass them far less time with far less money?

So you can see, I think, why I am so deeply skeptical of this whole thing, and dismiss it as overblown marketing hype (in which must of the media is, as always, deeply complicit).

And now my meds are starting to wear off, which means I'm either due for a bad time that may precipitate more meds, or a crash. In either case, time to log off for a bit.