Michael Crichton:
You're a brilliant scientist who's just created something that will revolutionize the world. Congratulations! It's now trying to eat you.
Michael Crichton:
You've stumbled upon a conspiracy involving [insert scientific field]. Now you're being chased by [insert government agency] while trying to explain complex scientific concepts to the reader.
Suzanne Collins:
You must choose between two brooding love interests while simultaneously overthrowing a totalitarian regime. Priorities!
Stephen King:
Welcome to small-town Maine, where the biggest threat isn't the weather, it's the [insert supernatural horror]. Don't worry, a writer will save the day.
Neil Gaiman:
Mythology crashes into modern life. You're either a god who's fallen on hard times or a regular person about to have a very weird Wednesday.
Margaret Atwood:
Society has taken a slight turn for the worse. Women are now [insert dystopian scenario]. This is definitely not a commentary on current events.
And perhaps my favorite:
George Orwell:
Big Brother is watching you. So is your toaster. And your pet. Trust no one, especially not the pigs.
Out of 30 generations, there were a few more that made me smile, but these were the main ones I enjoyed. Something I've noticed with statistical content generation is that it has a difficult time not being too "on the nose" -- almost like next-token-prediction is making it want to rush and get to the punchline a little too quickly. It has a hard time being subtle, and too often it felt like it was just a glib little summary of a story, rather than a sardonic take-a-step-back-and-look-at-the-big-picture sort of approach.
No major revelations, but just barely interesting enough to warrant commenting here. If there were a Dull Men's Club version of Hacker News, I would have posted this there.
Even more so, they provide commentary at the bottom on the weaknesses of the model, which is useful!