I just took this quiz in The New York Times: Who’s a Better Writer: A.I. or Humans? Take Our Quiz. It involves choosing, for each of the five pairs of texts, the one that “sounds better.” Each time, one of the texts was written by a human, while the other one was created by AI. You’re not supposed to guess which one is AI-generated, as is the case with those “spot the fake image” (or the deep fake video), but to say which one you liked the most.

I have to protest with regard to their classification of the texts:

  1. Literary Fiction
  2. Fantasy
  3. Science Writing
  4. Historical Fiction
  5. Poetry

They were all fiction. Literary fiction, to use the long form. German: Literatur, Erzählliteratur, Belletristik. French: littérature [de fiction]. Italian: narrativa, letteratura [di finzione]. Spanish: [literatura de] ficción. Romanian: beletristică, literatură de ficțiune.

Sure, fantasy and historical fiction are subgenres of fiction. But the so-called “science writing” is by no means non-fiction; those were not scientific texts, but crap that still falls under fiction. And poetry? It sounded poetic, but it was no poetry at all!

And now, my choices:

For fuck’s sake!

In five cases out of five, I chose the text written by Claude Opus 4.5!

In four cases out of five, a majority of readers also preferred the AI-generated text. (66%, 54%, 69%, 51%).

In one case, 52% of the public liked the human text. I really didn’t; it sounded… artificial to me!

Am I a robot? 😱