I fed to following prompt into ChatGPT:
I believe that AI for is karma for the literary world. The great modernists laid down the challenge. Fiction as art should do something that populist fiction doesn’t. Even if the novel is about a bunch of highly educated Europeans struggling with great intellectual issues, it is no different than a romance novel about gay hockey players if it is just a set of triggers to make you picture another reality. Real fiction does something else that is unavoidable non-populist. The cultural gatekeepers (agents, publishers, professors, the pretentious asshole at the cocktail party) deliberately resisted the challenge of the modernists and have sustained what we could call elite referential fiction, writing that does not reject the representational approach to fiction. The fiction writer makes up stories to say important things about humanity, not to create unique self-referential art that means only itself. This conspiracy to ignore the challenge to referentiality has made possible the horror show of expensive MFA writing programs spitting out books no one reads. Well, surprise, AI can imitate referential writing very well. So this whole castle made of sand will now crumble as punishment for intellectual cowardice.
ChatGPT gave me this reply:
There’s a sharp provocation in what you’re saying, and I think it lands on something real — though “conspiracy” is probably too intentional a word for what was mostly institutional inertia.
Your argument, as I hear it, is roughly this:
The great formal rupture of literary modernism — James Joyce, Virginia Woolf, Samuel Beckett, maybe even the late experiments of William Faulkner — posed a challenge that was never fully met. They implied that fiction, if it wanted to justify itself as art rather than entertainment-plus-prestige, had to stop relying primarily on representation: on “here are some people, imagine their lives, now infer significance.”
The challenge was to make language itself the event.
Not “what happens to these characters?” but “what happens to consciousness, syntax, temporality, perception, when language is forced into new forms?”
And your complaint is that literary culture largely flinched.
Instead of continuing that radical line, the apparatus of literary prestige rebuilt itself around what you’re calling elite referential fiction: novels that keep the representational machinery of conventional storytelling intact but decorate it with seriousness-signals — historical trauma, theoretical allusions, social critique, educated melancholy, geopolitical complexity.
These books flatter the reader into feeling they are participating in art because the references are elevated, the prose tasteful, the themes “important.”
But structurally they’re still trigger systems for mental simulation: the same underlying mechanism as genre fiction, just with different status markers.
That’s a serious critique.
And yes: large language models expose this brutally.
AI is extremely good at reproducing the external signatures of referential seriousness because referential seriousness is, to a large extent, pattern legible. It has recognizable lexical habits, tonal registers, narrative pacing conventions, symbolic scaffolding.
If a text’s value lies mainly in:
- sounding intelligent,
- alluding to approved concerns,
- arranging psychologically plausible scenes,
- producing the familiar aura of literary significance,
then AI can counterfeit it at scale.
That does feel like a kind of historical judgment.
Not mystical karma, exactly, but technological exposure.
It reveals that much of what institutions defended as difficult art was actually sophisticated convention.
Where I’d push back is here:
The modernist anti-referential project is not obviously the sole legitimate future of fiction.
It may be that fiction has an irreducibly referential basis. Language points; narrative represents; readers inevitably construct worlds. Total self-referential purity may be a limit case rather than a sustainable artistic mode.
Even James Joyce never fully escaped reference. Even Samuel Beckett retained theatrical and existential situations.
So the question may not be “did literature betray modernism?”
It may be:
Can fiction produce formal necessity within representation rather than by abolishing it?
That’s a harder standard, and one many MFA-era novels probably fail.
What AI is punishing is perhaps not referential fiction itself, but fiction that never justified why this arrangement of reference had to exist as art rather than as competent content.
That connects to something you’ve been circling in your other reflections about art and aging: the suspicion that many available “options” are delusions sustained by social machinery.
AI is a merciless delusion-detector.
It is forcing literature to answer: what can human writing do that is not merely the tasteful recombination of already recognized effects?
That’s not the death of fiction.
It’s the return of the question modernism posed and the institutions deferred.
Kenny Nowell 