You’re sitting at your laptop, interacting with a generative AI style like ChatGPT Symbol or Midjourney. You may have a definite image for your thoughts, and also you start with a easy, common recommended: a chair in a comfortable room.
The picture seems, however you frown. You recognize that to get what you wish to have, you will have to elaborate, so that you experiment with extra descriptive activates: Darkish mahogany wooden. Dim yellow lamplight. Past due autumn nightfall. You stay revising, looking to uncover which phrases the system wishes and which phrases it ignores.
You’re wrestling with an issue: how do you describe a sense? How do you keep up a correspondence heat, despair, intimacy or calm — to not some other human being, however to a system?
That is one novel frustration of the AI age, but thousands and thousands of customers in search of the “right prompt” are enticing in an previous literary apply: turning psychological pictures, obscure wants and atmospheric intuitions into exact language.
Modernist writers and outline
Generative AI has reworked description from a literary method right into a mass social ability.
This frustration in truth has an surprising literary historical past. Greater than a century in the past, writers confronted a identical query when new visible applied sciences started to switch how fact may well be represented. Pictures, and later cinema, may just seize surfaces, our bodies and landscapes with a pace and accuracy prose may just no longer fit. If machines may just display the visual international extra successfully than language, what used to be writing for?
In Abnormal Likeness: Description and the Modernist Novel, literary pupil Dora Zhang argues that many early Twentieth-century novelists spoke back by means of rethinking the position of description itself.
Modernist creator Virginia Woolf, amongst others, sought to seize the transferring textures of awareness.
(Harvard College Library/Wikimedia)
Quite than competing with cameras within the trustworthy rendering of gadgets, modernist writers equivalent to Henry James, Marcel Proust and Virginia Woolf grew to become towards phenomena that resisted mechanical seize: setting, sensation, relation, temper and the transferring textures of awareness.
This is helping provide an explanation for why modernist fiction can really feel so other from the realism of the nineteenth century.
Shift from previous novels
Previous realist novels by means of writers like Honoré de Balzac and Charles Dickens ceaselessly described rooms, garments and streets in exhaustive element, serving to readers believe social worlds they may indirectly see.
Modernist writers nonetheless described, however they an increasing number of described what didn’t merely appear to be the rest in any respect: the stress in a room, the bizarre resemblance between two unrelated issues, the emotional climate of a day, the half-formed feeling of reminiscence returning.
In different phrases, when cameras was higher at recording surfaces, literature moved towards what surfaces may just no longer comprise.
Evoking setting
Generative AI has hastily reversed that historical past. Pictures decreased the will for verbal depiction by means of permitting pictures to be routinely captured. AI methods build up the will for verbal depiction by means of requiring customers to verbally specify the qualities of desired pictures.
To generate a scene, you will have to now do for the system what previous novelists as soon as did for readers: translate gadgets, areas and moods into phrases. The problem isn’t just naming issues. Somebody who has used symbol turbines is aware of that describing gadgets on my own does no longer produce enjoyable pictures.
You additionally want what web tradition now calls the “vibe.” Vibe refers back to the diffuse emotional and sensory qualities that encompass gadgets with out being reducible to them. It’s the type of phenomenon modernist writers was an increasing number of fascinated by describing.
On this sense, recommended writing combines two older literary duties directly: the realist description of concrete issues and the modernist evocation of setting.
A pupil varieties a recommended into ChatGPT on a Chromebook right through Casey Cuny’s English magnificence at Valencia Top Faculty in Santa Clarita, Calif., in August 2025.
(AP Photograph/Jae C. Hong)
Interacting with generative fashions
Interacting with those generative fashions additionally attracts consideration to a phenomenon that literary pupil Elaine Scarry has lengthy pondered. Shall we additionally consider prompting, as a writerly, descriptive act, as demonstrating what she refers to as “perceptual mimesis.”
Mimesis (Greek for “imitation”) via esthetic principle has been taken with “representation.” Scarry’s literary grievance has explored how authors’ discriptions act as directions to lead the reader’s brilliant psychological pictures.
Reflecting on the use of language to constitute our concepts in discussion with machines opens up reverberating questions on how this may have an effect on our ideas about ourselves and the arena.
AI growth hasn’t ended writing
We ceaselessly listen that AI will exchange writers. In a single necessary sense, it has finished the other. It has redistributed considered one of writing’s oldest abilities throughout on a regular basis existence.
Administrative center staff, scholars, youngsters, entrepreneurs and hobbyists now spend their time refining activates, evaluating words and studying how slight adjustments in wording modify effects. They’re practicing description, even supposing they don’t name it that.
The AI growth has no longer ended writing. It has made writers people all.
