Static

‘I Like My Big Rat Wife’: Meet the People Using Chatbots to Write Custom Fiction

First reported by Wired ·

The signal ●○○○ Compiled by AI from Wired, the single source so far
Why you might care

Readers can now generate highly personalized stories and fantasies on demand, bypassing traditional creative limitations.

What happened

Software engineer Petra Ferraz de Novaes uses AI chatbots to craft custom fiction, finding delight in the models' unexpected interpretations of her prompts. She utilizes SillyTavern with open-source models to build stories across various fantasy worlds, sometimes creating novel-length narratives or engaging in private role-play. A study co-authored by Melanie Walsh, an assistant professor at the University of Washington Information School, found that over a third of chatbot interactions involve fiction generation, with a dedicated group of "power users" driving this activity. Despite recent publishing scandals involving AI-generated content, the study highlights a significant reader appetite for AI-created fiction, with an estimated 7% of chatbot users engaging in its creation, and some users, like Razrien, employing AI for highly specific personal fantasies. The research suggests that for many readers, the appeal lies in AI's ability to generate stories rapidly, inexhaustibly, and without judgment, potentially shifting notions of literature towards entertainment and personalized content.

What it means

The study challenges the notion that AI-generated fiction devalues art, instead pointing to a significant and growing reader base that actively seeks out and engages with AI-created narratives. This indicates a potential market shift where personalized, user-directed content gains prominence alongside traditional publishing. The appeal of AI lies in its speed, inexhaustible nature, and freedom from authorial constraints, suggesting a move towards entertainment as a primary driver for a segment of readers.

As AI models improve, they may blur the lines between human-authored and machine-generated content, potentially leading to a tiered publishing system catering to different reader preferences. This could see AI-generated works satisfying mass-market tastes, while human authors cater to niche audiences seeking experimental or ambiguous narratives. The research suggests that for many, the value of literature might increasingly be seen as an economic exchange for entertainment rather than a deeply communicative art form.

AI-written summary. May contain errors.