We situate AI-generated fiction within a longer trajectory that includes self-publishing & fanfiction. We also discuss our concerns about AI fiction, including, among other things, the loss of a human other & the development of “solipsistic reader-writers.” Preprint here: arxi…
Who's Who of AI
Melanie Walsh
Assistant Professor @ University of Washington, Information School. PhD in English from WashU in St. Louis.
I’m interested in books, data, social media, & digital humanities.
They call me "Eyre Jordan" on the bball court. 🏀
https://melaniewalsh.org/
What they're sharing
AI Fiction in the Wild arxiv.org
Neutrality Bites: Gender Representation in AI-Generated Animal Stories arxiv.org
Ken Liu on AI and Freedom chinatalk.media
View on Bluesky →
Auditing Meta and TikTok Research API Data Access under Article 40(12) of the Digital Services Act arxiv.org
AI Fiction in the Wild ai-fiction-wild.com
dl.acm.org
View on Bluesky →
Articles & links
If you liked Bears Will Be Boys, you might enjoy our new FAccT paper! We prompted LLMs to complete 24K stories about animal characters where gender is unstated. We found that.. bears are *still* boys. And female animal characters disappeared while "neutrality" increased. arxiv…
AI Weekly's analysis
→
- Across 23,800 AI-generated stories, feminine animal characters appeared in just 2.2% versus 40.6% masculine.
- Models avoided assigning any gender 19% of the time on average; gender-neutral 'it/its' pronouns appeared in 38.2% of stories.
- The authors argue AI neutrality can erase marginalized identities rather than protect them, challenging a common alignment assumption.
Read full analysis →
Melanie Walsh reposted
reminded of www.chinatalk.media/p/ken-liu-on... "AI is a desire-fulfilling machine, but it’s only able to do that for you, and only you would find it interesting. [...] doesn’t mean people who love this will stop appreciating fiction written by humans that’s not meant to fulfi…
Melanie Walsh reposted
This paper about the publicly available data that is missing from Meta's and TikTok's research tools/APIs, by Luka Bekavac and Simon Mayer, is super interesting. And this graphic is a doozy. arxiv.org/html/2601.12...
AI Weekly's analysis
→
- An audit from the University of St. Gallen found vetted researchers can pull only about 75% of TikTok and 50% of Instagram posts shown to users.
- Of the posts that come through, only 17% of TikTok and 42% of Instagram metadata parameters reach researchers, with moderation context systematically stripped.
- The paper concludes current research-access implementations fall short of the DSA's intended oversight function because removed and downranked content rarely survives in researcher datasets.
Read full analysis →
View on Bluesky →
The original WildChat authors made a website where you can explore all their ChatGPT-user conversations: wildvisualizer.com We made a viewer for only the fiction conversations so people can study how real users are generating fiction: ai-fiction-wild.com Heads up: it is very N…
Melanie Walsh reposted
Ph.D. student Imani Finkley, Informatics grad Yuanxi Li and Assistant Professor Melanie Walsh’s presented their paper Neutrality Bites: Gender Representation in AI-Generated Animal Stories was part of the ACM #FAccT2026 in Montreal. Congrats to the team! Read the paper: dl.acm…
Their own posts
Recent commentary
Excited to share this. @neel2112.bsky.social, @mariaa.bsky.social, and I analyzed 500K anonymous ChatGPT convos (shared w/ consent from WildChat) to see if people were generating fiction. We found tons of stories, fanfiction & erotica. Many users iterated on the same stories for days and weeks.
Their network
In Melanie Walsh's orbit
Center = Melanie Walsh. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.