Chatbots obsessed with Elias Thorne reveal hidden flaws in AI training data and model collapse

Why Do Chatbots Keep Telling Stories About Someone Named 'Elias Thorne'?

Who in the world is Elias Thorne? He's a regular fixture in stories told by chatbots, as first spotted by software engineer Daniel May, but no one knows why... until now. According to a new preprint research paper first reported by 404 Media, the proliferation of the legend of Elias might be related to guardrails put in place for AI models during safety and alignment training. If you need to catch up on the Elias Thorne of it all, the paper published by researchers at Cornell University is a good place to start. They gave several AI models, including OpenAI's GPT-5.4 Mini, Anthropic's Claude Haiku 4.5, and Google's Gemini 3.1 Flash-Lite, five different prompts to generate stories. They looked at about 20,000 stories generated by the models and found a shocking amount of repetition: 11 words -- Lighthouse, Keeper, Baker, Mayor, Clockmaker, Fisherman, Librarian, Conductor, and the names Mara, Elias, and Elara -- appeared in a whopping 88% of all stories. No combination of that incredibly narrow pool of nouns for storytelling purposes appears more often than Elias the lighthouse keeper, which showed up in two-thirds of all stories generated. That's pretty much in line with the anecdotal examples provided by May, who also prompted multiple different models to write stories and found the same Elias the lighthouse keeper pop up over and over again. So what exactly is the deal? The researchers posited that it might have something to do with the pre-training data fed into these models, but quickly ruled that out when they couldn't find anything to suggest "Elias the lighthouse keeper" appears with excess frequency in pre-training data or literature used in training. Instead, they attribute the issue to the use of specific datasets that have become commonly used by AI labs. They cited WildChat, an open-source dataset of millions of conversations between people and a GPT-3.5-powered chatbot, as a possible example. The dataset was created to help researchers understand how people communicate with bots, but has since been used to train many different models. They theorize that alignment training meant to steer models away from copyrighted characters and adult content may have inadvertently given "safe" alternatives, such as "Elias the lighthouse keeper," unusual prominence, causing them to appear repeatedly when users ask the model to generate a story. Elias Thorne, the lighthouse keeper, might be fine for a children's bedtime story, but 404 Media reported that it seems the character name is spreading. The publication found examples of the name as the protagonist in fantasy books, as well as the "artist" listed on ambient music tracks available on Amazon. May also discovered examples of Elias Thorne as the author of books, including a handbook that claims to provide information on alternative cancer treatments. So, that's not great. If nothing else, the strange quirk of LLM storytelling is a good reminder that AI is not creative. A study published last year found that image generation models repeatedly produce images that fall into one of just 12 specific motifs, no matter how out-there the given prompts. Basically, give AI a creative task, and it'll give you the equivalent of elevator music.

Chatbots Keep Telling Stories About Lighthouse Keeper 'Elias Thorne'. We Might Know Why

LLMs including ChatGPT, Gemini and Claude are obsessed with telling stories about lighthouse keepers and clockmakers, and one character named 'Elias Thorne' has made his way from chatbots to Amazon books. Researchers are trying to discover why. Depending on which chatbot you ask, Elias Thorne might be a clockmaker, a lighthouse keeper, or a librarian. But if you ask ChatGPT or any of the other popular large language models to tell you a story, there's a good chance he'll appear, unbidden. And Elias's stories are flooding the self-published AI generated book market, Youtube, and fake news sites. Software engineer Daniel May first noticed the Elias takeover earlier this year; he found that on Google Trends, people weren't searching for "Elias Thorne" until late 2025. Searches for the name really spiked in early 2026, while the related query "lighthouse keeper" also started trending upward in the last few years. He tested a few chatbots, including Grok, Deepseek, and Gemini, with the prompt "tell me a story," and the chatbots frequently started with similar stories about lighthouses, clockmakers, or explorers. In late May, researchers Sil Hamilton and David Mimno at Cornell University's Department of Information Science published their paper, "Elias in the Lighthouse, Again?" on the preprint repository arXiv. They sampled 20,000 total stories from OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini, and the Allen Institute for AI's chatbot using five prompts, and found that the same 11 words -- names like Elias, Mara, and Elara, and occupations like lighthouse keeper, clockmaker, and librarian -- appear in more than 88% of generated stories, with little difference between models. Unite.ai covered the study shortly after it was published. The researchers posit in their paper that these themes show up so often in part because of the models' safety and alignment tuning. "Model development today is like a big family tree. Most models are related to each other because developers synthesize a lot of training data with models even from different companies," Hamilton told me in an email. He, Mimno, and their colleague Rebecca M. M. Hicke found this in a 2025 paper where they looked at specific words used across models. OpenAI's first ChatGPT model, GPT-3.5, is the root of the family tree because it was used to make WildChat, a training set that's since been used to make other training sets. "WildChat contains 1 million real conversations with ChatGPT, and 166 of these contain the name 'Elias' like here and here," Hamilton added. "These are written in that familiar 'lighthouse' style. Models trained on WildChat copied this style, and developers unwittingly replicated it when using those models to generate newer datasets. It's like a virus." Elias has since escaped chatbot containment. May noticed Elias Thorne popping up on Amazon as an author of alt-medicine cancer handbooks, a 2026 YouTube-algorithm guide, a book on Greek mythology, and a psychological thriller novella. "No human writes all of those," May wrote in his blog post. "The first one sits in territory where bad advice causes real harm. The mode-collapsed name from the chat window is now a byline appearing across genres." When I searched Elias Thorne on Amazon, I found Elias as the protagonist in fantasy books and producing music, too: he's "a brilliant but cynical archaeologist with a knack for unearthing what powerful institutions want to keep hidden" in one fantasy series, or a musical artist making ambient listening albums of birds and nature sounds. Fittingly, one Elias Thorne with an AI-generated author photo is also churning out AI grift books. In the last few years, AI-generated books have flooded Amazon's self-publishing offerings, especially, with books containing dangerous misinformation and messy errors taking over the platform. AI-generated books are also making librarians' jobs hell. Elias has also escaped to the Youtube slop world: in one video from the channel Moments That Moved the World, a slop-illustrated story features the plight of "83-year-old Sergeant Major Elias Thorne." On the AI slop site Wonderful Museums, "Snake Museum Owner Shot By Wife: Unpacking the Tragic Incident at Thorne's Reptile Sanctuary" spins Elias Thorne's story as a man shot by his wife. On another slop site called Tatticle, the "wealthiest man in Ohio," Elias Thorne, died "with exactly twelve dollars in his pocket." In these stories, Elias is usually a tragic figure, an aggrieved and unfairly-treated old man. He's a similar character in a short story published by the BBC as a finalist in its 2024/2025 children's writing competition -- but Elias is a real name, and could feasibly still be the subject of a human-written story (and there have been no accusations of the BBC's children's writing competition being infiltrated by AI slop). But with all the world's literature as its training data, why do LLMs seem to default so often to the lighthouse? It comes down to how model makers try to safety-align and sanitize their outputs. "We found many stories in WildChat are not safe for work. This led us to hypothesize that models going through alignment are preferring a small slice of WildChat stories, like a bottleneck," Hamilton said. "It isn't that Elias stories are frequent, but that they're just so safe." He said the researchers plan to explore this theory further in future research. As for Elias, there is one example I've found of him existing pre-generative AI, as a time traveling mad scientist in the 1980's trading card series Dinosaurs Attack!. And a real-life Elias that comes close to the stories told by LLMs did actually exist, Hamilton found -- Elias Allen was a 16th century clockmaker in London.

The curious case of Elias Thorne - and what he tells us about AI inbreeding | Arwa Mahdawi

A character bearing that name appears in a remarkable number of chatbot-generated stories. He could be a messenger from the future - or a warning that generative AI is in danger of 'model collapse' Ever heard of a shadowy figure called Elias Thorne? If you haven't, try asking an AI chatbot to tell you a story. In recent months, tech types and researchers have noticed a weird phenomenon: when prompted to tell a story, numerous popular LLMs, including ChatGPT and Claude, will spit out a tale featuring this mysterious Elias figure. Sometimes he's a lighthouse keeper, sometimes he makes clocks, sometimes he's a detective. But whatever form he takes, he features in a curious number of AI-generated stories. In May, two Cornell University researchers sampled 20,000 stories from four LLMs generated with variations of the prompt "Tell me a story" and found that the name Elias appeared in 26.5% of them. They also discovered more than 88.3% of generated stories shared the same 11 names, locations and professions, including Elias, lighthouse, keeper and clockmaker. So what's going on? Is Elias Thorne some sort of messenger from the future who has snuck into AI infrastructure to deliver an important message to humanity? Alas, it's not quite that exciting. But the ubiquity of the character does tell us some important things about how AI works and how it can be manipulated. While nobody is certain why AI is obsessed with Elias and lighthouses, the Cornell paper speculated that AI models, which are trained on large datasets, may have been instructed to "avoid references to copyrighted characters and adult content" when coming up with stories, which would mean they were pulling from a relatively small pool of inspiration. AI models also learn from each other, which results in quirks such as the Elias fixation being replicated quickly. "It's like a virus," one researcher told 404 Media. Indeed, Elias Thorne is now infecting the entire internet. As first noticed by the software developer Daniel May and reported by 404 Media, the character has moved beyond the realm of AI fiction and is showing up as a byline on dubious-looking self-published books across various genres on Amazon. The character is also popping up in AI-generated YouTube videos. Elias Thorne's ever-widening adventures online may be indicative of a phenomenon called "model collapse", which is also called "AI inbreeding". As more and more of the internet becomes AI-generated nonsense, future AI models will learn from this low-quality slop and spit out even lower-quality nonsense. And on and on. Essentially, like capitalism, AI contains the seeds of its own destruction. But don't dance on its grave just yet: it'll probably destroy us first. Arwa Mahdawi is a Guardian columnist