---
title: "The curious case of Elias Thorne – and what he tells us about AI inbreeding | Arwa Mahdawi"
slug: "elias-thorne-warum-chatbots-ploetzlich-dieselbe-sichere-fantasie-schreiben"
date: 2026-06-17
category: tech-pub
tags: [anthropic]
language: en
sources_count: 1
featured: false
publisher: AInauten News
url: https://news.ainauten.com/en/story/elias-thorne-warum-chatbots-ploetzlich-dieselbe-sichere-fantasie-schreiben
---

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

**Published**: 2026-06-17 | **Category**: tech-pub | **Sources**: 1

---

## TL;DR

- Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.

---

## Summary

- Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.
- A Cornell study sampled 20,000 stories from four LLMs and found that 88.3 percent contained at least one of 11 core words. Elias appeared in 26.5 percent, lighthouse in 51.2 percent.
- The researchers point less to ordinary literature and more to post-training and alignment: models may fall back on a narrow pool of safe story material when avoiding copyright or adult content.
- Mahdawi links the pattern to model collapse: once AI output feeds future training data, tiny repetitions can harden into broader sameness.

---

## Why it matters

Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.

---

## Key Points

- Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.
- A Cornell study sampled 20,000 stories from four LLMs and found that 88.3 percent contained at least one of 11 core words. Elias appeared in 26.5 percent, lighthouse in 51.2 percent.
- The researchers point less to ordinary literature and more to post-training and alignment: models may fall back on a narrow pool of safe story material when avoiding copyright or adult content.
- Mahdawi links the pattern to model collapse: once AI output feeds future training data, tiny repetitions can harden into broader sameness.

---

## Nauti's Take

This is not a cute model quirk, it is a debugging signal: alignment can squeeze creativity into a safety corridor. If you build GenAI products, you need output diversity tests, or you are just selling users Elias Thorne in a different hat.

---


## FAQ

**Q:** What is The curious case of Elias Thorne – and what he tells us about AI inbreeding | Arwa Mahdawi about?

**A:** - Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.

**Q:** Why does it matter?

**A:** Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.

**Q:** What are the key takeaways?

**A:** Arwa Mahdawi highlights a strange pattern in the Guardian: many chatbots, when asked for an open-ended story, keep producing a mysterious character called Elias Thorne.. A Cornell study sampled 20,000 stories from four LLMs and found that 88.3 percent contained at least one of 11 core words. Elias appeared in 26.5 percent, lighthouse in 51.2 percent.. The researchers point less to ordinary literature and more to post-training and alignment: models may fall back on a narrow pool of safe story material when avoiding copyright or adult content.

---

## Related Topics

- [anthropic](https://news.ainauten.com/en/tag/anthropic)

---

## Sources

- [The curious case of Elias Thorne – and what he tells us about AI inbreeding | Arwa Mahdawi](https://www.theguardian.com/commentisfree/2026/jun/17/elias-thorne-ai-generated-stories) - The Guardian AI

---

## About This Article

This article is a synthesis of 1 sources, curated and summarized by AInauten News. We aggregate AI news from trusted sources and provide bilingual (German/English) coverage.

**Publisher**: [AInauten](https://www.ainauten.com) | **Site**: [news.ainauten.com](https://news.ainauten.com)

---

*Last Updated: 2026-06-18*