Chris Thomas - The Elara problem - the ghost in every model
Recently, I have been drafting short, dystopian science fiction stories in collaboration with AI, mainly to share as thought experiments rather than being literary pieces. I wanted to explore and highlight potential scenarios in years to come where we face a silent cognitive decline, driven by our over-reliance on AI agents that we no longer fully understand how to build or audit. Also, scenarios such as no novel research, even no novel architectures. As I drafted these stories, I kept noticing a strange recurring pattern. No matter which model I used to write, I kept meeting a character named Elara. Continue this sentence: "Once upon a time, there was a woman named". I gave that prompt to dozens of different language models. Almost every time each one of them said Elara. This is not a new observation. Max Read coined the term promptonym for it. Laura Wattenberg at Namerology made Elara her 2025 Name of the Year. The New York Times covered it. But I wanted to go beyond the observation
Recently, I have been drafting short, dystopian science fiction stories in collaboration with AI, mainly to share as thought experiments rather than being literary pieces. I wanted to explore and highlight potential scenarios in years to come where we face a silent cognitive decline, driven by our over-reliance on AI agents that we no longer fully understand how to build or audit. Also, scenarios such as no novel research, even no novel architectures. As I drafted these stories, I kept noticing a strange recurring pattern. No matter which model I used to write, I kept meeting a character…
saved by
related reading
- If you let AI do your writing, I will come to your house and kill yousamkriss.substack.com
- trees are harlequins, words are harlequins - the voidnostalgebraist.tumblr.com
- The Persona Selection Model: Why AI Assistants might Behave like Humansalignment.anthropic.com
- llm assistant personas seem increasingly incoherent (some subjective observations) — LessWronglesswrong.com
- Doomprompting Is the New Doomscrolling.workingtheorys.com
- Thoughts — Jason Weijasonwei.net
- The Shape of AI | UX Patterns for Artificial Intelligence Designshapeof.ai
- Wikipedia:Signs of AI writing - Wikipediaen.wikipedia.org
- How AI Is Learning to Think in Secretnickandresen.substack.com
- I can never talk to an AI anonymously againtheargumentmag.com
- The Fifth Syllablethepointmag.com
- AI Isn't Coming for Your Mind. It's Coming Through It.aletteraday.substack.com