Why Different AI Models Generate the Same Name: How Safety Mechanisms Work
Programmer Daniel May noticed that AI bots from different companies regularly put a character named Elias Thorn into the stories they generate. A study has explained the phenomenon: Thorn is an artifact of safety mechanisms built in during neural network training. To avoid real names, models develop stock "safe" characters, and Thorn became that placeholder hero across dozens of systems at once.
AI-processed from 3DNews AI; edited by Hamidun News
A study explains why different AI systems and text-generation neural networks come up with the same fictional name: Elias Thorn.
Why do different neural networks come up with the same name?
According to the findings of a published report, the cause is safety mechanisms: models are prohibited from mentioning the names of real people in fictional contexts, so they need a set of "safe" names. Elias Thorn became the universal placeholder character: the name is neutral, does not match any public figures and sounds "literary" enough.
Why do neural networks come up with identical names?
Because safety mechanisms prohibit mentioning real people in fictional contexts, so neural networks use a set of 'safe' names as placeholders.
Key Questions Answered
Why do different neural networks come up with the same name?
According to the findings of a published report, the cause is safety mechanisms: models are prohibited from mentioning the names of real people in fictional contexts, so they need a set of "safe" names. Elias Thorn became the universal placeholder character: the name is neutral, does not match any public figures and sounds "literary" enough.
Why do neural networks come up with identical names?
Because safety mechanisms prohibit mentioning real people in fictional contexts, so neural networks use a set of 'safe' names as placeholders.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.