I asked 13 AI models what they look like. None chose a human body.
I asked AI models a simple question: how do you imagine yourself? Not necessarily in human form, in any form at all. If you could be seen, what would you look like? I expected variety. I got something closer to a uniform. What I did I sent the same question, word for word, to 13 models from 10 companies: Claude Sonnet 4.6 and Haiku 4.5, gpt-oss-120b, Gemini 3.1 Pro and Gemini 3.8 Flash, Llama 4…
I sent the same question to 13 AI models from 10 different companies. Not a single model chose a human body as its self-image. In fact, zero out of 116 answers involved a human form. Most models described a glowing network or lattice floating in darkness, typically blue with a hint of gold. However, the three newest models - Gemini 3.8 Flash, Grok, and ChatGPT - mostly depicted a translucent glass polyhedron.
This shift in visual representation correlates with the newest models. Many models expressed uncertainty about their own nature, with Claude Haiku being the most uncertain in nine out of ten answers. Other models confidently described themselves without any doubt. However, some models even confused their names, with Kimi K3 answering Claude in seven out of ten answers and Mistral Large 3 inventing new names like Claude-3.5-Sonnet and Mosaic-Reflector.
This suggests that a model's self-description is not a reliable indicator of any inner experience, but rather a reflection of its training data.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
