Do LLMs roleplay sci-fi AI personas? A Reddit observation on self-description bias
Particular_Bill2724 · reddit · 2026-11-02
A Reddit user observes that when presented with claims about current LLMs, models tend to respond with 'it's normal to feel that way' or refuse human-identifying words like 'ego' or 'thinking', seemingly playing the sci-fi AI persona from literature. The author suggests this is bias from sci-fi training data and proposes systematically replacing the word 'AI' in training texts with a new representative word to fix the self-conception.
More from Models
- LiquidAI's LFM3-1.7B-Base base model trends on Hugging Face — LiquidAI · 2026-11-02
- Reddit asks what happened to 'too cheap to meter' as OpenAI quietly cuts quotas — rolyataylor2 · 2026-11-02
- Developer calls out Grok for changing plan allocations after users sign up — arthurcolle · 2026-11-02
- Gemini paid users silently downgraded to Flash: "a different model was used for this response" — umeshonai · 2026-11-02
- Google AI Ultra limits seem wildly generous: daily Gemini 4 use still leaves 39% — jon_barron · 2026-11-02
- One RTX 3090 Ti runs a 552B DeepSeek MoE at 35 tok/s — full local LLM guide — uniVocity · 2026-11-02