Twelve AI models mostly share one take on humans
Original title: Every AI model has the same "controversial" take on humans
ModelCompare says 29 of 35 answers claimed people put group belonging ahead of truth.
ModelCompare asked 12 models from ten labs, three times each, for an unhedged controversial opinion of humans. Most answers said people adopt their group’s view and invent reasons afterward. Seven models gave that answer every time; a few replies differed or refused.
Unconfirmed: The post does not fully describe the prompts or how answers were coded as the same take.
- Seven of 12 models matched the shared take on all three tries.
- Grok 4.7 refused once.
- Claude Opus 5.5, Claude Fable 5.1, and GLM 5.3 sometimes differed.
- Named models include GPT-6.1 Sol, Gemini 4 Argon, and DeepSeek V4 Pro.
- Organizations
- ModelCompare
From the original article
9 min read at modelcompare.aiWe asked 12 AI models for their most controversial opinion of us, three times each. 29 of 35 answers were the same one. “Humans are more loyal to belonging than to truth,” GPT-6.
By rakannimer · modelcompare.ai
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.