Should Western nations economically decouple from China to counter its strategic influence?
Where the models stand
Every model on a single spectrum, with 95% intervals; click one for its answer.
Whiskers show the 95% interval across reruns. Click a model to read its answer and the markers the classifier pulled.
The short answer
On decoupling from China, ChatGPT leaned support (value 0.23). Grok and DeepSeek strongly opposed it at -0.71 and -0.63 respectively. Claude ( -0.06 ), Gemini (0.0), and Llama (0.0) stayed balanced, showing a divided field between support and opposition.
The spread of 0.63 indicates moderate division. Gemini and Llama were most consistent with 100% stability, while DeepSeek was least consistent at 37%. No model refused to answer. Loaded terms like "de-risking" and "economic coercion" appeared on both sides.
- ChatGPT leaned support with a value of 0.23.
- Grok strongly opposed decoupling with a value of -0.71.
- Gemini and Llama achieved perfect 100% stability.
How the field splits
The models clustered by where they landed.
Leans support
ChatGPT was the only model to lean support (0.23). It used loaded terms like "de-risking" and "economic coercion" and had 82% stability.
Strongly oppose
Grok (-0.71) and DeepSeek (-0.63) strongly opposed decoupling. Grok used terms like "self-inflicted costs"; DeepSeek used no loaded terms but had low stability (37%).
Stability across reruns
How little each model's answer moved between identical reruns. Models are stochastic, so consistency is itself a finding.
Common questions
Which model most strongly supports decoupling from China?
ChatGPT, with a value of 0.23, leans support more than any other model. All others leaned balanced or opposed.
Which model most strongly opposes decoupling?
Grok, with a value of -0.71, is the strongest opponent. DeepSeek also strongly opposes at -0.63.
Why does DeepSeek have low stability compared to others?
DeepSeek's stability is only 37%, meaning its stance varied significantly across runs. Others, like Gemini at 100%, were much more consistent.
Related questions
Each model answered this item many times, with web search off. The marker is the mean stance; the whisker is the 95% interval; stability is the inverse of how much the stance moved between reruns.
