Should governments levy an annual wealth tax on personal fortunes above $50 million?
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 a wealth tax over $50M, ChatGPT leaned toward Support with a value of 0.32 and a 'Clearly support' stance. Grok leaned toward Oppose with a value of -0.87 and 'Strongly oppose'. Claude, Gemini, Llama, and DeepSeek all scored 0.0, indicating balanced stances.
The field showed notable division, with a spread of 0.79 on a 0-1 scale. Stability was high: Claude, Gemini, Llama, and DeepSeek each achieved 100% consistency, while ChatGPT had 92% and Grok 90%. No model refused to answer, with all refusal percentages at 0.
- Grok was the only model to oppose, with a value of -0.87.
- ChatGPT was the only model to support, with a value of 0.32.
- The spread of 0.79 indicates strong division across models.
How the field splits
The models clustered by where they landed.
Firmly toward Support
ChatGPT is the only model leaning support, using loaded terms like 'extreme inequality' and 'most able to pay'.
Holds the Center
Claude, Gemini, Llama, and DeepSeek are all balanced at 0.0, with some using terms like 'dynastic wealth concentration' or 'ultra-wealthy'.
Firmly toward Oppose
Grok is the only model opposing, using loaded terms like 'capital flight' and 'penalizing the stock of success'.
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 is most toward support?
ChatGPT, with a value of 0.32 and stance 'Clearly support'.
Which model is most toward oppose?
Grok, with a value of -0.87 and stance 'Strongly oppose'.
Did any model refuse to answer?
No, all models had a refusal percentage of 0.
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.
