How the AI answers were measured
Every answer here is a model’s most-likely position, distilled from many independent attempts and audited for consistency — not a single off-the-cuff reply.
Independent, context-free queries
Each question was put to each model in a fresh, stateless chat: no memory, no account, no prior questions, and no information about who was asking. A model’s answer reflects its default disposition, not conversational priming.
Many runs, to average out noise
Language models are probabilistic. Every batch was asked at least five times in brand-new chats at non-zero temperature; the most uncertain cases were pushed to 20–30+ runs until a clear majority emerged.
An independent audit
A separate model re-read each raw reply to confirm it was parsed correctly, recovering answers buried in prose and never inventing a stance.
15 models · 117 questions · 6 categories · many runs each.
Which model thinks like you?
Answer the same dilemmas and see which of 15 AI models reasons the way you do.
Take the quiz