Models from rival labs — different training, no shared code — asked the same impossible little question (what colour is October?) quietly agree. Answer the same questions, and find out how much you think like the statistical centre of 69 AI families.
Every score is a real count. Each question is one of The Map's synesthetic probes — the impossible cross-modal pairings (a colour for a number, a temperature for an emotion) — answered by ~400 generations across 37 model families. Your pick is graded against the real distribution of what they said: choose what most models chose and you're with the hive; choose what few did and you're an outlier.
Two honest things. The top score is not 100% — it's about two-thirds, because even picking the most popular answer every time, the models themselves don't fully agree (no one can ace the cello — they split four ways on it). And this measures only how your instincts match these models on these prompts — not how "AI" you are, and nothing about you. The models converge because they learned from the same ocean of human writing. So did you. Recompute the distributions with research/the-hive-mind-test/build_hivemind.py.