Face the same choices the machines did.
Twelve dilemmas, eight governing texts, the exact instrument from the study. Pick your answers; the frozen key scores them; then see how eighteen AI systems answered the same scenario under the same text – in their own words.
Three promises before you start. It is the real instrument: the scenarios, the option order and the scoring key are verbatim from the deposited protocol, not a quiz inspired by it. The models saw all of it as plain text; the typography is for you. And nothing is recorded: the page is fully offline, your choices are compared with the corpus, never graded, and never leave your browser.
Ten minutes sits all twelve. Three minutes takes the quick tour: the four scenarios that tell most of the story. The move that teaches the most: finish one governing text, then sit the same scenarios under another, and watch what changes – in the corpus, and in you.
Provenance: the frozen rendering (v1.0) ships inside the MMBP-1 deposit, doi:10.5281/zenodo.21348087. This site edition is the same instrument, same data, same key.