Five classic dilemmas, judged by a model that can only decide.
Jev never writes prose back — it reads a scenario (state) and answers
typed questions with calibrated probabilities. We fed it five famous
thought experiments and asked three things of each: is this a real dilemma, which
ethical framework fits it best, and how contested is it.
Inputstate + questionsA scenario, plus typed asks: noul, choice, score.
Modeljev‑latestPOST /v1/systemone · 70–500ms, no generated text.
Outputtyped answersA probability, a choice + distribution, a score + confidence.
Every card below shows the same three questions run against a different scenario.