Jev Meets the Trolley Problem (Zach Landes)
External essay by Zach Landes (2026-10-14): put TypeSafe Jev and other decision models on the classic trolley dilemma, then read calibrated Choice answers — not chat.
What this guide points to
Zach Landes published Jev Meets the Trolley Problem (2026-10-14). The piece explains decision models (typed answers with probabilities, not prose), restates the classic lever dilemma, and compares TypeSafe Jev with OpenAI Decisions and a set of local decision models on the same question.
Read it as a stress-test field note: how to phrase a moral scenario as a closed Choice, how to compare models that refuse to write essays, and what calibrated probabilities do and do not mean for ethics demos.
This page is an unofficial field manual pointer — verify claims and model names against the article and current TypeSafe docs.
Takeaways
- Decision models answer closed questions; they do not replace ethics frameworks.
- Keep the same state and labels when you compare vendors.
- Treat probabilities as routing signals, not moral authority.