Absurd Trolley Problems

What is the trolley problem?

A runaway trolley is heading toward five people tied to the track. You are standing next to a lever. Pull it and the trolley switches to a side track where it will kill one person instead. Do nothing and five die. What do you do?

The scenario was set out by the British philosopher Philippa Foot in 1967 and developed considerably by Judith Jarvis Thomson in the years after. It has become the most recognisable thought experiment in moral philosophy, largely because almost everyone has a fast confident answer and almost nobody can defend it consistently once the variants start.

How this version works

1. Read the scenario and pick a track.
2. See what happens, and how your answer compares.
3. The next scenario arrives, and it is stranger than the last.

Fifteen scenarios, escalating from the textbook version into territory the textbook never intended. Your running body count is kept, which is either the joke or the point depending on your mood.

Why the variants matter

Around ninety percent of people say they would pull the lever. Then comes the footbridge version: same five people, same trolley, but the only way to stop it is to push a large stranger off a bridge into its path. The arithmetic is identical, one death instead of five, and now most people refuse.

That gap is the entire reason the problem is famous. If you are a strict consequentialist counting outcomes, the two cases are the same and you should act in both. If you hold that using a person as a means is wrong regardless of outcome, you should refuse in both. Most people do neither, which suggests moral intuition is not running a single consistent rule. Work by Joshua Greene and others using brain imaging found that the personal, up-close version engages emotional processing far more strongly than the impersonal lever version, which is one proposed explanation.

Where it stopped being hypothetical

The problem left the seminar room when self-driving cars arrived, because someone genuinely has to write down what the vehicle does when harm is unavoidable. MIT's Moral Machine collected tens of millions of responses to exactly these dilemmas and found preferences varied significantly by country and culture, which is awkward for anyone hoping for one universal answer to encode.

The honest caveat is that real autonomous vehicles almost never face a clean five-versus-one choice. They face uncertainty, degraded sensors and split seconds, so the trolley problem is better understood as a tool for exposing your own reasoning than as an engineering specification.

If you liked being uncomfortable

The Password Game and I'm Not a Robot escalate an everyday irritation instead of a moral one, and Book of Answers hands the decision to an oracle so you do not have to reason at all. All free, no account, nothing stored.