Imagine you are standing beside a railway track. A runaway trolley is hurtling toward five people who are tied to the rails. They cannot move. They will die.
But you are standing next to a lever. If you pull it, the trolley will be diverted onto a side track — where only one person is tied. That person will die instead.
Do you pull the lever?
This is the Trolley Problem, first articulated by philosopher Philippa Foot in 1967 and later refined by Judith Jarvis Thomson. It is the most discussed thought experiment in the history of moral philosophy, and it has never felt more relevant than it does today — when we are building machines that will have to make exactly these kinds of decisions at highway speed.
The Basic Dilemma
Most people, when asked in surveys, say they would pull the lever. Five lives saved at the cost of one feels like straightforward arithmetic. This is the utilitarian position: the right action is the one that produces the greatest good for the greatest number.
But now consider a variation. You are standing on a bridge above the same track. The same trolley is heading toward the same five people. This time, there is no lever — but there is a large man standing next to you on the bridge. If you push him off, his body will stop the trolley. He will die. The five will live.
Do you push him?
Most people say no — even though the arithmetic is identical. Five lives versus one. The outcome is the same. Yet something feels profoundly different about pushing a person with your own hands versus pulling a lever.
This asymmetry is the heart of the Trolley Problem. It reveals that our moral intuitions are not simply utilitarian calculators. We care not just about outcomes, but about the means by which we achieve them.
The Philosophical Stakes
The Trolley Problem sits at the intersection of two great traditions in moral philosophy.
Consequentialism — most famously articulated by Jeremy Bentham and John Stuart Mill — holds that the rightness of an action depends entirely on its consequences. Pull the lever. Push the man. Save the five. The math is the math.
Deontology — associated above all with Immanuel Kant — holds that some actions are intrinsically right or wrong, regardless of their consequences. You must not use a person merely as a means to an end. Pushing the man off the bridge treats him as a tool, not a person. That is wrong, full stop, regardless of how many lives it saves.
The Trolley Problem does not resolve this debate. It sharpens it. It forces us to confront the fact that both positions capture something real about our moral experience — and that they sometimes point in opposite directions.
Why the Footbridge Feels Different
Judith Jarvis Thomson, who developed the footbridge variant in 1985, argued that the key distinction is between doing harm and allowing harm. Pulling a lever redirects an existing threat; pushing a man creates a new one. You are not killing him with the trolley — you are killing him with your hands.
Psychologists Joshua Greene and Jonathan Haidt have offered a different explanation. Brain imaging studies suggest that personal violence — pushing someone with your hands — activates emotional regions of the brain in a way that pulling a lever does not. Our revulsion at the footbridge scenario may be less a moral judgment than an evolved emotional response to direct physical harm.
This is an uncomfortable finding. It suggests that our most deeply felt moral intuitions may be artefacts of evolution rather than reliable guides to ethical truth.
The Doctrine of Double Effect
Catholic moral theology offers a third framework: the Doctrine of Double Effect. An action that causes harm is permissible if the harm is a foreseen but unintended side effect of achieving a good outcome — not the means by which the good outcome is achieved.
In the lever case, the one person's death is a side effect of diverting the trolley. In the footbridge case, the man's death is the mechanism by which the trolley is stopped. His body is the instrument. That, the doctrine holds, is what makes it impermissible.
This distinction — between harm as side effect and harm as means — runs through debates about everything from wartime bombing to end-of-life medical care.
The Trolley Problem in the Age of AI
For most of its history, the Trolley Problem was a classroom exercise. Then we started building self-driving cars.
An autonomous vehicle travelling at sixty miles per hour will, in rare circumstances, face genuine trolley-style dilemmas: swerve left and hit one pedestrian, or continue straight and hit five. Unlike a human driver, the car's response will be the result of a deliberate design choice made in advance by engineers and ethicists.
The MIT Media Lab's Moral Machine project collected over forty million responses from people in 233 countries, asking them to make trolley-style choices for autonomous vehicles. The results revealed striking cultural variation: Western countries tended to prioritise saving more lives; Eastern countries placed greater weight on sparing the elderly; some cultures strongly preferred protecting pedestrians over passengers.
There is no universal answer. And yet someone has to programme the car.
What the Trolley Problem Teaches Us About Life
I have been fascinated by thought experiments since long before I started writing fiction about them. What draws me to the Trolley Problem — and to the Chinese Room, and to the Ship of Theseus — is not that they have answers. It is that they do not.
They are mirrors. They show us what we actually believe when we are forced to choose, when the comfortable abstractions fall away and we are left with a lever in our hand and five seconds to decide.
The Trolley Problem teaches us that we are not consistent moral reasoners. We are creatures who feel our way through ethical life, guided by intuitions that evolved for a world very different from the one we now inhabit. Recognising that is not a counsel of despair. It is the beginning of wisdom.
It also teaches us that the questions we ask matter as much as the answers we reach. Philippa Foot did not solve the problem of moral philosophy when she invented the trolley. She gave us a better way of seeing it.
The Paradox Series Connection
The Trolley Problem is the philosophical engine of Book 2 of my Paradox Series. If you have not yet read it, you can find it on Amazon — and if you are new to the series, Book 1, The Chinese Room, is the place to start.
Book 3, The Ship of Theseus, is available for pre-order now and releases on April 1, 2026. It takes on the oldest identity paradox in Western philosophy — and, I hope, makes it feel as urgent as a trolley bearing down at sixty miles per hour.
Frequently Asked Questions
Who invented the Trolley Problem?
The Trolley Problem was first formulated by British philosopher Philippa Foot in her 1967 paper "The Problem of Abortion and the Doctrine of Double Effect." The footbridge variant was introduced by Judith Jarvis Thomson in 1985.
What is the correct answer to the Trolley Problem?
There is no universally agreed correct answer. The problem is designed to expose the tension between consequentialist and deontological moral frameworks, not to resolve it. Most people say they would pull the lever but not push the man — and the philosophical debate centres on whether that asymmetry is morally justified.
What does the Trolley Problem have to do with AI?
Self-driving vehicles and other autonomous systems may face genuine trolley-style dilemmas in emergency situations. How those systems are programmed to respond is one of the central ethical challenges in AI development today.
Is the Trolley Problem realistic?
The scenario is deliberately artificial — that is the point. Thought experiments strip away real-world complexity to isolate a single moral variable. The Trolley Problem is not meant to describe a likely situation; it is meant to reveal how we reason about harm, agency, and moral responsibility.
How does the Trolley Problem relate to the Doctrine of Double Effect?
The Doctrine of Double Effect, developed in Catholic moral theology, holds that it is permissible to cause harm as a foreseen but unintended side effect of a good action, but not permissible to cause harm as the means of achieving a good outcome. This distinction explains why many people feel comfortable pulling the lever (harm as side effect) but not pushing the man (harm as means).
