Thought Experiments As A Design Tool

An ordinary question asks what you think. A thought experiment forces you to think by placing you in a situation where your thinking becomes visible — to yourself and others — through what you choose, how you respond, what strikes you as obvious or horrifying.

The technique works because of a gap in human cognition: we often don't know what we actually think until we're confronted with a specific case. Ask someone "do you value fairness or efficiency more?" and they'll often say fairness. Put them in a resource-allocation scenario and watch what they actually choose. The thought experiment closes the gap between stated values and revealed preferences. It makes the implicit explicit.

This is why the best thought experiments are slightly uncomfortable. They're designed to push into territory where your instincts and your stated principles diverge. That divergence is the data.

Maxwell's Demon (1867)

James Clerk Maxwell imagined a tiny demon controlling a door between two chambers of gas. The demon lets fast molecules (hot) pass one way and slow molecules (cold) pass the other, creating a temperature difference from a uniform gas without doing apparent work — seeming to violate the second law of thermodynamics.

The experiment sat unresolved for 70 years until Leo Szilard and later Charles Bennett realized: the demon has to measure and remember the speed of each molecule. The act of acquiring information has an entropy cost. When the demon erases its memory to process the next molecule, that erasure generates heat — and the second law is preserved.

This thought experiment, over 70 years, forced the physics community to confront the relationship between information and thermodynamics. It ultimately contributed to the theoretical foundations of computing. The demon was fictional. The insight was real.

Schrödinger's Cat (1935)

Erwin Schrödinger designed his cat to be hostile. He was a critic of the Copenhagen interpretation of quantum mechanics — the view that particles exist in superposition until observed. He wanted to show that applying this interpretation to macroscopic objects produced an absurdity: a cat that is simultaneously alive and dead until someone looks.

The experiment didn't resolve the question — it sharpened it. It forced physicists to specify where the quantum/classical boundary lies and why. Some said the cat collapses the wavefunction itself; some said the observer is the key; some (the many-worlds interpretation) said both cat-lives are real in branching universes. The cat is not the answer — it's the pressure that forced answers to be given.

The Trolley Problem (1967, Foot; 1985, Thomson)

Philippa Foot introduced the trolley, Judith Jarvis Thomson refined it with the footbridge variant. Decades of research followed. What emerged was a map of human moral psychology:

- People treat direct versus indirect harm differently, even when outcomes are identical - People treat using someone as a means to an end differently from redirecting a harm already in motion - These intuitions are stable across cultures and appear early in child development - They conflict with utilitarian calculus in systematic, predictable ways

The trolley problem didn't answer what morality is. It revealed the hidden architecture of what morality actually is — in practice, in the human mind — as opposed to what we claim it is when we're not under pressure.

The Veil of Ignorance (Rawls, 1971)

John Rawls's A Theory of Justice used the veil of ignorance to derive principles of justice that no rational person could reasonably reject. His argument: behind the veil, not knowing your place in society, you would choose:

1. Equal basic liberties for all 2. Social and economic inequalities arranged to benefit the least advantaged members

Whether you accept his derivation or not, the tool is powerful. It provides an impartiality test for any policy, rule, or social arrangement: "If I didn't know which position I'd occupy, would I still choose this?" It cuts through self-interest and rationalization to something closer to principled reasoning.

Politicians, executives, and parents who actually applied this test — genuinely, not rhetorically — would think and behave differently.

The thought experiment is a tool you can wield for your own decisions. Here is the structure:

Step 1: Identify the assumption or value you're testing. Not the decision — the underlying principle. "Should I take this job?" is a decision. "How much do I actually value security versus growth?" is a value. Design around the value.

Step 2: Build the scenario. Create a situation that isolates your variable. Strip out everything else. Make it vivid — you need to actually feel the scenario, not just understand it intellectually. Use names. Use a specific place. Make it real enough that you have a genuine response to it.

Step 3: Push to the edge. The scenario that reveals the most is usually the extreme one. If you're testing a rule, take it to its limit. "Always keep your promises" — what if keeping the promise requires serious harm to a third party? Now you know what you actually believe. "Never lie" — what if honesty in this situation destroys someone who cannot change the thing you'd be honest about? The edge cases are where real values reveal themselves.

Step 4: Follow the experiment wherever it leads. This is the discipline. The entire value of the thought experiment rests on your willingness to continue when the result becomes uncomfortable, when it implies something about yourself or your position that you didn't expect or want. Stopping early because the conclusion is inconvenient is not thinking — it's theater.

Step 5: Interrogate the response. What did your gut say? What did your stated principles say? Where do they diverge? What does the divergence reveal? The divergence is the point of entry for genuine self-understanding.

Career and major choices

The "deathbed regret" experiment: imagine yourself at 90, looking back. Which choice would you regret not having made? This works because it removes the short-term fear that makes cowardly decisions feel reasonable. From 90 years, the cost of not trying usually looks much larger than the cost of trying and failing.

Value conflicts

Design a scenario where your two values cannot both win. "I value loyalty and I value honesty." Now: your close friend does something seriously wrong. Loyalty says cover it. Honesty says expose it. Which wins? You now know your hierarchy. You can't know that from the abstract statement of both values simultaneously.

Risk tolerance

"What is the worst realistic outcome of this decision? Can I live with it?" Not the worst imaginable — the worst that could actually happen, given what you know. If yes, the fear constraining you is not about the actual risk but about loss aversion or ego. If no, you've found genuine risk that deserves weight.

The single most important skill in thought experiments — and the rarest — is completion. Most people run a thought experiment until it produces a result they don't like, and then they introduce a complication, change a parameter, add a condition. They redesign the experiment on the fly to avoid the conclusion.

This is exactly what thought experiments are designed to prevent. The whole point is to go where the logic goes. If the experiment implies something you don't believe, the right response is not to change the experiment — it's to figure out which premise is wrong, explicitly, and revise from there. That's philosophy. That's actual thinking. The alternative — tweaking the scenario until it outputs the answer you came in with — is not thinking. It's performance.

Einstein didn't stop when riding alongside the light beam produced weird implications. He followed the weirdness to its conclusion and built relativity from there. The willingness to follow the experiment wherever it goes — especially into discomfort — is what separates thought experiments as a genuine design tool from thought experiments as decoration for pre-formed conclusions.