Thought experiments are a great way to understand yourself better, and to make explicit what is normally implicit. Ex: “If you must choose between letting a random baby or elderly person die, which would you choose?”
But thought experiments also suffer from problems:
1. It can be difficult to accurately imagine a world that incorporates the stipulations of the thought experiment.
Ex: “Suppose that you suddenly stopped experiencing any fear and any empathy – what would you change first about your life?”
2. They can incorporate constraints that may be incoherent, yet gloss over that incoherence making them seem sensible.
Ex: “Imagine that you are as much like yourself as possible, but a chicken. Would you then rather not exist, or live in a factory farm (and get slaughtered)?”
3. They make for unemotional analysis that may ignore what you’d really experience, leading to a false sense of what you’d do.
Ex: “If it was your only weapon, would you kill a terrorist with a spoon if you knew he planned to set off a bomb that would kill thousands of people?”
4. They usually ignore the fact that a described state of the world could be arrived at through many paths, and what path it got there through sometimes really matters.
Ex: “If you could remove all guns from people’s homes in the US, would you do it?”
5. They are often ambiguous, which may be necessary for simplicity’s sake, but which can obscure key elements (like what repercussions and second-order effects could occur).
Ex: “Should a doctor let a patient she can save die so that she can harvest their organs to save 2 others?”
Thought experiments are great as a tool, but they should be used cautiously and thoughtfully. There is an art to using thought experiments well.
This piece was first written on November 10, 2020, and first appeared on my website on September 25, 2026.
Comments