The trolley problem was first introduced by the British philosopher Philippa Foot in 1967. A runaway trolley is hurtling down a track toward five workers who can't escape. You're standing next to a lever that will divert the trolley onto a side track — where it will kill one worker instead. Do you pull it?

About 90% of people say yes. The math is brutal but obvious. One death is better than five.

Then in 1976, the American philosopher Judith Jarvis Thomson invented a variation. Same trolley, same five workers. But now there's no side track. The only way to stop the trolley is to push a large stranger off a footbridge into its path. His body will derail the trolley and save the five.

The math is identical. One death versus five. Yet only about 10% of people say they would push.

This is one of the most studied results in moral psychology. In one version, you're acting through a mechanism (a lever). In the other, you're acting with your hands (a push). Brain imaging by Joshua Greene shows that the two scenarios light up different brain regions. The lever activates a deliberative reasoning area. The push activates emotional aversion circuits.

The implication is uncomfortable. Our moral intuitions aren't a coherent ethical theory. They're a patchwork of inherited reflexes. We're built to be horrified at hands-on killing in a way we aren't horrified at distant deaths.

Self-driving cars and military drones make this real. They put us in the lever position routinely — far from the consequences, deciding through software. Whether that makes the decisions easier or more troubling depends on which philosopher you ask.