The Surprising Answer to the Trolley Problem
Kevin Maynard Co-Director of the Institute of Ethical AI
It was interesting to see the answer given on law.stackexchange regarding the legal take on the trolley problem: https://law.stackexchange.com/questions/1639/what-is-the-legal-take-on-the-trolley-problem
The trolley problem set out in law.stackexchange is as following:
“There is a runaway trolley barreling down the railway tracks. Ahead, on the tracks, there are five people tied up and unable to move. The trolley is headed straight for them. You are standing some distance off in the train yard, next to a lever. If you pull this lever, the trolley will switch to a different set of tracks. However, you notice that there is one person on the sidetrack. You have two options: (1) Do nothing, and the trolley kills the five people on the main track. (2) Pull the lever, diverting the trolley onto the side track where it will kill one person.”
The proposed answer from the legal point of view, which is the safer option, is to do nothing and let the five people die (assuming that you’re an innocent bystander, not an employee of the railway company or the train company, and have no duty to act). Copying the law stack discussions: if you do nothing, then it is unlikely that you would be charged with a crime — you had no duty to fulfil, and therefore not negligent. There’s little doubt that not pulling the lever is the safer option. Should you choose to pull the lever — then it’s probable that you would have charges of murder, or at least manslaughter, brought against you by the state as you actively caused the death of someone.
If the analysis in this case is correct, it makes you think whether there are situations where these sort of morally negative consequences occur in AI-based systems.
