By Peter Svarre

This article is published in connection with The Digital Ethics Compass project.

Imagine that you are a train driver and headed towards a tunnel. Inside the tunnel, there are five people working on the tracks. Suddenly the brakes on the train stop working and the train is going so fast that the five people working on the tracks do not have time to get away. It is certain that they will be killed by the train.

In the last moment, you discover a remote-controlled switch that changes tracks and you can use it by pushing a button on your console. If you press the button, the train will head towards a different tunnel. In this tunnel, there is only one person working on the tracks that will be killed.

You know nothing about any of these people.

Do you press the button?

Now imagine a different situation:

You are a doctor working at a hospital where five patients are terminally ill. They will die before nightfall if they, respectively, do not get a new heart, a new liver, a new kidney, new lungs and a new spleen. You notice a healthy nurse with the save blood type as the five patients. You are 100 per cent sure that she has the exact right kind of organs that the five patients need.

You can do two things:

1. You can sedate the nurse with chloroform, put her on an operating table and remove her organs. You will save five patients and kill the nurse.
2. You can do nothing. The nurse will live on and know nothing of what could have happened, but the five patients will die.

What do you do?

We have tested these dilemmas multiple times and in many different kinds of gatherings, and the conclusion is almost always that in the first example they choose to press the button and kill one person rather than five. In the second example, however, people choose to remain passive and let five people die rather than actively saving them by killing another.

These dilemmas are of course contrived and quite unrealistic, but in our experience, they really open up interesting discussions about ethics. For example:

1. Does it matter who dies? (Are some people considered ‘worth more’ than others?)
2. Does proximity/identification mean anything when deciding? (And what can digital designers learn from this when digital design is often about pushing interfaces in between people?)
3. What ethical tools do we use to solve the two dilemmas (people often used the ethics of consequence in the first example and the ethics of duty in the second one).

The train dilemma is often mentioned in connection with artificial intelligence and self-driving cars, and many companies working with AI have begun the processing of finding solutions to ethical dilemmas. They imagine that you could resolve 2,500 years of philosophical disagreements and somehow code the solution.

In The Digital Ethics Compass project, we have the exact opposite approach towards ethics. When we face an ethical dilemma, it is not about finding the “correct” solution. Having ethical competencies is about being extra aware that the world is not always black and white. There is no rulebook for ethics of duty and there is no Excel sheet for the ethics of consequence. Ethics is about being able to make choices despite being uncertain. It is completely irrelevant and illogical that when faced with the train hypothetical scenario you press the button while faced with the medical hypothetical scenario you spare the nurse – but it makes sense ethically. To some people, that is!