Here is the full transcript of Alessandro Lanteri’s TEDx Talk: Why Driverless Cars Need Philosophers at TEDxHultAshridge conference.
TRANSCRIPT:
Many great speakers start by asking their audience to close their eyes and think about something I can’t do that with you today, because if I ask you to close your eyes and think about philosophy, you’ll fall asleep.
And I know that because most people, when they think about philosophy, you know, have this image of a bearded man pondering about impractical solutions to irrelevant problems, and they do that, obviously, sitting on an armchair. That’s not a very nice thought, just let me tell you that, because I am a philosopher, and most importantly, you are a philosopher too.
Today I’m going to show you how everyone of you is a philosopher, how I’ve been studying Fork philosophy for the past ten years – and not from an armchair, from the lab – and how this philosophy can help us get inside into very interesting problems we’re facing now, like artificial intelligence and driverless cars.
So, first of all, are you willing to participate in a small philosophical task with me? Are you all here? All right, I’m going to ask you a short question. I just want to know how you feel about it.
We’ll start with a problem we call “the trolley problem.” Some of you probably have heard about it before. So the trolley problem, basically: there’s a trolley running at full speed down the track, and it’s going to run into five people and kill them all! Unless someone pulls a switch and diverts it onto a side track where only one person is standing and he’s going to, obviously, be killed, but the five are saved.
Now the question for you is: Which thing do you think is better? To pull the switch or not?
So if you think pulling the switch and saving the five is better, just raise your hands. Anyone? Okay, I can see that’s a good majority.
Not everyone but a good 70%. That’s the kind of observation we see most of the time. So you quite compare to the rest of the people. And the most interesting thing there is: why? Well, saving five, saving one, saving five is absolutely better and what you just did there, I want you to appreciate, is what we call a philosophical theory. You formalize a philosophical, a moral theory of that. A moral theory is any “why” answer to a moral judgement.
So here you told me, “We’ll do this because saving five lives is better than saving one.” And this makes you philosophers! So congratulations! Some of you will also identify this as being a theory very similar to what we call “utilitarianism.” So the idea that you have to do whatever ensures the greatest good for the greatest numbers.
So since we’re all philosophers, let me tell you another thing here. What you have experienced is what some colleagues of mine and myself like to call “experimental philosophy.” So, if the traditional tool of philosophy is that comfortable armchair, the symbol of experimental philosophy is the burning armchair.
We want to set that on fire. We don’t want any of that. And I love this because it gets you out of your comfort zone – now, that quite literally – and forces you to see what’s going on out there. What are the many nuances of philosophy that people think and experience. And what this experimental philosophy has as a starting point is controlled experiment with real people, like what we just did with you.
And that gives us insight into how Fork theory and Fork philosophy work, but also gives us insight into interesting contemporary problems. It won’t be hard for you to imagine, a few years from now, a driverless car driving down the road, about to run into five people, there’s no space to brake. Only way to save them: steer onto the sidewalk where it’s going to kill one person only. Since most of you said that was the right thing to do with the trolley, maybe we could just code the software of this driverless car to do precisely that. Won’t be that difficult, right? But if you think so, why? Is that because most people say so in an experiment of some kind? Now – that was a driverless car.
That’s the moment when I like to go back to the experiment and start twisting things around, and see if we can get a better understanding, and double-check our first intuition. So I’m going to need your help again. Are you ready? We’re going to look at this problem from a different perspective. Let’s go back to that trolley going down at full speed against the five people. This time, unfortunately there’s no side track, so these five people are going to be killed unless we put a big weight on the track to stop the trolley.
There just so happens to be a very big, tall and large stranger, just there. You could push this man onto the track and the trolley would run into him and save the five lives. Now let’s go back to the question I asked you earlier: Who would like to push a stranger onto the track? Just one? You’re a daring man. This is exactly what we observe all the time, and it’s good when we agree on ethical issues and we all share the same opinions.
There’s obviously a catch. A moment ago, I praised you for formalizing your first ethical theory, and now your ethical theory – quiet obviously – just failed. Then that makes you real philosophers, I guess. How do we explain the fact that we’re willing to be utilitarians in one scenario and not in another one?
Now as newly found philosophers, I encourage you to think about differences between the actions required or the two scenarios, refine your theories, improve them, and capture these nuances. I’m just going to share with you one explanation that I’m fond of. And that requires taking one further step away from that armchair because I really want to start peeking into people’s brains. I know some of you would think there is some nasty torture coming up. There’s no such thing. The only thing we do is just put some participants in an fMRI machine and keep track of the flows of blood in different areas of the brain to see which area of the brain becomes engaged in different moments.
What my colleagues found is that when you’re faced with the decision to pull the switch, an area of your brain becomes activated that is typically used for deliberation, rational thinking. In the case of the stranger to be pushed onto the track, the same area does not become engaged. A different one, associated with emotional and instinctive reactions becomes engaged, and we also see that the response times are different. When you need to choose whether you want to push a stranger, you know, right away, that you’re never going to do that, before you even become aware. When you need to pull the switch, it took you a couple of seconds before you raised your hands. That makes sense because deliberation takes a little time.
Now, what does that say about your initial intuition of saving the five lives by pulling the switch? And what does that say about our understanding of ethics? You know, I’m doing research now also on the role of emotions, like you’ve just seen, also on the role of luck on our moral judgments, and the role of gender with Ronda. And I find these things fascinating because they give us an insight into human psychology and human ethics, which is a subtle combination of deliberation and emotions.
In fact, I don’t think that you can be ethical without emotions, without feeling compassion when somebody suffers, or anger when there is an injustice. And computers and robots cannot experience that kind of emotions. We can program their software to replicate human deliberation, but I’ve just shown you that even human deliberation is affected by emotions.
And this leads me to the last point of today’s speech, and that is that our software engineers are going to have a very hard time designing and developing ethical driverless cars. And for that we’re going to need many more philosophers, like me, and, definitely, like you.