Drivers or Autonomous: I Choose Machine
Should we continue to trust our human instincts? Or can technology calculate risks and make better decisions?
I spent the morning arguing with a friend about whether mass penetration of AVs is really as close as predicted. “Everyone wants to drive,” he said. And then added the classic line I’ve heard so many times…. “I don’t trust technology to make life and death decisions for me.”
“Everyone wants to drive,” he said. “I don’t trust technology to make life and death decisions for me.”
Let me break it down for you with just two figures. One, the global road/traffic-related death rate per 100,000 people is 17.4 0 — less than 0.2%. This means around 1,904 people in the UK and 13,055 people in South Africa die each year due to road-related deaths. I guess the situation is more pronounced for me having come from a country where you grow up counting the number of taxis reported on the news as overturned on the way to school each week.
Just this Christmas break I headed from my farm (3 hours North of Johannesburg on the N1) to O.R Thambo International to pick my boyfriend up. I had planned to drive there and woke up at 6 AM to get going. I got to the kitchen and found my Oupa sitting there waiting for me to wake up. He couldn’t sleep, he said. Not only was he extremely concerned for my safety on the roads (there had already been 845 road-related fatalities in the country in just the first 19 days of December), but he was concerned about me getting pulled over, either interrogated by cops or fooled into helping someone fix a tire, and then being beaten, robbed or worse if I didn’t comply.“We’ve reached joke levels, like we’re not even shocked anymore,” says an emphatic Lady Skollie, a South African feminist artist and activist.
Mass penetration of AV technology means that I can get in the car, and not have to worry that I’m going to get pulled over because the car will be programmed to drive under the speed limit, or that I’m going to be involved in a road fatality since the car uses superhuman sensors that are simply more capable than our brains to detect and avoid danger.
How badly do you want to drive now? With AV technology…
Imagine a scenario…
So my question is this: You’re in the car and you have 5 seconds to act. What do you choose, to swerve and save the child — or to save yourself? You’ll probably say that you would swerve. I can understand that. When we imagine it, swerving seems like the more instinctual reaction (as we commonly do with animals in the road) but its also a decision that you’re making with limited information i.e. you haven’t foreseen how the whole incident plays out, and now you’ve swerved and you’re headed straight for the medium where you’ll ram your car and die.
We’re obviously all imagining a different scene here, but now that you’ve decided what you’re going to do in that situation, I ask — what would you actually do? Every single situation is unique, and you can’t ever really predict what you’ll do until you’re in that situation, can you?
Its the ‘not knowing’ that autonomous vehicles eliminate though — if you had the certainty of knowing where the impact would be greater (i.e. who could be saved in the situation rather than choosing who dies) you would opt to choose the outcome based on which causes the least amount of impact i.e. ram into the medium (because you have airbags to protect you as a first line of defence and the child is saved) or crash into the child (if it is calculated that the damage caused to you ramming into the medium is fatal, but only life-threatening to the child for example).
Secondly, surely the driver should always die before the pedestrian? That’s how it goes today anyway. If you kill yourself, everyone calls it an ‘accident’ — it was totally unforeseen and is such a tragedy, but what a HERO for saving the child. Yet if you kill the child, you’re a monster and have to live with that, right? It was your choice to get into the car, knowing the risks after all (and again with autonomous technology, not much has changed).
All that’s changed with AV technology is that an outcome can be predicted with a certain interval of confidence before anything happens: you are actually in a position where the choice is made based on logical rules, e.g. the extent of impact dictates that it is either i) going to result in a critical injury or ii) death for at least one party. It then makes a logical decision based on an algorithm that factors in both a weighted aggregation of mass public opinion of ethical responsibility in such a situation, as well as the choice which would result in causing the least impact on any human involved in the incident, as opposed to leaving it to pure chance and calling it ‘bad luck’ or ‘an accident’.
Nothing is accidental. We have been playing God since the inception of vehicles being on the road, and now that we shift the responsibility from instinct to facts & logic we’re throwing our hands up in the air and saying “we don’t trust it”: Why?
This is a game changer because, by an aggregating mass public opinion of ethical responsibility, you have then also pleased a majority of the onlookers or convinced a jury, again, with a high enough certainty/confidence interval that your decision or ’the technology’s decision’ is justified.
Would you rather:
a) not know who was going to die and leave it to fate (or instinct) to decide; or
b) have a machine decide based on facts that are both measurable (i.e. impact or damage due to cause) weighted against public opinion of what the most ethical thing to do would be.
Scenarios:
1. Ram into medium: You die, child is saved.
2. Ram into child: Child dies, you are saved.
3. Ram into medium: You are fatally injured, child is saved.
4. Ram into child: Child is fatal injured, you are saved.
The technology, if smart enough (i.e. if it has ‘learned’ enough over the years) would be able to calculate the outcomes to a certain degree of accuracy by taking in factors of weight, speed, distance, force and the angle of impact (probably among other factors); eliminating at least two of the above options (i.e. at least one person will die, or at least one person will be critically injured), and it would obviously always favour a critical injury over death.
For women around the world — and of course men too — having an understanding of how technology can impact our lives is so essential to using it to empower ourselves, all the way from being able to push an emergency button in an armed robbery; to recording your experience of rape in a safe and secure manner; to developing and adopting AV technology that gets us safely from point A to point B. More important is being part of the conversation on how it can help you rather than why you don’t trust it, and shifting your mindset to problem-solving with technology rather than waiting for the right technology to arrive and then still hesitantly adopting it.
Our future world is going to be filled with technology, and I understand that the unknown is scary, but at least it is predictable. And based on all of the points above, I still choose machine.
