Over a decade ago, Mobileye could hardly conceptualize of a future where cars would drive themselves. Today, they are at the forefront of a transportation revolution.
In a meeting with ASIC Manager Mois Navon, one of the key people blazing the trail at Mobileye, the discussion around the history and progress of Mobileye threw up fresh ethical issues that every self-respecting commuter should be aware of today.
Mois Navon cuts a stately, almost imposing, figure, but his welcoming smile and warm tone could make you believe for a moment that he was inviting you into his very home. This is probably no coincidence, because along with developing Mobileye’s image-processing ASICs (think super-advanced made-to-order microchips), Mois is also an ordained rabbi. Mois tells the story of Mobileye affectionately like he’s telling the story of his own childhood. In a way, it is a story of about the growth of Israel and her start-up economy – an Israeli bildungsroman.
Mobileye began just before the dotcom bust in 1999, when Professor Amnon Shashua leveraged his research on monovision distance detection to start a company. This company was Mobileye. Mobileye produced the EyeQ chip, a custom-made microchip running proprietary software. Mois uncovered the seldom-discussed setbacks and pivots made during the early years of Mobileye, an encouraging story for any prospective entrepreneur, chalking up Mobileye’s success to a unique mixture of genius, skill, hard-work and providence. Since then, Mobileye’s vision has evolved from Cruise Control, to Advanced Driver-Assistance Systems, and finally, the holy grail of driving technology – self-driving cars.
Why should you care about self-driving cars? Well, imagine the world without combustion engines, wheels or airplanes. Soon, you won’t be able to imagine the world without self-driving cars. It’s a fundamentally disruptive technology.
Here are 5 reasons why:
- Self-driving cars could make the concept of “traffic” obsolete. Instead, everything could be run by a central computer that seamlessly integrates your car into the mega-flow of other cars without the need for stop lights, intersections, annoying things like U-turns and route-planning. Not to mention instead of relying on a chain reaction of start-stop human reflexes, long chains of cars could start and stop simultaneously. (Watch this video to nerd out about it.)
- Road and transportation infrastructure could be completely redesigned to use far less space. Our current infrastructure is designed for human reflexes and human fields of vision. With self-driving cars, we could drastically reduce the land necessary for roads and parking spaces, especially in built-up urban spaces, thus freeing up billions of dollars in precious real estate. For a land-scarce city like Singapore, with 12% of our land still occupied by roads, this could prove to be a gamechanger.
- Self-driving cars make possible the model of paying only for car-usage, instead of car-ownership, and make it radically cheaper. This means less cars on the road, and more efficient usage of cars that do exist.
- Self-driving cars are potentially far safer than any human could be. The simple reason is because the human brain is a single blob of genius, but general, processing power. Self-driving cars can have more than one eye on the road, never get distracted, or inebriated, or need to text their mother.
- Everything surrounding transport policy will change. This includes traffic tickets, road taxes, ERP, COE, etc. The government will be able to provide the usage of cars to more people at equitable rates.
So it appears that the upside of self-driving cars is too good to ignore. But are we ready for them, and the potential ethical and moral issues that could arise? In 1967 ethicist Philippa Foot published the infamous Trolley Problem: you see a runaway trolley careening into a group of five, but you have the option of changing its course so that it only kills one person. Do you do it? Since then, various philosophers, including Judith Thomson, have extensively analyzed and iterated on this issue. The relevance could not be more obvious; just replace “trolley” with “self-driving car”.
Mois presented us with two broad positions, the deontological and the consequentialist. Deontologists are simply concerned about following rules governing actions, whereas consequentialists focus on the – wait for it – consequences of the action. Mois walks us through other iterations of the problem, like the Tunnel Problem (Google it), or the Fat Man Problem. Being an ordained rabbi, he also put these ethical problems in the conversation with Judaism’s rich ethical tradition.
But what did Mobileye decide to do?
The answer is to put the question in the hands of the governments, because it’s their job to make difficult decisions about things like ethics and moral opinion. Mois says in his article about the ethics of self-driving cars and Judaism that “the final word on this issue is still being debated and much is left to be said”.
The question for us as Singaporeans, then, is how will the Singapore Government decide?
We have been (and prided ourselves on being) non-ideological, non-religious and austerely pragmatic. Should we look back to rich ethical traditions, and if not, how do we negotiate between different ethical points of view? These questions remain open, and the only thing that is clear at this point is that we will benefit from more discussion, not less.
I will end with offering my own point of view. Perhaps what we need to do is to reframe the discussion from an infrastructural perspective, not an ethical one.
“In the case of the autonomous vehicle, there is no driver, there is a program that is being executed according to some predetermined code. That code was written days, months, or, in all likelihood, years before it encountered this Trolley Dilemma,” says Mois.
From that perspective, it would be technically incorrect to consider the autonomous vehicle an ethical being, making decisions in the moment. It is a machine, and it is programmed to perform a function, not to choose between lives. Any system could cause deaths if not managed properly, but in my opinion, if self-driving cars become a reality, the moral weight will have to be borne by the managers of that system who are ethical agents. Each pedestrian must take responsibility for their own safety, knowing that there are cars on the road. Do we punish the train conductor when someone dives onto the tracks? No, we build better infrastructure to ensure freak accidents are less probable.
But nobody can engineer the workings of the human heart.
Photos and visuals courtesy of MobileEye.