Imagine being a kid in the 1960s. The United States and Russia are battling over space dominance; at breakfast time, a new TV show, The Jetsons, is broadcasted; Stanley Kubrick is somewhere filming his masterpiece ‘2001 – A Space Odyssey’, and a father, on the couch, is reading ‘Do Androids Dream of Electric Sheep?’. All these things have in common the anticipation of a (near) future of perpetual innovation and radical changes in the daily life of Western citizens, possibly spreading to Communist countries after the defeat of the USSR. They picture a world in which technology has prevailed, for good or worse.
The optimistic feelings of those living in an era of continuous growth, labour productivity increases and technological marvels, however, started to tumble with the economic crises of the 1970s, suggesting that the real output growth that the world economy had experienced so far was slowing down. Despite that, in the following decades, life-changing technologies have been introduced, impacting every industry, from molecular medicine to communication, transportation to food production.
A Story That Isn’t Over Yet
The Internet revolution, sparked at the beginning of the new millennium, is only the last piece of a puzzle that hasn’t been completed yet; considering that not even close to everyone in the world has started benefiting from the technological innovations of the last 20 years, it’s reasonable to show at least some optimism.
Furthermore, researches by J. Fernald of the San Francisco Federal Reserve and S. Basu of Boston College suggest that there might be a lag of around 15 years before investments in information and communication technology guarantee the desired increases in productivity.
Not everyone, however, is of the same opinion as Fernald and Basu. A few years ago Peter Thiel, PayPal co-founder and early investor in Facebook, revealed his frustration towards what he deemed a concerning slow-down in innovation. According to the entrepreneur, the introduction rate of groundbreaking technologies (in transportation, biotechnology, energy) is being obstructed by stricter governmental regulations and an excessive focus on internet-based technologies, rather than the real world.
On the one hand, it’s curious to hear such remarks from the first professional Facebook investor, but it’s understandable considering the nonsensical fuss that was created in the last few months around the Snap Inc IPO, especially in the light of his memorable quote “we wanted flying cars, instead we got 140 characters“.
Thiel’s reasonings sound extreme, but are more reasonable than might seem. Every kid born from the 60s onwards has probably dreamt of driving a flying car one day, as well as going to Mars. And if the odds of seeing people on Mars in this generation are constantly increasing, it’s difficult to believe that flying cars will be the future of Earth transportation any time soon, especially in the consumer segment. It is likely that these kind of innovations would first be available to premium segments or fleet operators, car-sharing and rental car companies.
An Unexpected Path
Granted that Blade Runner’s world is not the one the current and next generations are going to live in, what is, however, undeniable, is that transportation has been radically changing in the last few years, and energetic entrepreneurs like Elon Musk have been fostering a process of democratising the high-end car industry. The case of autonomous vehicles is not only meaningful from a technical standpoint, but it is also of great interest because it exemplifies human evolution and the future of the relationship between mankind and technology.
While it is true that self-driving cars are somehow a less sophisticated version of what was predicted to be the standard in the third millennium, they represent more than that, because they bring with themselves a set of issues that will shape the society of the future and how people relate with machines in general – not just cars.
Take improvements in AI, for instance – the new car driver. In 2004 a Defence Advanced Research Project Agency (DARPA) study ended with no driverless car being able to finish the predefined route, yet today autonomous vehicles in the consumer segment are coming close to reality. A WEF report recently highlighted intensifying concerns regarding the development of AI technologies, not only in terms of a lack of regulatory provisions, but also, in more practical terms, the undesired role they will potentially have in people’s lives.
While many commentators are positive regarding the automation opportunities in applications such as driverless cars and production robots, others (such as Stephen Hawking, Elon Musk and Bill Gates) fear that uncontrolled developments in sectors such as the autonomous weapons industry could escalate into unforeseeable scenarios, where it will no longer be understandable how machines have been designed and to what extent decisions are being taken consciously by the machines itself.
The regulatory body related to self-driving cars and AI is still at the infant stage. While governmental bodies like the US National Highway Traffic Safety Administration (NHTSA) are in charge of proposing safety regulations and promoting awareness related to the different kinds of autonomous vehicles today present on the market (in the US), it is still up to individual countries how they regulate the market and to what extent they allow self-driving cars to run on public streets. And such decisions do not seem the easiest.
At the same time, technological disruption in the car industry will affect also other important players in the industry other than the producers: insurance companies, for instance.
According to figures cited in a recent KPMG study, the frequency of accidents has fallen 50% in the last 15 years and will continue doing so, especially considering that more than 80% of car crashes are caused by human error. Recent data from the Insurance Institute for Highway Safety suggests that the design of autonomous vehicles and their safety systems are extremely successful at reducing accidents.
While this will probably claims damages and liability exposures by billions (and, of course, save thousands of lives) it will be crucial for insurance companies to understand how they could re-engineer their business processes and actuarial models to survive a world where cars will be less and less frequently involved in accidents. How will premiums be calculated? What kind of insurance products will be sold? It’s reasonable to think that if insurance costs for the consumer are going to decrease, prices for repairs in such technologically advanced cars are going to skyrocket.
A Web of Issues
At the same time, if the vehicle is going to be able to make autonomous decisions according to its software and learning capabilities, who is going to be responsible in cause of failure and potential accidents? As technology advances, new legal issues and opportunities are born for the same companies and for the legal system as well.
What about moral issues? When talking about an inanimate object driven by software, one of the first things coming to mind is the ‘trolley problem’, a well known philosophical dilemma that makes the individual choose between acting or avoiding action and the effect on other people’s lives.
MIT has developed a platform to investigate how people would answer to different scenarios in which a car, with no driver at the wheel, has to decide whom to kill in case of mortal accident.
Not so surprisingly, people are somewhat consistent in their utilitarian analysis of who should deserve to die in such an accident (fewer people is better than more, older is better than younger; criminals are better than honest people, and so on) only as long they are not the passenger that gets sacrificed during the accident.
So, who is going to code how the car should act in such a difficult situation? And who is going to buy a car that would eventually decide that it’s better to kill the passengers rather than people crossing the road? So far, such situations haven’t taken place, but it must be considered how the vehicle should react in such unforeseen cases. After all, how else will it learn?
Nature versus Machine
When human nature meets the machine, there’s no black and white and, as a consequence, it’s always difficult to compromise. Ethical questions are always going to be among the most relevant aspects to be considered – what about hacking and privacy, for example – and they will probably extend as the technological tide involves more sectors and affects multiple constituencies.
The case of autonomous vehicles and the few examples outlined above give a taste of what is going to be crucial for the development of society in the following decades, regardless of the field of interest. It suggests that it’s crucial to avoid underestimating the power of innovation, even if it seems to follow different paths or is slowing down from those once forecasted. It suggests that there is much more than the introduction of a new technology, when this completely redefines precedent paradigms. The future is now, and the world must be well-prepared to face it.