Three decades ago, the internet was just beginning to revolutionize human communications. Little did the world know how much power would fall into the hands of a few technocratic elites as a result. Autonomous vehicles likewise will transform human transportation in the same way; the skill of helming the wheel will no longer be necessary in about a decade or two, just as the art of writing on paper has all but ceased to exist.
Recent news of a so-called Apple Car project has done little to bring positive attention to the possibilities of a self-driving revolution. In poll-after-poll, nearly half of Americans say they would not use an autonomous taxi or ride-sharing service. According to PAVE, an advocacy group for autonomous vehicles, 2 out of 3 people believe the disadvantages of this technology will outweigh the advantages.
As the founder of an AV startup, I believe this technology will save many lives. It not only represents a fledgling industry of innovators that can challenge the expanding footprint of Big Tech, but a thriving culture primed to draw us closer to public acceptance.
Polarizing as it might be to weigh the net benefits of “writing” one’s thoughts on Facebook or Twitter, there is one worthwhile reason self-driving cars will be transformative in a good way: they are significantly better drivers. In study-after-study, they are proven to be superior decision-makers when the chance of human error is reduced to zero. Progeny will remember the millions that died anually in preventable car accidents as tragic figures tolerated for too long by a less advanced society.
This presents a major moral dilemma: the autonomous vehicle is perceived as an existential danger, thus technologies that have the capacity to make our lives safer, will be underutilized because of a cultural backlash against their use. To resolve this problem of public acceptance of AVs, we must first address the bad ethics rampant within the industry.
#1 We’re Ignoring a Strong Distaste for AV
The neural networks powering autonomous vehicles have encroached so frighteningly close to surveillance, some developers have refused to continue working on them. Alexey Bochkovskiy, a lead machine learning engineer for a computer vision program called YOLO — which some AV geeks consider a major breakthrough in building autonomous cars — has said he no longer believes it is a technology for the greater good.
“I stopped doing [Computer Vision] research because I saw the impact my work was having. I loved the work but the military applications and privacy concerns eventually became impossible to ignore.”
Ask anyone you know if they’d use a self-driving car, and if you have the same experience I’ve had for many years, you will get a near unanimous “no” with a multiplex of reasons behind it. This apprehension is usually not the robot overlord paranoia some feel exists within the public concious, but reasonable political and social attitudes toward automation, and its consequences: mass unemployment for Uber/taxi drivers, more wealth and power concentrated in Silicon Valley, paranoia that A.I. will make life or death decisions during an accident etc.
The Pew Research Center in 2018 conducted an expansive survey of some of the top experts in Artificial Intelligence. From MIT engineers to the co-founder of the ARPANET, all were asked what worried them most about the technological changes made by 2030. The number one concern was not mass unemployment, mass surveillance, or even drone warfare. Their biggest concern was the rapid loss of human agency.
People’s blind dependence on digital tools is deepening as automated systems become more complex and ownership of those systems is by the elite. — Respondent to Pew Research Study on A.I.
Whether through expert analysis or on-the-ground discussions, the dilemma of autonomous vehicles presents itself vividly in this equation: people have become heavily reliant and disillusioned by the technological advancements of smartphones and the internet, and know the same promises are being made in a new field. It is no wonder the public perceives a dystopia where cars are no longer symbols of freedom or independence, but cogs to Big Tech’s scaled services.
If the rise of the internet can be learned from, it is that breakthroughs in software do not in themselves accelerate acceptance of a technology. It is when developers focus on the nature of human-to-machine design. Improved design in our case requires a rethinking of how technology companies will reshape urban environments. As Voyage, a fellow startup focused on self-driving systems for retirement communities proves, smaller more locally-oriented buisnesses may effectively resolve community disdain.
#2 The Growing Moral Crisis of Robotaxis
This central cyncism about autonomous vehicles — that they are inherently disempowering — has proven to be legitimate, as the biggest competitors share an identical vision of their modus operandi, one that requires customers to be fully dependent on them. Like the IoT that’s grown to run so much of our daily lives, Big Tech companies like Amazon (Zoox), Google (Waymo) and Apple also envision the car as service, or what is more commonly referred to as a robotaxi.
The current mindset of the biggest developers indicates these machines will become a major source of disenfranchisement, one on a vast socioeconomic scale. Waymo was the first to launch a robotaxi service in a major metro area last year, and it continues to be the leading competitor in the U.S. Meanwhile in China, where AV innovation is happening at a seismic pace, government-sponsored Pony A.I. and WeRide are mass-scale robotaxi projects bent toward a market philosophy that an automized future will be a world without personal ownership.
“AI will continue to concentrate power and wealth in the hands of a few big monopolies based on the U.S. and China. Most people — and parts of the world — will be worse off.”
— Mark Surman, executive director of the Mozilla Foundation
Waymo’s success has emboldened Big Auto firms like Volkswagen and General Motors to pursue a similiar track of mass-scale CaaS, as CNET reports, “that has led many car companies to partner up with smaller, self-driving car-specific startups. Ford and Argo or Honda and Cruise, for example.”
But this fleet-based model is unnecessary in the AV market, stifling the potential for innovation, and fundamentally disenfranchising.
What people fear about the self driving car is becoming the industry-wide expectation. It is one imagined to be a world of robotaxis and the corporate behemoths that run them. The enthusiasm gap will continue to widen between the few who see the technology as a lifesaver on a global scale, and the many who only see expansive ecosystems they must submit and pay into, with no sense of possession.
XEV, an Italian EV startup that has 3D printed a car with a price tag under $7500, signifies the potential for a new kind of model in industrial production. What is refered to as additive manufacturing will enable nice-oriented problem solvers, where the expediency can lead to accelerated production of autonomous vehicles that are specialized to suit the complexities of local environments and their communities.
#3 “A.I” Can Be Vauge, Negligent, and Deadly
Self-driving cars are software with wheels. They’re systems extending from an arms race known as Computer Science, but consumers and developers a like hold an ambiguous idea that Artificial Intelligence is a kind of magical, self-aware creature that will choose who lives or dies in a car crash.
As a developer, it is troubling to hear news of the fatal car crash involving Tesla’s autopilot program. A U.S federal agency had warned that Tesla was essentially using its customers as guinea pigs. This is unfortunately too familiar within the industry. Just two years ago, Uber was responsible for a fatal car crash as well, where investigators concluded negligence and narrow-mindedness led to the loss of a human life.
My startup is working on a self-driving prototype designed to solve the problem of so-called “edge cases,” or rather, any encounter that may escalate into a life or death situation. Edge cases are fundamentally errors of coding, where something as simple as the lack of light, or a signal loss can lead to a deadly error in computer vision (as was the case with Uber).
In reality, what we refer to as A.I. — Machine Learning, Deep Learning, Neural Networks— are convoluted levels of programming. Most AV developers work with the traditional fundamentals of the C+ Language, and thus build a routine in very traditional ways: how should it handle any and all possibilities of inputs, and when will it need to reform (or “learn”) newer ways of processing them?
In other words, self-driving cars cannot make choices, only humans can. What is branded as A.I. is in fact operating on the same pretense of a predetermined set of logical choices; each reflects the capacities and limitations of humans and their ever-increasing computing power.
Neither developers or consumers should live on faith that Artificial Intelligence is perfect, nor always better. We can however work with a strong rational: humans are limited because emotions tend to blur the line between rational and irrational choices. The art of A.I. must be to ensure that a computer program is not defined by the weaknesses, but the strengths of our imagination.
If you sort of step back, the car, in a lot of ways, is a robot. An autonomous car is a robot, and so there’s lots of things you can do with autonomy. And we’ll see what Apple does.”
— Apple CEO Tim Cook, with Kara Swisher on the “Sway” podcast.
Currently, some of the biggest AV develpers have given up. Uber scrapped it’s program altogether, while automakers like Tesla and Volkswagon have turned the volume down on their promises for mass-scale Level V Autonomy. Google’s Waymo is ostensibly in the lead, while the opaque project by Apple presents the possibility of a game changer.
But we must ask ourselves if the future of autonomous vehicles is inherently good, especially in an increasingly centralized, technocratic social order where a handful of corporate empires decide what is next on their list for market valuation and expansion. Is the human condition truly prioritized in a world determined by a handful of bottom lines? A world where human agency is gradually made insignificant?
Like the moral crisis we face with internet services — one in which ideas of privacy, censorship and free speech have come into stark focus — if car culture morphs into yet another Big Tech apparatus, we’ll enter a world where millions will lose their sense of freedom to a preeminent system of intelligent machines.
I believe without a doubt bold and creative startups can transform the public acceptance of self-driving cars. With readily available and increasingly affordable technologies like Lidar and 3D Printing, we hold an opportunity to rethink the web of problems awaiting in our urban environments. And when we move forward knowing it is in our scope to address the daunting ethics of autonomous vehicles, only then will the power of this technology truly rise to the greater responsibility of improving human transportation.