Our civilization has, in an unprecedented time, run the distance between the desktop computer and a simple navigation app, self-driving cars and the ever more dominant use of artificial intelligence in our lives. It’s a race we’ve run without sight of the finish line, and the question of whether humanity is saving or damning itself with the help of AI remains. I have no firm stand on either position. It’s without any doubt, however, that if we want to find a safer place in outer space, we will need tons of AI to get us there.
As I was driving back to DC last Christmas, I needed the navigation application Waze to get me out of Brooklyn, onto the Verrazano bridge, then across Staten Island into New Jersey, and finally to the I-95 that leads to DC. By the time I got on the NJ Turnpike, it was dusk, but the sky above us was illuminated by the long line of airplanes in the sky, stretching far to the south. The planes were cruising slowly in their automated approach, waiting for permission to land at Newark Airport. Getting on the straight highway road south, I felt I could do with an automatic pilot, too. The I-95, with a couple of exceptions like the Delaware Memorial Bridge and Baltimore tunnel, is such predictable and boring drive that you wish something would happen. I forgot that I still had Waze on when a robotic female voice ordered me to keep right and then use the part of the highway that is also used by trucks and buses. Why not the one reserved for passenger cars only (and does Waze know my vehicle type)?
In my total boredom, and based on my previous experiences with Waze, I gave in. So I joined the trucks and buses and others like me (using Waze?) on the part of the highway that was evidently busier than the one next to us. So, with NPR humming in the background, my wife and now this one-sided dialog with Waze were the only things to make the drive more interesting. I tried to enter into Waze’s mind, attentive to what might possibly reveal the reason for this particular and seemingly incoherent Waze order. Was it a traffic jam somewhere down the highway, a car accident I couldn’t yet see? I had no indication of why Waze wanted me on the road with more dense traffic, yet I followed it anyway. The traffic on the parallel highway was traveling at the same speed until about 50 miles later when the two highways merged again.
Is Waze, therefore, not only navigating its users but also an entertaining them by generating some kind of dialogue? One which involves other passengers in the debate on what is or is not the right road to take? Whenever Waze or Google maps, for that matter, are set on speaker mode, there is no need to explain your driving to anyone; that voice in your vehicle is your pilot, and it’s the voice of the future.
For those who might not know what Waze is, let me explain that is one of the fastest growing navigation apps worldwide. Created by a bunch of Israeli guys, it was bought by Google in 2013 for $1.2 billion. While Waze is available in only 13 countries, and Google maps are universally known and used, the two-run separately. Waze is community-based; it requires a data connection and offers a high level of customization, allowing users to indicate to fellow drivers potholes, upcoming speed traps, accidents, and traffic. Google Maps is instead more data-based; it offers walking, driving, public transport and biking options for any route, predicts estimated times for future routes, and its data, including business data such as menus, hours, and phone numbers, is available offline. This was not the first time I used Waze, of course. I used to use tomtom (to my shock, they still sell them), but with the rise of smartphone navigation, first with Google maps and now Waze, I must admit that these navigators are highly effective and convenient. Not only because of my frequent traveling, but also somehow because of my advanced-tech family, and the all-female cast of navigation voices. Now, other than the fact that navigators often complicate your life with inaccuracy (tomtom), or are too pedantic in their time-saving (Waze), they also save you when winding in and out of the urban traffic jungle.
Would I ever see the butcher in Ridgewood–deep in Queens–when I need him for sarma cooking without a navigation app? Never. At least not in time to find the shop still open. So there are many situations in which navigating apps are useful and Waze sometimes even gets surprisingly creative, as this little fragment from one of my previous travels shows. But whether I trust these navigating apps is another question. Would I like to see that self-driving car be made available as soon as possible, based on a similar technology? Just recently, while I was at a writers conference, I heard Bret Stephens saying that people should not oppose self-driving cars just because they will take away the jobs of the drivers. “Be creative and think what you would be able to do while a car would drive you to the address you gave it. You could have a massage in the car, for instance,” Stephens argued. I would not be against a self-driving car, if affordable. To my mind, navigation apps and Uber have already introduced us to the idea of sitting in a self-driven car. But do I trust the Uber driver? Not entirely, as I will not entirely trust a self-driving car. Perhaps this is why Stephens was trying to convince us with a massage, the ability to relax en route. I don’t mind being driven around, but then again, why would one buy a car that one cannot drive? So promoting a self-driving car in the public service, individual or collective, would be the right idea.
Let’s see how close we are to the moment when we will be able to hand over our steering wheels to AI. There about 20 companies around the world that are seriously working on their autonomous car projects. Some of these driving robots might hit the road by the end of this year. That is if the local authorities will provide the companies with the adequate permissions, and only when the producers prove absolute safety for this kind of cars by meeting the standards of regular traffic. Recently, there have been a couple of minor accidents of tested Tesla and GM self-driving cars. The mishaps may slow down the procedure, and require more safety tests before the automated vehicles get the green light.
Their celebrated birth is not far, and the appetite of the industries is growing. In case you cannot wait for that moment to arrive, you can read or listen to how it feels to ride Google’s Waymo self-driving car, Chrysler Pacifica, in this report by Alexis Madrigal for the Atlantic. I suggest listening to the recording of the piece with Madrigal’s voice. It makes the article sound more remote, like the writer’s voice is of one of the old radio plays from when the world was still black and white. In the piece, there is a science-fiction-like description of the secret Waymo course where the car has been tested, and also an explanation of the software that may become the heart of self-driving cars. They call it Carcraft.
However, as Madrigal wrote his report six months ago, we may assume that in this age of rapid technological innovation Waymo has made further progress, even if no longer in the lead. Close to the finish may be a company like Tesla, which in recent weeks seems to be lagging behind but only because it’s testing a different technology than most of the competitors.
And while Google and Uber are at loggerheads for the presumed theft of documents by former Waymo engineer Anthony Levandowski, who passed them to Uber, competition is also increasing in Asia. As the South China Morning Post recently reported:
In the hyper-competitive industry for AI-based technology applications, the recruitment of top talent has become the most significant challenge amid a severe shortage of qualified people. Globally the AI talent pool is only about 300,000, far short of the millions required by the rapidly growing industry, according to a report by Tencent Research Institute. Competition for the top tier of talent is the most intense, as there are fewer than 1,000 people in the world considered capable of steering the direction of AI research and development, according to the report.
So when the two legendary programmers James Peng and Lou Tiancheng left Baidu to found Pony.ai, an artificial intelligence company that seeks to become the major Chinese competitor of Waymo, they had no problem collecting one hundred million dollars in investments for their start-up. They became talent magnets in a crucial battle for the best and brightest in AI. As SCMP reported, three of China’s top five coders now work for Pony.ai, and the company, according to Lou, owns the world’s best autonomous driving team. Pony.ai is moving fast and “plans to launch robot taxi services in Guangdong by the end of the year, where users will be able to hail an autonomous ride through a booking app like Didi or Uber,” Lou? said.
While the hottest project at the moment, self-driving cars are not the only product that relies massively on the development and implementation of AI. Look around and count the devices and gadgets in your everyday life that use AI, that new member of your family and social circle, on which (or whom) we depend increasingly, gradually making us more futile.. Once you count all the devices in your possession that utilize any form of AI, open this link, which innumerates the complete list. Check the boxes, and you will be surprised at what you’ve missed. I was, for instance, surprised that AI has been used in the financial sector and that robo-advisory services have funds under management estimated at $30 billion globally, with Hong Kong experts predicting that this figure could mushroom to $3 or $4 trillion by 2020. That is five percent of the world’s wealth management market. And with the development of voice and image recognition and other applications, AI will revolutionize many industries, from manufacturing to agriculture, medicine and security, that is, surveillance.
Last year, Vanity Fair magazine hosted Maureen Dowd, a columnist at the New York Times. Describing the dialogue between Demis Hassabis, a leading creator of advanced artificial intelligence, and Elon Musk, a doomsayer about the perils of artificial intelligence more famous for his SpaceX and Tesla projects, here is Dowd:
They are two of the most consequential and intriguing men in Silicon Valley who don’t live there. Hassabis, a co-founder of the mysterious London laboratory DeepMind, had come to Musk’s SpaceX rocket factory, outside Los Angeles, a few years ago. They were in the canteen, talking, as a massive rocket part traversed overhead. Musk explained that his ultimate goal at SpaceX was the most important project in the world: interplanetary colonization.
Hassabis replied that, in fact, he was working on the most important project in the world: developing artificial super-intelligence. Musk countered that this was one reason we needed to colonize Mars—so that we’ll have a bolt-hole if A.I. goes rogue and turns on humanity. Amused, Hassabis said that A.I. would simply follow humans to Mars.
As explained later in the piece, Elon Musk, who just a few days ago launched the Heavy Falcon, the most powerful rocket ever created, three years ago warned about the possibility of A.I., and our dependence on it, running amok. As in the magazine, some people at the DeepMind company are convinced that the extinction of the human race is likely to occur at the hands of technology, though not immediately. We might perhaps live to be 150 years old, but we’ll have machine overlords. Or, as Musk put it: “Let’s say you create a self-improving A.I. to pick strawberries, and it gets better and better at picking strawberries and picks more and more, and it is self-improving, so all it really wants to do is pick strawberries. So then it would have all the world be strawberry fields. Strawberry fields forever. No room for human beings.”