9. The Future History of Transportation
Chunka Mui
Futurist and Innovation Advisor @ Future Histories Group | Keynote Speaker and Award-winning Author
Imagine a world where everyone had rapid access to safe, cheap, and dependable autonomous ride-sharing services. It's possible.
People will no longer have to buy cars, which are typically their second biggest purchase after a home—even though the car sits unused for more than 95 percent of the time. There’s no need for insurance for cars that don’t hit anything. Hospital emergency rooms, happily, will lose a lot of business. There'll be no need to design our homes, stores, office buildings, and cities around garages and parking lots needed for personal vehicles, because ride-shared vehicles don't need convenient parking. And, the huge inefficiencies associated with today’s car sales, licensing, maintenance, and repair will diminish because they will be dealt with by companies managing fleets.
This new world of transportation would help tackle poverty and loneliness, because access to affordable transportation is one of the biggest factors in economic opportunity and quality of life. It would greatly improve health, giving people better access to care, among many other benefits. It would help ensure equal access to transportation for the millions with disabilities and no option to drive themselves. And, it would be much better for the climate, since such services would be built on electric vehicles rather than carbon-polluting gas vehicles.
Because of these massive potential benefits, we think it would be crazy not to take advantage of the Laws of Zero related to Transportation to make inexpensive, autonomous, electric vehicles widely available by 2050.
In this week's serialization of "A Brief History of a Perfect Future," coauthored with Paul Carroll and Tim Andrews , we offer a Future History of Transportation to illustrate what is possible and explain why that future is attainable.
We hope you'll take a read or a listen, give us a "like," and share your thoughts.
CHAPTER 9 — The Future History of Transportation
Future History Scenario: One Hundred Billionth Ride for Driverless Cars, July 17, 2035
MORGAN HILL, CA — Steve Mahan had a long and distinguished career, including as the executive director of a center that helped the visually impaired in Silicon Valley. He may be better-known, however, as Google’s Self-Driving Car User #0000000001. Yes, Mahan was introduced as the “driver” in the very first demonstration video of Google’s self-driving cars, back in 2012.[1]
Today, as the AV industry is poised to log its 100 billionth ride, let’s look back at that first trip and the wild (but almost entirely safe) ride AVs have taken since then.
The world met Mahan as he slid into the driver’s seat of a Toyota Prius with a strange, flower-pot-shaped contraption bolted onto a rack on top of the car. After some banter, he and his passengers decided to go get some tacos and burritos. But Mahan never touched the steering wheel or gear shift. He merely told the car the destination, and it took them to a Taco Bell drive thru. Next, Mahan “drove” to his dry cleaner. At that point, the voiceover narration noted that Mahan is blind.
Mahan was living in the future — which has now caught up.
According to a study by the Ruderman Foundation back in the 2010s, six million Americans with disabilities had difficulty getting the transportation they needed.[2] They struggled with the inability to drive themselves. They struggled with public transportation, too; there weren’t enough accessible buses and trains, and many of the sidewalks and elevators that were needed to get to those buses and trains weren’t accessible, either. People with disabilities struggled with the high cost of cars that could accommodate them. They struggled with human-driven paratransit services, which were far too expensive for most transit agencies to provide adequately and which were non-existent in rural areas, sometimes referred to as “transportation deserts.”?
But, thanks to the Laws of Zero, all the necessary retrofitting that cost hundreds of thousands of dollars on that 2012 Prius, including the flower-pot-shaped lidar on its top that used spinning lasers to track all nearby objects, now cost a tiny fraction of their 2012 prices and have been easily incorporated into all cars on the road. Every person, blind or otherwise, gets to “drive” an AV now.
By the time Mahan appeared in that 2012 introductory video, Google AVs had already safely driven 200,000 miles. By 2020, Google had spun out a subsidiary named Waymo to commercialize the technology, had a fleet of 600 test cars, and had logged more than 20 million road miles in more than two dozen cities, plus billions more miles in sophisticated computer simulations. Waymo had even launched its first commercial service in Phoenix.
The biggest challenge for the commercialization of these AVs turned out not to be having them learn how to drive. Throw any road, weather, or mechanical scenario at both a human and AV driver, and the AV will do better every time. The biggest challenge was getting AVs to learn to deal with humans, and vice versa. Pedestrians, cyclists, and other human drivers can do some really hard-to-predict (okay, crazy) things. Passengers — including the inebriated, the messy, and the overly amorous — proved to be quite a challenge, as well, as services scaled up.
So Waymo and early competitors like Amazon, General Motors, Aurora, and Uber spent a lot longer ramping up than any originally expected. But they all followed a mantra of “think big, start small, and learn fast” and developed a steadily larger set of services in a growing array of markets. This approach let service providers explore aggressively, learn constantly, and deploy as they were ready, but not before.?
Meanwhile, service providers felt less pressure to rush to market because the massive investment in consumer-focused AV technology enabled a much narrower but quite profitable business model: trucking. Driving on highways doesn’t typically involve passengers or pedestrians, and, on such roads, the driving and the other drivers are much more predictable. Developers and regulators had been wary of trucking in the early stages — “one video of an AV semi crushing a minivan, and we would have been done,” one analyst commented at the time — but perception shifted by 2021. With so many safely driven miles already logged by its cars and with growing public and regulator acceptance, Waymo pivoted and soon dominated trucking, a $700 billion-a-year business at the time.?
The market for autonomous trucking has now exploded. For safety reasons, drivers had been limited to 11 hours a day on the road in 2020, but the introduction of AVs in the trucking realm meant that rigs could now run 24/7 — and without the expense of a driver. A cross-country trip that once took a human driver five days now requires only two. Trucking no longer has to be optimized to fill a full trailer to justify the expense of a driver. “Trucks” of all sizes can now be dispatched based on need, not on the size of a standard container. As a result, many shipments have eliminated the old, intermediate step, where they’d go to a warehouse and be repacked into other vehicles for local delivery. Businesses and people are happy to get their items much more quickly.
The supply chain for businesses was reinvented.[3] Oh, and that chronic shortage of truck drivers pre-2021? No one even remembers that problem existed.
The profits from the burgeoning autonomous trucking business fed back into the consumer efforts and created a virtuous circle. Small, geo-fenced efforts like Waymo’s service near Phoenix back in 2020 gradually expanded to cover whole metropolitan areas. AV offerings then moved beyond such weather-friendly areas to all major and even medium-sized markets, and urban redesigns have been integrating AVs into the fabric of life, making them even more valuable. We’re now to the point where the federal government is debating whether to mandate universal access to AVs, akin to the universal access requirement for telephone service in the early 1900s.??
Already, just 15 years after Waymo launched in Phoenix, it and the competitors it inspired are running AV transportation services in and are connecting more than 500 towns and cities in the U.S. Tens of millions of disabled, elderly, and economically disadvantaged persons now have access to cheap, fast, and accessible transportation services.
“The driverless car changed my life,” Mahan said. “It gave me the independence and flexibility to go to places I both want to and need to go, when I need to do those things.”
How to Build that Future:
Without breakthroughs in transportation, we’ll keep coasting along toward a polluted, congested future. While electric vehicles will grow as a percentage of cars on the road, the vast majority will still be powered by internal combustion engines. While we’ll build lots more roads to try to alleviate congestion and limit the number of people sitting in rush hour traffic cursing their fates, they’ll still be stuck in traffic, just in sleeker versions of the cars they sit in today.
We think it would be crazy not to have inexpensive, autonomous, electric vehicles widely available in 2050.
While autonomous vehicles may seem to have come out of nowhere in recent years, they, like many innovations, actually have a long history that involved untold numbers of smart minds — and somehow trace back to Leonardo da Vinci. What didn’t that man foresee? A sketch he drew circa 1478 of a gear-laden, three-wheel cart has been shown to be a workable, self-propelled cart with programmable steering.[4] (The cart bears a striking resemblance to the Spirit rover used by NASA on Mars.)?
General Motors took up the AV cause at the 1939 New York World’s Fair, unveiling a 20-year vision of radio-controlled electric cars powered by circuits embedded in the road.
Almost on schedule, a 1956 Saturday Evening Post advertisement updated GM’s vision. It depicted a family of four playing dominos in a car configured more like a living room than a car, with a caption that promised “no traffic jams… no collisions… no driver fatigue.”
The ad wasn’t fantasy. Around the same time, a 400-mile stretch of highway was built outside Lincoln, Nebraska, that included electronic circuitry. Retrofitted cars had radio receivers that communicated with that circuitry and controlled steering, acceleration, and braking. The system worked well enough that eminent researchers proposed building electronic superhighways.[5]?
The complexity of the systems and the cost of revamping all roads killed the idea soon enough, but two major things happened in the decades since. First, the Laws of Zero made it possible to imagine putting intelligence in the cars, rather than having to rebuild all the roads. This prospect propelled steady, significant advances at universities and in research labs in multiple countries.
Second was the Iraq war.
The U.S. suffered such heavy casualties from ever-more-sophisticated roadside attacks on its vehicles that Congress funded an aggressive program for developing AVs that could move without putting people in harm’s way. The goal was to make one-third of vehicles used by ground military forces autonomous by 2015.
DARPA, the U.S. Defense Advanced Research Project Agency, took up the challenge — literally. Rather than rely just on awarding typical Defense Department contracts (which were also done separately), DARPA set up a Grand Challenge in 2004 where it offered a $1 million first prize for autonomous vehicles competing on a 150-mile course in the Mojave Desert near Barstow, California. Twenty-five groups entered, and 15 were invited to attempt the course. Most didn’t make it out of sight of the starting line. One, a motorcycle, fell over right at the start. The farthest any vehicle went was 7.4 miles. Within a few hours, the challenge was shut down, and everybody went home. An article in Science referred to the contest as “DARPA’s debacle in the desert.”[6]
In fact, the challenge was a master stroke.
Importantly, DARPA had opened the competition beyond the defense contractors that were already working on the problem, and the competition stoked media attention. This helped capture the imagination and participation of a wide swath of corporate labs, university researchers and students, and even some bright high schoolers who had ideas about what they could do with a $1 million prize.
The attention grew when DARPA announced it would hold the competition again in 2005 and doubled the top prize to $2 million. This time, the field exploded to 198 applicants, with 23 finalists invited to attempt the course. Five vehicles (including an updated version of the motorcycle) completed it. A team from Stanford University won first place, while teams from Carnegie Mellon took second and third, winning a combined purse of $1.5 million.
In 2007, DARPA held another challenge on a 60-mile urban landscape course, complete with driving regulations, traffic signals, congestion, and merging traffic. This time, a Carnegie Mellon team took the $2 million first prize, while a Stanford team won the $1 million second prize.
The race was on. With just a few million dollars spent on prize money, DARPA had produced a proof of concept for autonomous vehicles and generated extraordinary interest in the idea.?
Larry Page took the next big step. Page, co-founder of Google, had been a spectator at the Grand Challenges and, with roughly all the money in the world at his disposal, hired the leader of the Stanford team, Sebastian Thrun.[7] Thrun then recruited many of the best alumni of the winning teams, including Chris Urmson, who led the rival Carnegie Mellon teams, and launched a self-driving car project at Google.?
The combination of unmatched Google assets and, arguably, the world’s greatest collection of expertise produced amazing results. The Laws of Zero in computing, communications, and information helped, too. Just look at lidar — that distinctive device on the roof of the early Google cars that used a rotating laser to track everything around the car. Even after the experimental versions went into production, a single lidar sensor still cost $80,000 as recently as 2017. But new, solid-state sensors will cost $500 apiece — a price drop of more than 99 percent — and numerous suppliers think they can cut that price to under $100 before too long. Suddenly, all the equipment needed for autonomy becomes affordable. One might almost say “free.”[8]
Tie it all together — breakthroughs in sensors and artificial intelligence, tested through billions of miles of driving in simulators, millions of miles of driving on actual roads with safety drivers, and tens of thousands of miles in fully autonomous mode — and by 2019 Google’s Waymo had 300 AVs operating in an approximately 100-square mile service area that includes the towns of Chandler, Gilbert, Mesa, and Tempe, Arizona. By early 2020, Waymo was operating an Uber-like car service handling 1,000 to 2,000 rides every week, five to 10 percent of which were without human drivers in the car as backup.
What’s more is that tiny bit of initial seed money from DARPA, amplified by Google’s moonshot and assisted by the Laws of Zero, created competition that has pushed the private sector to invest many tens of billions of dollars in AVs and the technologies that will support them. The next 30 years of AVs will surely make the last 15 look prehistoric.
While AVs will drive, if you will, the vast majority of change in transportation in the Future Perfect, two other innovations will also provide a great deal of progress, and some others may contribute.
Personal mobility devices, such as the e-scooters that are starting to proliferate, will play a significant role. At the moment, they have a neither-fish-nor-fowl problem. They’re too fast to integrate with pedestrian travel on sidewalks, but they’re too slow for use among cars and expose riders to far too much danger on roads. In the Future Perfect, where the vast amount of space used for parking has been converted into other uses, safe lanes for scooters will be available, and the scooters themselves will incorporate the kinds of safety features that are built into AVs.
Transportation will also go three-dimensional. Why limit ourselves to roads and sidewalks when all that air space is available and when you don’t have to pave the air to be able to use the space? Yes, crashes up in the sky could be catastrophic in ways that those on the ground are not. There’s no such thing as a fender-bender at 1,000 feet, and congestion will become an issue in the skies at some point. But, remember, we have unlimited computing power and communication capabilities available to us in the Future Perfect, along with an extraordinary array of sensors, so we’ll have an air traffic control system that will prevent almost all crashes. The move to 3-D will begin with delivery drones but will expand from there. By 2050, you might even have your very own flying car.
Hyperloops might also be available, based on Elon Musk’s idea for vehicles moving so fast underground you could get from San Francisco to Los Angeles in half an hour. We’re not sure whether we’ll need to do all that digging and build all that track when AVs have removed almost all considerations of time and distance for travel on roads and when flying vehicles may be available, but we’re happy test projects are under way. We’ll see what happens.
We expect breakthroughs in fuels that will remove the pressure to reduce air travel, which contributes to climate change. Yes, fuels will still need to be burned to create power, but the Law of Zero for energy means we’ll be able to go to extraordinary lengths to create fuels that can be burned without hurting the environment. Progress is being made on using hydrogen as jet fuel (and safely — plenty of people still remember the photos of the hydrogen-filled Hindenburg burning in 1937). Electrolysis can separate water into hydrogen and oxygen, and then the byproduct of burning hydrogen is water, so there’s no problem there. (Many see this “green hydrogen” as a way to take advantage of any solar and wind power that exceeds the needs of the grid at that moment and even as a sort of battery to cover times when the grid needs more electricity — you let the solar panels and wind turbines generate all they can and turn anything the grid doesn’t need into green fuel. We may also have access to hydrocarbons that are created, using renewable energy sources, based on carbon dioxide that’s extracted from the air. So, while burning these fuels would create carbon dioxide, it’d be recycled into the next round of fuel that the planes would use.
Whatever the technologies that become available over the next 30 years, we’re going to need to take out a clean sheet of paper and show considerable creativity in redesigning our environment — and history suggests the redesign will be hard, because our minds are typically so immersed in the present. Alexander Graham Bell imagined the telephone[9] as a way of broadcasting symphonies, not as a new form of two-way communication. Early TV shows were just radio shows conducted in front of a camera. Early airports resembled train stations. Unless we want to keep repeating those mistakes, we’ll need to start with the assumption that what we’ll have in the Future Perfect will be far more than a better version of today,
Think of all the constraints that will disappear by 2050. We’ve already talked about all the issues of time and distance that go away and about all the space that will open up in cities once street parking and garages can be repurposed. But those are just the beginning.
People will no longer have to buy cars, which are typically their second biggest purchase after a home — even though the car sits unused for more than 95 percent of the time. That switch from a fixed cost to a variable cost for transportation will be profound, freeing up all kinds of resources for people. That’s especially true because AVs will provide all sorts of efficiencies. There’s no need for insurance for cars that don’t hit anything. There’s no need to insure your life or health against a car accident that won’t happen. Now, cars will fit the task at hand — you can summon a big car when you need one but use a car for one or two people when that’s all you require. And, if you think about, the vast majority of trips truly are for one or two people, without cargo. The huge inefficiencies associated with today’s car sales – mainly, the tens of billions of dollars of inventory sitting on dealers’ lots – will go away because purchasing will be done by companies managing fleets. The car itself will look different. Without all the parts and weight of the driving mechanisms, and without having to worry about accidents, designers will be free to innovate on shape, size, and use case. Want a pink, bouncy house on wheels? A Batmobile? Maybe we’ll even end up with that living room of a car that GM envisioned in the 1950s, where a family of four was sitting around playing dominoes.
With all sorts of constraints gone, we can design ways to attack some of today’s seemingly intractable problems. For one, the new world of transportation could greatly improve health, giving people better access to care, among many other benefits. Transportation could also tackle poverty: Today’s public transport often doesn’t get people close enough to their jobs, but tomorrow’s will take people right to the doorstep – no more changing public buses twice and walking the last mile in the cold or the rain.
领英推荐
We’ll also surely see some positives we can’t currently predict, just as we did with the first iteration of cars in the early 1900s. The car revolutionized production, creating the need for a motorized assembly line and for standardized parts, while also rewiring supply chains worldwide. The car created all kinds of secondary markets, too. For instance, the car created the need for roadside motels and hotels and, of course, auto insurance. The driverless car, likewise, might produce whole new categories of jobs, such as long-haul “drivers” who oversee numerous trucks from their homes. AVs also might spur other gains, such as transformation of local electric grids to support all the electric vehicles.
There are certainly a lot of operational challenges when it comes to running what’s essentially a fleet of hotel rooms on wheels, as we learned in detail during our work on a consulting project with a major player in the AV space. Infrastructure that’s currently paid for by taxes on gasoline will have a problem because, in the Future Perfect, there will be no gasoline; those projects will need to find new sources of funding. To get the full benefits in the Future Perfect, the transportation system will have to be multi-modal, meaning there will need to be work on integrating AVs with public transport, scooters, bikers, and pedestrians, rather than just pushing everything but cars off the roads, as typically happens now. We’ll also have to be sure to not become victims of our own success: Making transportation free(ish) could easily encourage people to use so much of it that roads are overwhelmed with traffic.
As we’ve said, just because we can imagine a Future Perfect doesn’t mean we won’t wind up with a Future Pathetic. Creativity and care are required.
To that end, here’s the sort of future history those thinking about transportation and its implications might write to help them both envision an ideal future and then start building pieces of it to test and prepare for all the advances that will be available to us in 2050. This builds on a real event: Toyota’s announcement in 2020 that it’ll start building a prototype model city near Tokyo to explore how advances in transportation might enable radically new conceptions of what a city is.
Future History Scenario: A Living Laboratory?for the Future of Transportation, March 24, 2050
WOVEN CITY, Japan — A drone dips over the rooftop of an apartment building, edges over to the four-foot square landing pad on the third-floor balcony, gently sets down a jar of gochujang, and then flies away. Paula Endo, who had forgotten to include the Korean hot pepper paste in her order that an AV dropped off earlier that day, steps outside to retrieve the jar, returns to her kitchen, and slips back on the virtual reality goggles she was using to watch a chef demonstrate how to prepare tteokbokki, a spicy, rice cake-based dish Endo was making for a Korean friend coming over for lunch.
With lunch under control, Endo calls her elderly mother and asks if she can “come visit.” When her mother responds cheerfully, Endo engages her mother’s service robot and “enters” it. Through the VR goggles, Endo can now see through the robot’s “eyes” and, by walking around in her own apartment, moves the robot around in her mother’s. Putting on special gloves, Endo takes control of the robot’s “hands” and opens cupboards to check on medicines, then tidies up a bit for her 104-year-old mother, who is mentally very sharp but has lost some manual dexterity and tires more easily than she once did. The mother smiles at her daughter’s face on the high-resolution screen on the front of the robot, and the two have their usual, daily, 10-minute chat.
Endo’s friend Jeon Jimin arrives on the helipad on the roof of her building in the one-person mobility drone he took from the train station. After arriving by high-speed train from Tokyo, he spent an hour in the gym at the station, then showered and headed over to Endo’s. (Free gyms are located at transportation handoff points throughout the Woven City to encourage people to exercise.) Endo and Jeon share the tteokbokki and head out into the crisp, clear winter afternoon to go for a walk through Woven City,[10] which was built for just such strolls.
The 175-acre site near Tokyo was designed back in the early 2020s to put people, not cars, first — even though it was the car maker Toyota that financed the project. Streets in the city were split into three zones: one for zero-emission AVs; one as a promenade shared by pedestrians and slower forms of personal mobility, such as scooters; and one as a tree-filled park for pedestrians only. As much cargo traffic as possible was routed underground, so trips such as for grocery deliveries wouldn’t clog streets. The remaining traffic was designed to mesh seamlessly – hence the name Woven City.
“Toyota began as a loom manufacturer,” said Daisuke Toyoda, president of Toyota Motors Corp. “We didn’t begin by making cars. Now, we have woven all the threads, not just cars, into a beautiful design for a city of the future.
“Thirty years ago, when many of our colleagues in the automotive and technology sectors were focused on perfecting autonomous technology to reshape the car industry, my father [Akio Toyoda, who was himself president of Toyota] was more worried about how the technology would reshape society. That’s why he built the Woven City as a living laboratory with residents rather than building a test track for robot cars. The question of how to best serve society as a whole has guided our work ever since.”
Amal Kumar, an analyst with XYZ Research, agreed: “In the 2010s and 2020s, other companies like Waymo, General Motors, and Uber were focused on testing AVs in real-world environments, such as Phoenix, Arizona, and the interstate highways of Texas. Akio Toyoda recognized the real potential of AVs was in the context of everything connected to everything else. To understand the art of the possible, he decided to build a connected, sustainable citywide laboratory from scratch. Everyone else was trying to change the present; he wanted to live in the future.”
The Woven City’s entire infrastructure is below-ground, including its hydrogen power storage (with energy fed from all the solar panels on roofs and in windows in the city) and water filtration systems, so Endo and Jeon are free to enjoy the ambience as they take in a spectacular view of Mt. Fuji and appreciate the accents from the all-wood buildings. (Wood was chosen because it’s carbon-neutral and allows for artisans to showcase the traditional Japanese craft of wood-joinery, which is now augmented with robotic production techniques.)
"One of the biggest breakthroughs in the process was accepting that the city of the future doesn’t have to look futuristic but rather that the opposite should hold true," said Leon Rost, one of the original designers of the city. "Technology at its best is invisible and is secondary to human interaction and its connection to nature."
Endo and Jeon bump into some friends — Woven City was designed to promote interaction via “chance” meetings — and they chat for a while about how the city is about to be designated as a historic district, given its pioneering role in urban design.
As the sun starts to drop off the shoulder of Mt. Fuji, Jeon realizes it’s time to head back to Tokyo, so he summons an AV shuttle to take him to the train station. He’ll be back in his apartment in an hour, after having another personal mobility drone carry him from the Tokyo station to the roof of his apartment building. He’ll get some work done on a client presentation on the way.
Endo walks for a bit longer. She’s felt a bit claustrophobic in her apartment during the winter, and she’s enjoying the sunshine. Her health monitor has also been prodding her lately to walk more to get her blood pressure down. After 30 minutes, she walks over to the vehicular lane and catches a shuttle that drops her off in front of her apartment building.
She settles in on her couch. She considers ordering in for dinner but decides she’ll have some leftovers from lunch. That tteokbokki had turned out rather well.
End of Chapter
Other parts of this serialization (Subscribe to be notified of upcoming chapters as they are released):
A Brief History of a Perfect Future: Inventing the world we can proudly leave our kids by 2050 by Chunka Mui, Paul B. Carroll, and Tim Andrews
Part One: The Laws of Zero
Part Two: The Future Histories
Chapter 9 Transportation
Chapter 10 Health Care
Chapter 11 Climate
Chapter 12 Trust
Chapter 13 Government Services
Coda What is the Future Isn't Perfect?
Part Three: Jumpstarting the Future (Starting Now)
Chapter 14 What Individuals Can Do
Chapter 15 What Companies Can Do
Chapter 16 What Governments Can Do
Prologue: Over to You
Footnotes:
[3] The reinvention of the supply chain will resemble what happened when businesses reorganized itself around the container in recent decades, as described in this article in the New York Times: https://www.nytimes.com/2006/03/23/business/the-container-that-changed-the-world.html The idea of batches will fade, because transportation will be so cheap and so automated. There will still be a long-haul transportation network, but far less effort will need to go into conforming to it. On the front end, trucks and trains will be organized more flexibly — rather than having to fill a standard container or assemble a long train, shippers will be able to dispatch much smaller loads, as soon as they’re ready. On the back end, elaborate delivery routes will be less necessary; AVs and drones will be able to grab a delivery and take it straight to the recipient.?
[5] https://spectrum.ieee.org/cars-that-think/transportation/self-driving/the-electronic-highway-of-1969
[7] Thrun says he was motivated to work on autonomous driving because his best friend as a teenager died in a car accident — the sort of harrowing personal story that has focused many on the work and that AVs should almost eliminate in the Future Perfect.
[8] Urmson tells us he very much counted on the Laws of Zero in the early days of AVs. He incorporated all of what he considered to be the best technical approaches, including lidar, even though doing so meant that his early vehicles cost hundreds of thousands of dollars apiece. He was confident that the costs of all his electronics would head toward zero. Google’s approach caused a sort of arms race, pushing GM, Ford, and many others to also invest in the most robust technology possible, despite the expense. Others — notably, Tesla — relied solely on cameras and radar, which were already inexpensive and which could thus be used quickly to enhance safety. The devices Tesla used were also already miniaturized, so they wouldn’t interfere with its cars’ sleek look, as a flower-pot-like device on the roof would. Tesla’s approach, while effective in the short run, is now being exposed as timid: While Tesla claims self-driving capabilities, they don’t and can’t approach what lidar-based systems can do.
[9] The telephone is a good example of how a big enough quantitative change can produce a qualitative shift. The work that led to the phone began as a way to multiplex telegraph lines. Initially, a line could carry just one string of dots and dashes at a time, but techniques were developed that allowed for two signals at a time, then four, then eight… until enough of those magic doublings had happened that inventors like Bell could imagine having enough capacity to carry a voice signal. Et voila! iPhones for everyone!
Chief AI Engineer Global / US, PwC
2 个月As a big fan of the "Future of" series you have created, this one I struggle with. I see all the benefits for sure, but I do still love the feel of driving on an open road in the hills.. :) Maybe not good for environment, but sometimes good for my sanity?