By David Galland, for The Rational Optimist Society
Table of Contents
The Great AV Disruption—Continued
The Remaining Challenges
Investment Opportunities
Conclusions
In Part 1 of this series, we set the stage for the disruptions to be caused by the widespread adoption of autonomous vehicles (AV)—or self-driving cars, as they are often called. If you are new to the party, you can read Part 1 here.
With thanks to the many ROS members who wrote in with comments and critical commentary about Part 1, we move on…
***
The Great AV Disruption—Part 2
Continuing our necessarily brief overview of the many sectors of the economy that will be impacted by autonomous cars, we must include…
Fueling and Parking Infrastructure
I still have reservations about electric cars and laugh at the notion they don’t pollute, given that the electricity they rely on comes from somewhere.
That said, the substantial investment and effort being directed toward electric vehicles and related technologies, particularly battery advancements, make it almost certain gasoline-powered vehicles will become a rarity in the foreseeable future.
Self-driving cars will greatly accelerate this transition for the simple reason that electric vehicles are mechanically far simpler, making them much easier to maintain.
And the entire supply chain and infrastructure related to recharging them is exponentially less complicated than for gasoline or diesel. As you don’t need me to tell you, those require trucking flammable liquids to gas stations around the country, where they need to be safely stored and regularly replenished.
As the number of EVs increases dramatically, significant new infrastructure will be required, including more and improved charging stations and specialized facilities for autonomous vehicles to park, be cleaned, recharged, and inspected before returning to service.
The current recharging bottleneck is likely to be resolved with some or all of the following innovations on the horizon:
Fast charging. There are already companies reporting they’ll soon be able to charge an electric car to drive over 200 miles in five minutes or less.
Super batteries. Imagine what a game-changer it will be when the technologists produce a battery able to power a car for 1,000 miles between charges. A lot of very well-funded companies are hard at work making that a reality.
Quick-swap battery packs. In the same way you might swap out the battery in your portable electric drill, companies—maybe repurposed gas stations—will offer a service to swap out your spent battery pack with a fully charged one in the time it takes the robot attendant to wash your windows.
Parking infrastructure will also be greatly impacted. Parking takes up around 30% of the available land in a typical city. Much of that will no longer be needed and can be freed up for more interesting purposes.
The city-dwellers who insist on owning their own autonomous cars—versus using ride-hailing services—will still not want to deal with parking issues.
That gives rise to new technologies— like the high-tech parking garages already found in places like China, Germany, Japan, and elsewhere—where your car will automatically deliver itself when you are done with it.
Energy Companies
Global revenues ~ $4 trillion
For the reasons mentioned, I think self-driving cars should all but eliminate the use of gasoline and diesel fuel for driving.
Even so, I don’t believe for a moment “Big Oil” is going out of business or even facing major disruption. While providing fuel for private cars, public transportation, and trucking may account for up to 30% of their current revenues, hydrocarbons are essential to nearly every tangible product we humans manufacture—from fertilizers and medicines to clothing and high-tech materials like carbon nanotubes, and much more.
I wouldn’t short oil companies based on the idea they’ll be acutely disadvantaged by the arrival of self-driving electric cars, but I’d probably keep an eye on them as the switchover begins.
The Used-Car Industry
Global revenues ~ $2 trillion
Widespread adoption won’t happen overnight. But in time, virtually every non-AV will need to be disposed of. Initially, many will be sold to developing countries lagging in AV adoption. But in time, a high percentage will be recycled into raw materials.
We can also expect companies to retrofit standard cars into self-driving ones. However, once the new generation of purpose-built self-driving cars arrives, designed to repurpose the driver’s space into a lounge-like environment for passengers, the demand for retrofitting is likely to decline rapidly.
Of course, used-car dealers will happily sell used self-driving cars as they become available. But the overall quantity of cars will fall precipitously, as will their revenue.
Automobile Repair Shops
Global revenues ~ $1.3 trillion
Should Morgan Stanley’s estimate of a 50% reduction in the number of cars on the road prove correct—coupled with 95% fewer accidents—a lot less revenue will go toward the hundreds of thousands of automobile repair shops around the world.
In addition, electric vehicles are mechanically so much simpler, which means no more oil changes or all the other ways that repair shops make money.
Of course, the savvy operators will pivot to repairing autonomous cars and, specifically, the technology that makes them tick. But the amount of work available will be significantly reduced, as will the number of employees.
Human Transport Companies
Global revenues ~ $10 trillion?
Given the number of people involved in transporting humans from here to there and back again, it is hard to discern just how big this industry is… but it’s big.
The list of pros to be derived from automating cars is long, from saving lives and preventing injuries to dramatic reductions in personal transportation costs and insurance. And let’s not overlook the time and efficiency savings as traffic jams and long commutes disappear.
The flip side is that hundreds of thousands, and maybe millions, of individuals currently making their livelihoods by driving will have to switch careers. History has shown that significant technological advancements such as Full Self-Driving (FSD) cars spawn a multitude of new businesses with their attendant employment opportunities, and we can only hope that is the case in this instance.
This is a topic we will address another day.
Naturally, taxi and ride-hailing companies will scramble to adapt to the disruptive tsunami of autonomous driving.
Recognizing the threat, both Uber and Lyft are moving aggressively to not be left behind, with Uber recently announcing a partnership allowing Waymo—for a share of the fees—to tap into Uber’s massive network of 116 million monthly app users.
Uber is also looking to create large service hubs where fleets of autonomous cars can be recharged, cleaned, and parked while waiting for the next call. Those hubs could become big business.
That said, Tesla is clearly focused on entering the ride-hailing space. One initiative, the Tesla Network, will allow Tesla owners to earn income by putting their FSD cars into a pool for the network. In addition, Tesla recently debuted its Robotaxi, a futuristic two-seater FSD.
According to Tesla, the Robotaxi will start operating in Austin this year and then expand from there.
In sum, while the current ride-hailing leaders, with their large installed rider base, provide some short-term advantages, they do not have a very deep moat to protect themselves once FSD taxis become ubiquitous.
Parcel Delivery Companies
Global revenues ~ $500 billion
Not having to pay for human drivers will prove a huge boost to the bottom line of parcel delivery companies.
That said, the packages still need to be delivered to the front door of customers. This is one of those problems that some smart person will ultimately come up with a solution for—maybe a robot, maybe automated receptacles that receive packages delivered by an “arm” extending from the delivery trucks? Maybe drone delivery services will become widely accepted?
Or, more likely, something we can’t even imagine now.
In the meantime, hiring lower-cost hourly workers to ride along in the trucks to handle the final delivery will provide much-needed jobs while still greatly reducing expenses.
Real Estate
US revenues ~ $2.4 trillion
In the same way the initial adoption of horseless carriages transformed the urban landscape, the widespread adoption of driverless cars will have a similar effect on both the countryside and cities.
Numerous studies have shown that the human mind is largely responsible for traffic congestion. Simply put, we humans are curious creatures, and the sight of a car in a breakdown lane invariably triggers an instinctive need to slow down and gawk, resulting in unnecessary traffic jams.
Now, imagine a world where cars flow smoothly, unhindered by the human mind.
A savvy real estate investor whose name escapes me once said that to be successful in projecting future development, imagine how far a city might expand in 10 years and then double that. With autonomous cars, that level of expansion is likely to be accelerated.
Of course, perhaps the entire notion of cities will change as well, given that—absent the congestion, clutter of parked cars, honking taxis, and fumes (I’m looking at you, NYC)—they could evolve into tranquil, pleasant places to live.
Meanwhile, back in suburbia, driveways will be turned into lawns, garages converted to hobby shops (or whatever floats your boat), and… and…
Simply put, there will be widespread ripples impacting where and how we live, opening up any number of new opportunities in real estate and housing.
Other Considerations
While emphasizing once again that the widespread adoption of autonomous cars will impact virtually every corner of society, here are a few additional areas likely to be affected.
Tire manufacturers. A more consistent—and generally lower—rate of speed will cause tires to last longer. In addition, given driverless cars can’t change their own tires, expect run-flat tire technology to only get better.
Car interior design. Without the need for a driver or a steering wheel, expect consumers—and car designers—to shift their focus from things such as horsepower and ease of steering, in favor of a complete rethink of the interior design of cars, with a premium paid for comfort, luxury, and versatility.
This is where the luxury carmakers have a chance to shine once they have adopted the FSD car technology.
Senior and disabled mobility. As it now stands, once a person reaches a certain age and can no longer drive—or shouldn’t—it is greatly restricting of their mobility. With automated car services, that issue vanishes.
A pilot program conducted by Valley Metro—in collaboration with Waymo and Arizona State University (ASU) between September 2019 and March 2020—aimed to assess the feasibility of integrating autonomous vehicles into Valley Metro's RideChoice program, which provides transportation services for seniors and individuals with disabilities in the Phoenix metropolitan area.
Despite older people being labeled as resistant to change, participants in the study reported a preference for automated cars over traditional taxis or rideshare options, citing increased safety and convenience.
Additionally, the participants in the trial program were much more likely to engage in activities outside the home, which any number of studies confirm has a very positive effect on mental and physical health.
Hospitals and emergency services. Every year in the United States, 1.3 million people are injured in car accidents, with about 40,000 fatalities. A 95% reduction in accidents means saving a lot of time and resources throughout the emergency response chain, from the police and ambulance drivers called to accident sites to emergency room doctors and staff, nurses, etc.
All of which translates into lower healthcare costs.
Road Infrastructure. While technologists are working to solve the challenges poor weather conditions create for self-driving cars (more on that in a bit), many of those challenges could be mitigated with changes to the road system.
For instance, AV-only lanes on major highways are likely to be introduced to reduce interactions with far less reliable human drivers.
High-contrast signs and road markings—perhaps with transmitters that work even in heavy snow or rain—would be helpful, as would “smart” sensors in traffic lights that communicate directly with the cars in the vicinity to determine the appropriate signal timing, greatly reducing congestion. China is already retooling roads specifically to support FSD cars.
Fortunes will be made implementing these highway improvements (much of it under the table to local bureaucrats in charge of awarding the contracts), but in the end, the roads will be just that much safer.
Restaurants. The whole notion of a DUI will go away, freeing people to head out to their favorite restaurant for a more relaxed evening without having to worry about that extra glass of wine. As liquor sales are a big contributor to the bottom line of restaurants, this will be a welcome boost. Of course, overall liquor sales should do better as well.
I am sure I have missed various pockets of society and businesses that will be impacted by AVs, but it’s time to turn our attention to the remaining challenges to be surmounted before we get to the Great AV Disruption.
The Remaining Challenges
Challenge 1: Weather
Perhaps the largest technological hurdle autonomous cars will need to clear is their ability to navigate in adverse meteorological conditions. Snow, ice, fog, heavy rain, dust storms, etc.
Notably, one of Tesla’s few fatal accidents came about when a car in FSD mode hit a pedestrian trying to direct traffic around a broken-down truck on the side of an Arizona highway. The cause was determined to be the angle of the sun and the resulting glare, which blinded the car’s cameras.
Waymo would have had an advantage here, as its LiDAR would have been unaffected by the glare.
Here’s an overview of how each of the two companies are trying to overcome weather challenges…
Tesla
Tesla’s FSD system includes a feature designed to detect adverse weather conditions. Previously, the system would produce an audible alarm to alert the driver that its functionality might be compromised. However, with the introduction of the FSD 13.2 update, the alert now appears only as a pop-up without the accompanying beep.
This change reflects Tesla’s growing confidence in its technology, as the lack of an audible alarm suggests it trusts the system to handle most weather challenges without causing unnecessary concern for the driver.
This increased confidence may also stem from the January 2023 upgrades to Tesla’s cameras, which now feature heating elements designed to prevent snow and ice from forming. These enhancements ensure the system can maintain safe operation even in challenging weather conditions.
Supporting that contention, there are many videos showing Tesla’s FSD handling bad weather conditions, including snow. While there will be instances where things go awry, the same can be presumed for human drivers in similar weather.
Here are some links to weather-related videos:
The reimplementation of radar—which doesn’t rely on vision—in the latest Tesla models will also help mitigate weather (and glare) issues.
Clearly, Tesla is moving toward true FSD versus the current compromise—a driver handover in the event of weather-related issues—but is not quite there. Even so, progress is being made, and its FSD is already operating well in most subpar conditions.
What About Waymo?
Like Tesla, Waymo’s working to refine the car’s suite of artificial intelligence (AI)-managed sensors to improve performance in bad weather.
While Waymo has some advantages—starting with an additional array of sophisticated sensors—it also has something of an Achilles heel in this regard… and maybe two.
On the plus side, Waymo’s cars are fully autonomous, i.e., no driver required. Tesla’s cars, while capable of operating autonomously in FSD mode, still require a driver to always remain ready to take over the controls.
Yet, because Teslas in their millions are sold and operated all over the US—including the northern states—the company is currently capturing massive quantities of bad-weather driving data, allowing it to rapidly adapt and improve its FSD function.
Waymo, in contrast, is currently operating only in sunnier climates and with a far smaller number of vehicles. The company will soon open operations in Atlanta and announced plans to open in Miami in 2026, but it will be another year of testing there before it’s able to do so. And, in both Atlanta and Miami, the most serious weather issue will normally be rain.
In addition, because Waymo is fully autonomous, it doesn’t have the option of handing control over to a human driver should sensors begin to fail. As a passenger in a snowstorm, that might concern you.
Currently, Waymo claims it can drive in snow but has not yet announced plans to expand to northern climes. Realistically, it will be at least two years—at the earliest—before it begins operating in snowy climates and benefits from data provided by driving in those conditions.
That gives Tesla a massive head start in terms of gathering data and refining its technology to handle poor conditions. While only time will tell, that head start may turn into an insurmountable competitive advantage for Tesla.
Given the importance of the weather safety issue, at the end of this Deep Dive, you will find an addendum with further details on the topic.
Challenge 2: The Human Factor
Taking advantage of the accident-avoidance features requiring FSD cars to automatically stop if they come to an obstruction in the road, there have already been several cases where poor examples of humanity have stood in front of a Waymo taxi, forcing it to stop so they could harass the occupants.
Then there’s the potential for someone to put up a roadblock or even just use cars to creatively block an autonomous truck full of high-value goods—say, the next shipment of iPhones—on a remote section of highway in the Midwest.
I don’t think these challenges are much of a worry because solutions will be found as these problems arise. Perhaps loud warnings broadcast from external speakers, emergency calls made to a specialist team back at HQ with access to all the vehicles’ cameras, and, with the push of a button, the local police.
Criminals don’t like loud alarms, or being filmed, or for the police to be called.
A much bigger challenge will be implementing FSD in chaotic traffic environments, such as those in places like Vietnam, which you can see here. (Thanks to member Frank W. for sending along the video I lifted the image from.)
Dealing with this sort of chaos will require a multi-pronged approach, including AI models trained specifically on chaotic environments with a more aggressive “negotiation strategy” for complex, unstructured traffic.
Step in front of an FSD car and, perhaps, after a toot on the horn, it will start inching steadily forward, making loud noises and sending film to the authorities. If you wish to be run over, keep blocking its way. Otherwise, stepping aside would be the smart move.
The solution for dealing with chaotic congestion will likely come out of China, where the government is not nearly so concerned about the occasional “failure” when FSD cars meet reckless pedestrians or bicyclists. In fact, Baidu’s Apollo Go service has already logged millions of driverless miles in dense cities like Wuhan and Chongqing.
This is, again, one of those “solutions in the waiting” that will present itself in due course.
Challenge 3: Hacking
Hacking is a frequent concern brought up by AV skeptics. Could someone hack your car, turn it into a speeding weapon, and use it to plow into a crowd?
Or could the truck full of iPhones be electronically hijacked and guided to a remote barn to be relieved of its cargo?
Absolutely. However, today’s leading AV companies—Tesla and Waymo (owned by Google)—are among the largest, best-financed, and savviest technology companies in the world. If they can’t hack-proof their vehicles, no one can. So, I think this is an unnecessary worry.
Challenge 4: Government Regulations
Of course, before AVs are released into the wild, the regulators will need to sign off.
Already, 25 states have adopted regulations related to AVs. For example, Florida has enacted laws allowing fully autonomous vehicles to operate without a human driver present, provided they meet federal safety standards and can achieve a minimal-risk condition in case of system failure.
Ultimately, it will come down to federal agencies—the Department of Transportation (DOT) and the National Highway Traffic Safety Administration (NHTSA)—to approve federal safety standards that all AVs will have to meet to operate without a driver.
Under the Biden Administration, Congress considered various proposals for comprehensive AV legislation, e.g., the SELF DRIVE Act and the AV START Act, but as of this writing, none have been enacted.
This lack of federal guidance leaves significant regulatory gaps that states are filling on their own, leading to a patchwork of rules and holding back progress.
That could all soon change, given the business and innovation themes underlining the new Trump administration and the president’s close association with Tesla’s Elon Musk.
The odds are good this regulatory morass will be dealt with in fairly short order, which could very well be the green light for the Great AV Disruption.
Investment Opportunities
It is not in the purview of The Rational Optimist Society to offer investment tips, but in the case of AVs, the most likely winner of the prize—and it will be a very big prize—remains Tesla.
In addition to its self-driving cars, Tesla shareholders also own the company’s battery and charging solutions divisions, as well as the Optimus robot technology that’s garnering a lot of attention these days.
There is another interesting source of potentially game-changing revenue for Tesla. Earlier, I commented that car manufacturers that fail to quickly adapt to AVs—or fail to impress when they do so—face an existential threat.
The fact is that, outside of China, none of the major car manufacturers are remotely close to Tesla in terms of self-driving technology and the data that supports it. But what if Tesla decided to spin out its car manufacturing business into a separate company and then license its self-driving technology to other car companies?
Were it to do so, the other car companies could get back to business designing and manufacturing cool (self-driving) automobiles, while Tesla’s new AV tech division provides the AV equipment, software, and—most important—the regular data-driven updates. It would enshrine Tesla as the FSD industry leader.
Is Musk Lying?
Before continuing, it is worth asking whether Tesla’s ambitions are realistic… fantasies created by Elon Musk… or some combination of the two? There’s no question Musk is prone to bouts of over-enthusiasm—some would say dishonest hype—when it comes to promoting his various enterprises.
After all, it’s a matter of record that in 2019, Musk predicted the company would have 1 million robotaxis on the road before the end of 2020. Five years later, there are still none. At a recent event showcasing the Robotaxi, he claimed they will be on the road before 2027, then added the caveat, “I tend to be a little optimistic with time frames.”
And when he talks about selling tens of millions of Optimus Robots a year, starting in just a couple of years, one might want to look for a handy saltshaker.
Other critics point to Tesla’s data skeptically, opining the company is nowhere near achieving true FSD capabilities.
So, what to make of all this? For starters, much of the criticism I have come across appears to be out of date or unsupported by reliable data. For example, the article linked to just above uses crowdsourced data—a tiny sample size which, while informative, can’t be taken as hard proof.
In my opinion, Musk has proven himself to be a once-in-a-generation visionary, nearly single-handedly creating the EV sector, revolutionizing space launches—and, in the process, jump-starting the space economy—advancing brain-computer interfaces, developing machines for cost-effective tunneling, and now pioneering humanoid robots.
While his critics have credible points to be considered, most notably his well-documented issues with punctuality, I wouldn’t believe the harsher accusations of fraud leveled against Tesla’s FSD. It is coming; just don’t accept Musk’s forecast on timing as a sign to invest right away.
Though Tesla is currently valued at over $1 trillion dollars, looking at its financial metrics, it is not outrageously expensive. That said, before investing, you might want to set up Google keyword alert related to Tesla and FSD and, over the next year, monitor the technological and regulatory progress the company is making before taking a chance with a small investment.
In time, Tesla may need to add additional sensors—perhaps due to regulatory requirements—but, ultimately, whatever it takes will be done because the FSD movement is now unstoppable.
Other investment opportunities are embedded in the list of disruptions above. Those point toward sectors to avoid—much like video rental companies when streaming was just over the horizon—and sectors where you might find an opportunity.
And there will be a lot of opportunities.
Parting Thoughts
It is no longer a question of whether autonomous vehicles will become the norm but when.
The technology has already progressed beyond proof-of-concept to practical implementation.
The fact that some insurance companies are already offering low-priced policies for autonomous cars is an indicator the “powers that be” are actively preparing for their arrival.
Which brings us to the question: “How long?”
Specifically, how quickly will the regulators allow both Tesla and Waymo to operate in FSD mode all the time throughout the United States?
At what point are the steering wheels and pedals no longer offered in new cars, thereafter, to be relegated to history?
Even with a favorable wind in the sails provided by the Trump administration, widespread adoption won’t happen overnight but rather will likely follow distinct patterns in different regions and use cases.
Already leading the transition are southern states, where the weather challenges are minimal, as well as dense urban areas with high ride-sharing demand.
Long-haul trucking routes through sparsely populated areas represent another early adoption scenario, where the operational benefits are clear and the driving environment is more predictable.
Of course, the road to widespread adoption will not be without potholes, and we can expect there to be widely publicized failures resulting in the occasional fatality.
While these incidents will receive a lot of dramatic media coverage—because that’s how the media companies butter their bread—they will also serve as learning opportunities for the AV technologists.
But in the end, when you compare the butcher’s bill, it is clear cars driven by humans—with their 40,000-plus fatalities and million-plus injuries each year—are far, far more dangerous than AVs.
And the safety advantages of AVs, which are already approaching 95% over humans, will only get better from here.
Adding appeal to the AV movement is that, in the United States, transportation is the third-largest budget item for the average family, after housing and taxes. Transportation expenses—including car maintenance, high insurance premiums, the cost of fuel, and maintenance—will be greatly reduced.
Most pundits I reviewed predict full adoption will take 7–10 years, and they may be right. However, given that most such projections are based on current technology and computational power, they tend to be overly pessimistic.
That is especially true given the shift in mood in Washington, DC.
How will you know the Great AV Disruption is upon us?
It will be when asked, “Would you allow your children or grandchildren to ride to school in an autonomous car?”
… and you answer with a resounding “YES!”
Until next time…
David Galland
For The Rational Optimist Society
Would you ride in an Autonomous Vehicle? Will you stop driving yourself once they are widely available? All comments and thoughts are welcome… you can share them directly with me at Galland@rationaloptimistsociety.com.
Addendum: Q&A on the Bad Weather Challenge
1. Has Tesla sufficiently proven its camera-based technology works in bad weather?
I would argue yes. From what I can tell, while it has hiccups, its technology appears to be improving. Regardless, its vehicles are still safer than cars driven by humans in the same conditions.
2. Does Waymo’s business model trump Tesla’s when it comes to improving its FSD in bad weather?
Emphatically no. Waymo requires areas to be mapped beforehand to operate; Tesla does not. Waymo cannot begin operating in bad weather conditions without having “solved” the weather problem because, unlike Tesla, it cannot give control back to the driver. In the interim, Tesla’s data gathering and resultant improvements are only going to accelerate.
3. Is Tesla in trouble if its cameras are not as good as Waymo’s technology?
Maybe. But so far, the data shows it’s more than holding its own. However, should that begin to change, Tesla can adapt and retrofit additional sensors to all current vehicles or just offer them on new cars moving forward.
4. Is Waymo screwed if Tesla’s FSD tech continues to be proven safer than its tech?
Not absolutely, but likely. Waymo will still need to continue mapping each area it moves to and develop its tech to work better in those areas, while Tesla can steadily improve with nationwide software updates.
Once Tesla gets to “true” FSD and is permitted to unleash its robotaxis, Waymo has no real competitive advantage. This fact may have also contributed to General Motor’s decision to abandon its FSD ambitions.
5. Have either Tesla or Waymo solved bad weather conditions while driving?
No. But Tesla can currently operate in bad weather conditions even if it has issues that may require manual takeover. Waymo will soon operate in rainy conditions, but Tesla can already do that for the most part—and Waymo still has a lot of work to do before it can begin serious testing in those conditions, let alone snow and ice.
6. Finally, is it a certainty Tesla and/or Waymo will fully solve the weather problem?
Yes. I have faith in their technical capabilities and the growing power of AI. And, in the case of Tesla, probably sooner rather than later.
Sources and Resources: