The Robotaxi Apocalypse Is a Lie

 

They say robotaxis will kill millions of driving jobs. The numbers tell another story: the great self-driving apocalypse may be Silicon Valley’s most expensive scare campaign.

I keep hearing that self-driving cars are coming for everybody. First, they will take over the streets. Then they will wipe out taxis. Then Uber and Lyft drivers will join the unemployment line. After that, apparently, Americans will surrender their steering wheels, sell their cars, climb into robotaxis, and happily place their lives in the hands of computers.

I have only one problem with this beautiful science-fiction funeral. I am not convinced the corpse is dead.

I don’t get the hype about self-driving cars. Yes, the technology is impressive. A car drives itself. Cameras watch the road. Radar and lidar scan the surroundings. Computers make driving decisions. The steering wheel moves without human hands.

Fine.

But after I finish admiring the circus trick, I still have a simple question.

What exactly has been revolutionized for the passenger? If I need to travel 10 miles across town, a self-driving car does not magically turn those 10 miles into 5. It sits in traffic. It waits at red lights. It gets trapped behind construction. It cannot tell rush hour, “Excuse me, I am artificial intelligence. Please move.”

A robotaxi stuck in traffic is still stuck in traffic.

That is my first problem with the prophecy.

My second problem is price. Self-driving technology does not automatically guarantee that owning or using a vehicle becomes dramatically cheaper. Somebody still has to manufacture the vehicle, insure it, maintain it, clean it, charge or fuel it, repair it, replace tires, operate the fleet, maintain software and pay for the enormously complicated technology wrapped around the machine. Removing the driver removes one cost. It does not remove economics.

My third problem is availability. A conventional automobile can operate almost anywhere there is a legal road and a capable driver. Fully driverless commercial service remains geographically limited. A technological revolution that works beautifully inside selected operating areas is impressive, but it is not the same thing as replacing hundreds of millions of ordinary vehicles across cities, suburbs, mountains, rural roads and miserable weather.

And then there is safety. Yes, self-driving cars are safe. Waymo reported that through March 2026 its vehicles had traveled 220.6 million rider-only miles. Its data showed substantially lower crash rates than human-driver benchmarks, including an 81.93% reduction in crashes involving reported injuries and a 94.05% reduction in crashes involving serious injury or worse. Earlier research covering 56.7 million miles also reported major reductions in injury crashes involving pedestrians, cyclists and motorcyclists. Those are serious numbers, and I am not going to hide them under the carpet because they inconvenience my argument.

But those numbers prove that autonomous driving can become safer in certain operating environments. They do not prove that the privately driven automobile is preparing for its funeral.

That distinction is where the robotaxi apocalypse starts losing air.

Real technological disruption usually has a weapon. Not a slogan. Not a TED Talk. Not a billionaire standing beside a futuristic machine. A weapon.

The weapon is overwhelming advantage. The Internet had it. Consider the letter. I could write something in Baltimore, put it in an envelope, buy postage, send it across the country and wait days for delivery. Or I could press “send” on an email and deliver the message in seconds.

Game over.

That was not a 10% improvement. It was a slaughter.

The evidence is sitting in America’s mailbox. U.S. First-Class Mail volume reached about 103.7 billion pieces in 2001. By 2025, it had collapsed to about 42.2 billion. The Internet did not ask traditional mail for permission. It walked into the room, pulled out speed and convenience, and started shooting.

Look at shopping.

Before e-commerce, I had to drive to a store, find parking, walk inside, search shelves, stand in line, pay and drive home. The Internet said, “Why?”

Suddenly I could sit on my couch at 11:47 p.m., wearing pajamas that should probably be investigated by federal authorities, order what I wanted and have it delivered to my door. In the 4th quarter of 1999, e-commerce represented only 0.7% of U.S. retail sales. The transformation that followed was enormous because online shopping removed friction from the transaction.

Banking got the same beating.

Once upon a time, paying bills meant writing checks, stuffing envelopes, buying stamps and trusting the mail. The Federal Reserve estimated 42.6 billion check payments in 2000. By 2021, that number had fallen to 11.2 billion. Meanwhile, electronic payments exploded. Debit-card payments alone rose from 8.3 billion in 2000 to 87.8 billion in 2021.

Why?

Because the new method was dramatically easier.

The Internet did the same thing to newspapers. I no longer had to wait until morning to discover what happened at midnight. News became immediate.

It did it to maps. I no longer had to unfold a paper map across the passenger seat like Christopher Columbus looking for Ohio. Digital navigation could tell me where I was, where I needed to go, which road was congested and when I would arrive.

It did it to entertainment. I no longer had to drive to a video store, wander through aisles, discover that the movie I wanted was gone and return another day. Streaming brought huge libraries directly into the living room.

It did it to travel booking. Instead of calling agents and waiting while somebody searched schedules, consumers could compare flights, hotels and prices online themselves.

It did it to communication. A business meeting that once required airline tickets, hotels, taxis and 2 wasted travel days could suddenly happen through an Internet video call.

That is disruption.

The old way says, “This takes time, money and effort.”

The new way says, “Not anymore.”

Now bring the self-driving car into the interrogation room.

“Can you get me across town dramatically faster?”

“No.”

“Can you eliminate traffic?”

“No.”

“Can you eliminate parking problems if I own you?”

“No.”

“Can you teleport me?”

“Don’t be ridiculous.”

“Then why exactly are you going to destroy the automobile as we know it?”

Silence.

That does not mean autonomous vehicles are useless. Far from it. They could become enormously important. They may improve mobility for elderly people, people with disabilities and people who cannot drive. Robotaxis could reduce labor costs for transportation companies. Autonomous systems could reduce certain kinds of crashes. They may transform particular transportation markets.

Waymo is already demonstrating that this is not laboratory fantasy. The company said it delivered more than 14 million trips during 2025 alone. That is real commercial activity, not a PowerPoint presentation.

But replacing something is a much higher bar than competing with it. We have seen this movie before.

Television did not kill radio.

Airplanes did not eliminate trains.

Cars did not eliminate bicycles.

E-books did not eliminate printed books.

Online shopping did not eliminate physical stores.

The smartphone absorbed cameras, maps, calculators, music players and dozens of other functions because putting those functions into 1 pocket-sized device created a massive convenience advantage. The self-driving car offers something narrower: I ride in a car without personally operating it.

That is useful.

It is not teleportation.

And the unemployment prophecy looks especially shaky when I examine the government’s own numbers. If taxi and ride-hailing drivers are standing on the edge of technological extinction, somebody forgot to tell the U.S. Bureau of Labor Statistics. The BLS counted about 204,000 taxi drivers, including ride-hailing drivers, in 2024. Instead of projecting their employment to collapse, it projects employment to reach about 226,600 in 2034—a gain of 22,600 jobs, or 11%. For taxi drivers, shuttle drivers and chauffeurs combined, employment is projected to rise from 447,900 in 2024 to 486,800 in 2034. The BLS expects about 58,800 openings per year across those occupations.

Read that again.

While headlines are preparing the undertaker, the federal employment projections are ordering more chairs.

Could those forecasts be wrong? Absolutely. Forecasting technology 10 years into the future is dangerous business. History is full of confident predictions that eventually became expensive comedy.

But that is precisely my point.

We keep turning possibilities into certainties.

“Robotaxis exist” becomes “robotaxis will dominate.”

“Autonomous vehicles are improving” becomes “nobody will drive.”

“Some driving jobs can be automated” becomes “millions of drivers are doomed.”

Slow down.

Technology does destroy jobs. Anyone denying that is denying history. Telephone automation devastated the old world of manual switchboard work. Digital photography crushed much of the film-processing business. ATMs changed bank employment. E-commerce hammered certain retailers. Online travel booking changed travel agencies. Machines have been stealing particular tasks from human beings since the Industrial Revolution.

But technology also creates jobs, changes jobs and sometimes expands the market it supposedly came to destroy.

The smartphone did not merely kill products. It created app developers, mobile marketers, social-media managers, rideshare drivers, content creators and businesses that barely existed before.

Even Uber itself is a delicious piece of irony. Technology was once supposed to eliminate traditional transportation jobs through efficiency. Instead, smartphone technology created an enormous new category of app-based driving work.

Now another technology is supposedly coming to kill the workers created by the previous technology.

Capitalism has a sense of humor.

Will some drivers lose work to autonomous vehicles?

I think so.

In carefully mapped urban markets where autonomous fleets can operate efficiently, routine taxi trips are obvious targets. If a company can safely operate a vehicle for long hours without paying a driver, the economics will eventually become tempting. Pretending otherwise would be foolish.

But “some displacement” is not “driver-driven cars are finished.”

Those are completely different claims.

The ordinary automobile possesses an advantage futurists sometimes underestimate: humans already know how to use it, the infrastructure already exists, it operates almost everywhere, people own hundreds of millions of them, and many people actually like driving.

The robotaxi therefore does not merely have to work. It has to make the old system sufficiently inferior that millions of consumers willingly abandon it.

That is what the Internet did to many old technologies.

Email looked at postal speed and laughed.

Streaming looked at scheduled television and laughed.

Digital banking looked at the paper check and laughed.

Online shopping looked at closing hours and laughed.

GPS looked at the folding road map and laughed.

Video conferencing looked at flying 2,000 miles for a 1-hour meeting and laughed.

What does the self-driving car look at when it sees my regular car? Mostly the steering wheel. That is the weakness buried beneath the hype. Self-driving technology may change transportation. It may capture substantial market share. It may eventually become safer than human driving across many environments. It may reduce the number of professional drivers needed in certain cities. It may even become ordinary enough that our grandchildren wonder why humans were once trusted to control 4,000-pound machines while angry, sleepy, drunk, texting and eating hamburgers.

But that future does not automatically mean the death of human-driven cars. Technology does not win because it is futuristic. It wins because it solves an old problem so brutally well that going backward starts to feel stupid.

The Internet did that repeatedly. Self-driving cars have not crossed that line yet.

So when somebody tells me that Uber drivers should start packing their bags because robotaxis are coming, I am not running to organize the funeral. I am looking at the road. The robot is coming. The human driver is still there. And, according to the current employment numbers, he may be there for quite a while.

For now, the robotaxi apocalypse has encountered the oldest enemy known to transportation.

Traffic.

 

On a different but equally important note, readers who enjoy thoughtful analysis may also find the titles in my  “Brief Book Series” worth exploring. You can also read them here on Google Play, or in Barnes & Noble bookstore: Brief Book Series.