Uber Just Ordered 10,000 Driverless Taxis from Rivian. The Dream of Automated Cars Was First Sold to the Public at the 1939 World's Fair.

Key Takeaways

• •Uber's 10,000-vehicle order from Rivian is the largest single order of purpose-built autonomous vehicles in history
• •The vehicles are purpose-built EVs designed for autonomous ride-hailing, with no steering wheel or pedals
• •GM's 1939 Futurama exhibit was the first mass presentation of the concept of self-driving cars to the public
• •Waymo currently operates in San Francisco, Phoenix, Los Angeles, and Austin with fully driverless vehicles
• •Uber's plan targets deployment in 10 US cities by the end of 2028

Timeline

Uber has placed an order for 10,000 autonomous electric vehicles from Rivian.

Read that sentence again, because the magnitude is easy to miss in the noise of a news cycle that moves at the speed of social media. This is not a memorandum of understanding. It is not a "strategic partnership" press release designed to boost both companies' stock prices. It is a purchase order. Ten thousand vehicles. Purpose-built for autonomous ride-hailing. No steering wheel. No pedals. No human driver. Delivery beginning in late 2027.

It is the largest single order of driverless taxis in history.

Ten thousand vehicles. Not concept sketches. Not pilot programs. Not a press release with 'plans to explore.' Ten thousand actual vehicles, purpose-built, with delivery dates.

— ROOT•BYTE analysis

The vehicles will be built on Rivian's R2 platform, modified with a comprehensive autonomous driving sensor suite including lidar, radar, cameras, and ultrasonic sensors. Uber is not developing the self-driving software itself; it is partnering with Aurora Innovation, whose Aurora Driver system has been in commercial testing on freight trucks in Texas since 2024. The Uber-Rivian-Aurora triad represents the most ambitious commercialization attempt in the autonomous vehicle industry to date.

Uber's plan calls for initial deployment in Miami, Dallas, and Atlanta in late 2027, expanding to seven additional cities by the end of 2028. If the timeline holds, there will be autonomous Uber vehicles operating without human drivers in ten major American cities within three years.

This is real. But the road to this moment has been extraordinarily long. Longer, in fact, than most people realize.

On April 30, 1939, the New York World's Fair opened in Flushing Meadows, Queens. The theme was "The World of Tomorrow." The most popular exhibit, by an overwhelming margin, was General Motors' Futurama.

Designed by the visionary industrial designer Norman Bel Geddes, Futurama was a massive, 35,738-square-foot scale model of what the United States might look like in 1960. Visitors sat in moving chairs that glided over the diorama while a narrator described the world below. The centerpiece was a network of automated express highways where teardrop-shaped cars moved at 100 miles per hour, guided by radio control, spaced with mathematical precision, and entirely free of human error.

No traffic jams. No accidents. No road rage. The cars drove themselves.

In 1939, visitors to Futurama were given a pin that said 'I Have Seen the Future.' Eighty-seven years later, the future they saw is finally ordering its first fleet.

— ROOT•BYTE analysis

Twenty-eight million people rode through Futurama during the fair's two-year run, making it the most attended exhibit. As visitors exited, they were given a small blue-and-white pin that read: "I Have Seen the Future."

The future, it turned out, was running late.

GM's vision of automated highways by 1960 did not materialize. Nor did it arrive in 1970, or 1980, or 1990. The technical challenges were far greater than Bel Geddes had imagined. A car that drives itself needs to see the world, understand what it sees, and make decisions in real time. In 1939, the concept of computer vision did not exist. Digital computers did not exist. The transistor would not be invented for another eight years.

The first serious attempts at autonomous vehicles began in the 1960s. Stanford's Cart, built in 1961, was a small wheeled platform with a camera that could navigate around obstacles using early computer vision algorithms. It moved at roughly one meter per second, pausing for fifteen minutes between moves to process what it saw. It was proof of concept, nothing more.

Progress was glacial for decades. The real breakthrough came from an unexpected source: the United States military. In 2004, DARPA, the Pentagon's research agency, announced the Grand Challenge: a $1 million prize for the first autonomous vehicle to complete a 142-mile course through the Mojave Desert. Fifteen vehicles started. None finished. The best performer, from Carnegie Mellon, made it 7.3 miles before getting stuck.

The following year, DARPA doubled the prize and shortened the course. This time, five vehicles finished. The winner was Stanley, a modified Volkswagen Touareg built by a Stanford team led by Sebastian Thrun. Stanley completed the 132-mile course in six hours and 53 minutes.

Thrun later joined Google, where he led the company's self-driving car project starting in 2009. That project became Waymo, which is currently the only company operating a fully driverless commercial taxi service in the United States, running in San Francisco, Phoenix, Los Angeles, and Austin.

But Waymo's approach has been cautious and capital-intensive. Each Waymo vehicle costs hundreds of thousands of dollars. The service operates in geofenced areas that expand slowly. Waymo has been at this for seventeen years and serves a tiny fraction of the ride-hailing market.

Uber's approach is different. Rather than building and operating its own fleet, Uber is leveraging its existing platform, its 150 million monthly active users, its routing algorithms, its demand prediction models, and adding autonomous vehicles as a new supply type alongside human drivers. The Rivian vehicles will coexist with human-driven cars on the Uber platform. Riders in supported cities will be offered the option of a driverless ride, likely at a lower price point.

The economics are compelling. The single largest cost in ride-hailing is the driver. An autonomous vehicle eliminates driver wages, which typically account for 60-70% of the fare. Even with the higher capital cost of the vehicle and the technology stack, Uber projects that autonomous rides will be 30-40% cheaper than human-driven rides within two years of deployment.

There are enormous open questions. Regulatory approval in each city is required and not guaranteed. Liability in the event of an accident remains legally complex. Public trust is fragile; a single high-profile accident could set the industry back years. The Aurora Driver system, while extensively tested, has not operated at the scale of 10,000 vehicles in dense urban environments.

But the order is placed. The vehicles are being built. The future that was promised in 1939, the future that 28 million people were told they had seen, is finally being manufactured.

Eighty-seven years late. But arriving.

(Sources: Uber Technologies investor presentation, Rivian Automotive SEC filing, Aurora Innovation technical blog, Norman Bel Geddes Papers at Harry Ransom Center, DARPA Grand Challenge records, Waymo safety reports)

Enjoy This Article?

RootByte is 100% independent - no paywalls, no corporate sponsors. Your support helps fund education, therapy for special needs kids, and keeps the research going.

How did this make you feel?

Recommended Gear

Keep Reading

Mobility Tech

The Electric Car Was Invented in 1832. Then Oil Killed It for 100 Years.

9 min read

Mobility Tech

Ride-Sharing Was Invented During World War I, Not by Uber

5 min read

Mobility Tech

Electric Cars Existed 50 Years Before Gasoline Ones, Then the World Forgot

7 min read