Tesla and Waymo are Chasing the Robotaxi Dream - but the Company Spending the Most to Win Builds no Cars at All

Uber is quietly writing $500 million checks to lock in robotaxis as Waymo threatens to leave it behind

Tesla's autonomous robotaxis use the company's Model Y vehicles. The promised Cybercab is still in development.

Tesla investors are paying a premium for a robotaxi fleet of just 20 cars

If you own Tesla stock, much of what you are paying for above the value of a carmaker is a bet on autonomy and artificial intelligence that has barely reached the income statement: full self-driving software, the Optimus robot and a robotaxi network.

The robotaxi is the nearest-term and most testable piece of that bet, and this spring it amounted to about 20 driverless Tesla Model Y vehicles in Austin, Dallas and Houston. Value the car business the way investors price any other automaker, and it accounts for only a fraction of the stock; the rest is the market's bid on that future, a premium no ordinary carmaker could carry. What is new is that the bet is finally testable against operating data rather than projections.

This is not only a Tesla $(TSLA)$ story. Autonomous driving has been separating investors from their money for years - from the self-driving Google cars promised to the public by 2017 to the roughly $10 billion that General Motors $(GM)$ poured into Cruise before shutting it down.

What is different this time is that the money has moved from research budgets to balance sheets, with contracts, milestones and deployment targets attached, across Tesla, Uber Technologies (UBER) and Alphabet $(GOOGL)$ $(GOOG)$ alike. Each company's progress, or lack of it, is now something investors can finally mark to market.

Tesla's bet is real, but still small

Tesla's robotaxi bet is a vertical one, with a single company owning the vehicle, the software and the network. The hardware is no longer theoretical: CEO Elon Musk said Cybercab production had begun at Giga Texas in April, though he warned that output would follow a "stretched out S-curve" - his term for a manufacturing ramp that stays slow far longer than usual before turning sharply higher, with little volume expected until late in the year. Musk has cautioned that material Cybercab revenue is unlikely before 2027.

Tesla's shareholder update in late 2025 listed Phoenix, Las Vegas and three Florida cities - Miami, Orlando and Tampa - as part of its planned first-half 2026 Robotaxi coverage. But with that window nearing its end, those five markets have still not moved into service. In its first-quarter update, Tesla no longer attached the first-half timing to those cities and instead described them as "Preparations Underway," while spring reporting pointed to a rollout that was slipping rather than accelerating.

The spread between Waymo's half-million weekly rides and Tesla's roughly 20 active unsupervised robotaxi vehicles is, in effect, what the market is being asked to absorb every time it assigns Tesla a robotaxi premium.

Uber in the fast lane

If Tesla is the bet on owning everything, Uber is the bet on owning almost nothing and renting demand to whoever builds the cars.

If Tesla is the bet on owning everything, Uber is the bet on owning almost nothing and renting demand to whoever builds the cars. That model is getting more expensive to defend. On June 3, Reuters reported that Uber has committed close to $500 million to self-driving software startup Nuro, far more than previously disclosed. The publicly known piece was Uber's participation in a $203 million funding round that valued Nuro at $6 billion. On top of that, Uber quietly made a second, larger investment and agreed to release further capital as Nuro cleared defined technical and commercial benchmarks.

The structure matters as much as the headline number. Nuro has already hit its first milestones on schedule, unlocking an initial tranche of performance-linked capital. The rest is gated behind driverless testing later this year, paid passenger pickups before year's end and expansion through 2027.

The Nuro commitment sits inside a three-way arrangement with electric-vehicle maker Lucid Group $(LCID)$ to deploy 35,000 autonomous taxis built on Lucid Gravity SUVs and future midsize models, running Nuro's autonomy stack, and dispatched through Uber's network. Uber has separately invested $500 million in Lucid, and Saudi Arabia's Public Investment Fund added another $550 million in convertible preferred stock to support the program.

Add the rest of the ledger and the pattern is unmistakable. Uber has committed up to $1.25 billion to Rivian Automotive (RIVN) through 2031, tied to milestones, for as many as 50,000 autonomous versions of Rivian's R2, with the first commercial rollouts planned for San Francisco and Miami in 2028. Uber has partnerships with Baidu $(BIDU)$, Wayve and Waymo, as well as a Munich robotaxi project with Nvidia (NVDA) and Israeli AI startup Autobrains - Uber's first autonomous ride-hailing venture in Europe. Taken together, Uber's committed and contingent autonomy capital now exceeds $2 billion.

That is the tension investors should be pricing. Uber's pitch, repeated in its 2026 proxy statement, is that autonomy amplifies what it already owns: tens of millions of daily trips and decades of real-time trip-matching data, with the AV spending framed as disciplined capital allocation.

The counter-argument is that a genuinely capital-light platform would not need to keep writing nine-figure checks to carmakers and software developers to secure supply. For Uber shareholders, the real risk is less a weak quarter than a re-rating: The market awards asset-light platforms a far richer multiple than it gives capital-heavy fleet owners, and every check Uber writes nudges it toward the second group. Buying equity stakes rather than owning depreciating fleets softens that, but the line between "demand layer" and "co-investor in hardware" thins with each deal.

Waymo in the way

Waymo is targeting 1 million weekly rides by year's end.

The reason Uber is paying up is visible in the numbers of Alphabet's Waymo. The company crossed 500,000 paid rides per week in March, a tenfold increase in less than two years, across 10 U.S. metro areas after adding limited public service in Dallas, Houston, San Antonio and Orlando in February. It raised $16 billion at a $126 billion valuation, and it has kept opening markets since: It added Nashville in April as its 11th U.S. city and took Miami and Orlando fully public the same month. A planned 2026 London launch, where road testing is already underway, would be its first international commercial market, with Tokyo still in early-stage testing.

Waymo is targeting 1 million weekly rides by year's end. None of this moves Alphabet's stock on its own, since Waymo is still a rounding error beside search and cloud. Its weight here is competitive rather than financial: It sets the pace that every rival, Uber included, now has to match.

Here is the uncomfortable detail for the Uber thesis: Waymo rides are available exclusively through the Uber app in Atlanta and Austin, but Waymo's four-city expansion announcement in February made no mention of Uber at all. The largest supplier of autonomous capacity in America is increasingly going direct through its own app. If the best fleets can generate their own demand in dense markets, Uber's aggregation value shrinks to secondary cities and overflow capacity. That disintermediation risk is precisely why Uber is paying to lock in supply from Nuro, Lucid, Rivian and Wayve: It needs credible alternatives to the one operator that may not need it.

One trade, many tickers

For investors, the checklist for the next 12 months is concrete.

What used to be a venture-capital story now runs through public markets at every layer of the stack. Nvidia is a Nuro backer, supplies the Munich project, and its DRIVE Hyperion 10 platform is becoming the reference design for automakers building autonomy on someone else's computing.

That makes Nvidia the closest thing to a neutral wager on the category: It sells into Tesla's rivals, Uber's partners and the Chinese operators alike and collects whether or not any single network wins. Chinese lidar maker Hesai $(HSAI)$, Hyperion 10's primary lidar partner, just posted the industry's first full year of GAAP profitability, holds more than 40% of the long-range automotive lidar market and raised its 2026 shipment outlook to 3 million to 3.5 million units, with robotaxi orders from Pony.ai (HK:2026) $(PONY)$, WeRide, (HK:800) Baidu's Apollo Go and DiDi $(DIDIY)$. Among the listed names it is the most direct bet on robotaxi-sensor demand, and the rare one already turning a profit on it.

Lucid and Rivian, for their part, get guaranteed fleet demand for cars they would struggle to sell at scale to consumers. Amazon.com's (AMZN) Zoox rounds out the full-stack operators.

For investors, the checklist for the next 12 months is concrete. Watch whether Nuro begins driverless testing on schedule later this year and carries paying passengers before year's end, since both events release more Uber capital and validate the licensing model. Watch whether Waymo's march toward 1 million weekly rides continues to bypass Uber in new cities.

And watch unit economics. Waymo's Jaguar-based robotaxis are estimated to cost around $150,000 each, while Tesla is targeting a price below $30,000 for the Cybercab - a figure it has yet to prove in production. That gap, together with falling sensor costs, will help determine whether the lasting margin sits with the vehicle maker, the autonomy provider or the ride-hailing network.

For all the attention on Musk and Tesla, the company spending the most to win robotaxis is the one that builds no cars at all. Uber is writing half-a-billion-dollar checks so that whenever the technology finally scales, it is standing at the register.

Jurica Dujmovic is a MarketWatch columnist. Follow him on X @JuricaDujmovic.

-Jurica Dujmovic

MW Tesla and Waymo are chasing the robotaxi dream - but the company spending the most to win builds no cars at all

By Jurica Dujmovic

Uber is quietly writing $500 million checks to lock in robotaxis as Waymo threatens to leave it behind

Tesla's autonomous robotaxis use the company's Model Y vehicles. The promised Cybercab is still in development.

Tesla investors are paying a premium for a robotaxi fleet of just 20 cars

If you own Tesla stock, much of what you are paying for above the value of a carmaker is a bet on autonomy and artificial intelligence that has barely reached the income statement: full self-driving software, the Optimus robot and a robotaxi network.

The robotaxi is the nearest-term and most testable piece of that bet, and this spring it amounted to about 20 driverless Tesla Model Y vehicles in Austin, Dallas and Houston. Value the car business the way investors price any other automaker, and it accounts for only a fraction of the stock; the rest is the market's bid on that future, a premium no ordinary carmaker could carry. What is new is that the bet is finally testable against operating data rather than projections.

This is not only a Tesla (TSLA) story. Autonomous driving has been separating investors from their money for years - from the self-driving Google cars promised to the public by 2017 to the roughly $10 billion that General Motors (GM) poured into Cruise before shutting it down.

What is different this time is that the money has moved from research budgets to balance sheets, with contracts, milestones and deployment targets attached, across Tesla, Uber Technologies (UBER) and Alphabet (GOOGL) (GOOG) alike. Each company's progress, or lack of it, is now something investors can finally mark to market.

Tesla's bet is real, but still small

Tesla's robotaxi bet is a vertical one, with a single company owning the vehicle, the software and the network. The hardware is no longer theoretical: CEO Elon Musk said Cybercab production had begun at Giga Texas in April, though he warned that output would follow a "stretched out S-curve" - his term for a manufacturing ramp that stays slow far longer than usual before turning sharply higher, with little volume expected until late in the year. Musk has cautioned that material Cybercab revenue is unlikely before 2027.

Tesla's shareholder update in late 2025 listed Phoenix, Las Vegas and three Florida cities - Miami, Orlando and Tampa - as part of its planned first-half 2026 Robotaxi coverage. But with that window nearing its end, those five markets have still not moved into service. In its first-quarter update, Tesla no longer attached the first-half timing to those cities and instead described them as "Preparations Underway," while spring reporting pointed to a rollout that was slipping rather than accelerating.

The spread between Waymo's half-million weekly rides and Tesla's roughly 20 active unsupervised robotaxi vehicles is, in effect, what the market is being asked to absorb every time it assigns Tesla a robotaxi premium.

Uber in the fast lane

If Tesla is the bet on owning everything, Uber is the bet on owning almost nothing and renting demand to whoever builds the cars.

If Tesla is the bet on owning everything, Uber is the bet on owning almost nothing and renting demand to whoever builds the cars. That model is getting more expensive to defend. On June 3, Reuters reported that Uber has committed close to $500 million to self-driving software startup Nuro, far more than previously disclosed. The publicly known piece was Uber's participation in a $203 million funding round that valued Nuro at $6 billion. On top of that, Uber quietly made a second, larger investment and agreed to release further capital as Nuro cleared defined technical and commercial benchmarks.

The structure matters as much as the headline number. Nuro has already hit its first milestones on schedule, unlocking an initial tranche of performance-linked capital. The rest is gated behind driverless testing later this year, paid passenger pickups before year's end and expansion through 2027.

The Nuro commitment sits inside a three-way arrangement with electric-vehicle maker Lucid Group (LCID) to deploy 35,000 autonomous taxis built on Lucid Gravity SUVs and future midsize models, running Nuro's autonomy stack, and dispatched through Uber's network. Uber has separately invested $500 million in Lucid, and Saudi Arabia's Public Investment Fund added another $550 million in convertible preferred stock to support the program.

Add the rest of the ledger and the pattern is unmistakable. Uber has committed up to $1.25 billion to Rivian Automotive (RIVN) through 2031, tied to milestones, for as many as 50,000 autonomous versions of Rivian's R2, with the first commercial rollouts planned for San Francisco and Miami in 2028. Uber has partnerships with Baidu (BIDU), Wayve and Waymo, as well as a Munich robotaxi project with Nvidia (NVDA) and Israeli AI startup Autobrains - Uber's first autonomous ride-hailing venture in Europe. Taken together, Uber's committed and contingent autonomy capital now exceeds $2 billion.

That is the tension investors should be pricing. Uber's pitch, repeated in its 2026 proxy statement, is that autonomy amplifies what it already owns: tens of millions of daily trips and decades of real-time trip-matching data, with the AV spending framed as disciplined capital allocation.

The counter-argument is that a genuinely capital-light platform would not need to keep writing nine-figure checks to carmakers and software developers to secure supply. For Uber shareholders, the real risk is less a weak quarter than a re-rating: The market awards asset-light platforms a far richer multiple than it gives capital-heavy fleet owners, and every check Uber writes nudges it toward the second group. Buying equity stakes rather than owning depreciating fleets softens that, but the line between "demand layer" and "co-investor in hardware" thins with each deal.

Waymo in the way

Waymo is targeting 1 million weekly rides by year's end.

The reason Uber is paying up is visible in the numbers of Alphabet's Waymo. The company crossed 500,000 paid rides per week in March, a tenfold increase in less than two years, across 10 U.S. metro areas after adding limited public service in Dallas, Houston, San Antonio and Orlando in February. It raised $16 billion at a $126 billion valuation, and it has kept opening markets since: It added Nashville in April as its 11th U.S. city and took Miami and Orlando fully public the same month. A planned 2026 London launch, where road testing is already underway, would be its first international commercial market, with Tokyo still in early-stage testing.

Waymo is targeting 1 million weekly rides by year's end. None of this moves Alphabet's stock on its own, since Waymo is still a rounding error beside search and cloud. Its weight here is competitive rather than financial: It sets the pace that every rival, Uber included, now has to match.

Here is the uncomfortable detail for the Uber thesis: Waymo rides are available exclusively through the Uber app in Atlanta and Austin, but Waymo's four-city expansion announcement in February made no mention of Uber at all. The largest supplier of autonomous capacity in America is increasingly going direct through its own app. If the best fleets can generate their own demand in dense markets, Uber's aggregation value shrinks to secondary cities and overflow capacity. That disintermediation risk is precisely why Uber is paying to lock in supply from Nuro, Lucid, Rivian and Wayve: It needs credible alternatives to the one operator that may not need it.

One trade, many tickers

For investors, the checklist for the next 12 months is concrete.

What used to be a venture-capital story now runs through public markets at every layer of the stack. Nvidia is a Nuro backer, supplies the Munich project, and its DRIVE Hyperion 10 platform is becoming the reference design for automakers building autonomy on someone else's computing.

That makes Nvidia the closest thing to a neutral wager on the category: It sells into Tesla's rivals, Uber's partners and the Chinese operators alike and collects whether or not any single network wins. Chinese lidar maker Hesai (HSAI), Hyperion 10's primary lidar partner, just posted the industry's first full year of GAAP profitability, holds more than 40% of the long-range automotive lidar market and raised its 2026 shipment outlook to 3 million to 3.5 million units, with robotaxi orders from Pony.ai (HK:2026) (PONY), WeRide, (HK:800) Baidu's Apollo Go and DiDi (DIDIY). Among the listed names it is the most direct bet on robotaxi-sensor demand, and the rare one already turning a profit on it.

Lucid and Rivian, for their part, get guaranteed fleet demand for cars they would struggle to sell at scale to consumers. Amazon.com's (AMZN) Zoox rounds out the full-stack operators.

For investors, the checklist for the next 12 months is concrete. Watch whether Nuro begins driverless testing on schedule later this year and carries paying passengers before year's end, since both events release more Uber capital and validate the licensing model. Watch whether Waymo's march toward 1 million weekly rides continues to bypass Uber in new cities.

And watch unit economics. Waymo's Jaguar-based robotaxis are estimated to cost around $150,000 each, while Tesla is targeting a price below $30,000 for the Cybercab - a figure it has yet to prove in production. That gap, together with falling sensor costs, will help determine whether the lasting margin sits with the vehicle maker, the autonomy provider or the ride-hailing network.

For all the attention on Musk and Tesla, the company spending the most to win robotaxis is the one that builds no cars at all. Uber is writing half-a-billion-dollar checks so that whenever the technology finally scales, it is standing at the register.

Jurica Dujmovic is a MarketWatch columnist. Follow him on X @JuricaDujmovic.

-Jurica Dujmovic

(MORE TO FOLLOW) Dow Jones Newswires

June 24, 2026 07:50 ET (11:50 GMT)

MW Tesla and Waymo are chasing the robotaxi dream -2-

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

June 24, 2026 07:50 ET (11:50 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.