Uber Wants to Turn Its Millions of Drivers Into a Data Factory for Self-Driving Cars

Welcome back to SavvyMonk, your one-stop for AI and tech news that actually matters.

Uber just revealed a plan that reframes everything we thought we knew about who wins in the autonomous vehicle race. Spoiler, it might not be the company with the most self-driving cars.

Let's get into it.

Late last year, a Klimt sold for the highest price ever paid for modern art at auction.

An outlier sure, but it wasn't a fluke. U.S. auction sales grew 23.1% in 2025. The $1-5mm segment even grew 40.8% YoY.

Meanwhile, Apollo’s chief economist Torsten Slock said to expect ‘zero in return in the S&P 500 over the coming decade.’

So, that's where it gets interesting.

Each environment is unique, but after dot-com, post war and contemporary art grew about 24% annually for a decade. After 2008, about 11% for 12 years.

It’s also had near-zero correlation with the S&P 500 since ‘95.*

Now, Masterworks lets you invest in shares of artworks featuring legends like Banksy, Basquiat, and Picasso.

Shares can sell quickly, but my subscribers can skip the waitlist:

Skip the waitlist

^{*Investing involves risk. Past performance is not indicative of future returns. See important Reg A disclosures at masterworks.com/cd.}

In 2018, Uber sold off its self-driving unit to Aurora under enormous pressure after a fatal crash in Arizona and mounting financial losses. At the time, many saw it as a strategic retreat. Co-founder Travis Kalanick has since called it a big mistake.

But Uber's current leadership appears to have reframed that exit entirely, and what they're building now looks less like a consolation prize and more like the winning position.

Praveen Neppalli Naga, Uber's Chief Technology Officer, revealed the new direction at TechCrunch's StrictlyVC event in San Francisco on May 1st. The plan, he explained, is a natural extension of a program called AV Labs that Uber quietly launched in late January.

The core idea is to eventually equip Uber's millions of human drivers' cars with sensor kits, turning the entire fleet into a distributed, real-world data collection machine for autonomous vehicle companies.

How It Works

Right now, AV Labs operates on a small, dedicated fleet of sensor-equipped cars that Uber owns and runs itself. That's the starting point. The long-term ambition is to roll that out to the broader driver network, which spans millions of vehicles globally, collecting the kind of messy, unscripted, real-world driving data that AV companies are desperate for.

The data feeds into what Naga calls an "AV cloud," a library of labeled sensor data that partner companies can query and use directly to train their models. A company could, in theory, request footage from a specific intersection in San Francisco at a specific time of day, and Uber's system could surface it. That level of specificity and scale is something no single AV company can replicate on its own today.

But Uber isn't stopping at data storage. Partners can also run their trained models in shadow mode against live Uber trips, simulating how an autonomous vehicle would have performed on a real route without putting a single self-driving car on the road. It's a way for AV companies to stress-test their software against actual traffic conditions before deploying anything physical.

Why It Matters

Naga put the core problem plainly at the event. "The bottleneck is data," he said. AV companies need enormous volumes of real-world scenarios, including rare edge cases, to train reliable models. A single robotaxi fleet might take years to encounter enough of these situations. Uber's network, completing millions of trips daily across dozens of cities, captures them constantly.

There's also a geographic argument. Self-driving companies like Waymo tend to concentrate their fleets in a handful of test cities. Uber already operates across hundreds of markets worldwide. That diversity of roads, weather conditions, traffic patterns, and driving behavior is something that can't easily be synthesized or replicated in simulation.

The regulations aren't figured out yet. Naga acknowledged that equipping third-party driver vehicles with sensors raises real questions about data-sharing rules that vary state by state. But the direction is clearly set.

The Current Partnership Landscape

Uber already has partnerships with 25 AV companies, including Wayve, which operates in London. The company has also committed $10 billion toward its robotaxi strategy, and has made equity investments in several AV players. Giving partner companies both proprietary training data and a marketplace to reach riders puts Uber at two critical leverage points in the AV ecosystem at once.

Naga framed the data ambition as non-commercial. "Our goal is not to make money out of this data," he said. "We want to democratize it." Whether that positioning holds as the asset grows more valuable remains to be seen.

The Bottom Line

Uber's play here is genuinely clever. It lost the race to build self-driving cars, so it's positioning itself to be indispensable to everyone who is still running that race. A dataset built from millions of daily trips, spanning hundreds of cities and every kind of road condition, is the kind of moat that takes years and billions to challenge.

The sensor grid isn't live yet, and the regulatory path is uncertain. But if Uber pulls this off, it won't need to own a single self-driving car to matter enormously in the AV industry.

Category: Competitive Strategy Analysis

"I run a company that was disrupted out of a market by a stronger competitor. Help me identify three strategic positions I could occupy in the same industry without competing directly. For each position, describe what assets I already have that could be leveraged, what new capabilities I'd need to build, and what the key risks are. My company is [Company Name] and the industry is [Industry]."

Uber's story is a useful reminder that losing a race isn't the same as losing the market. The company walked away from self-driving cars and is now positioning itself as infrastructure for the companies still building them. The most durable business positions are sometimes the ones nobody fights over because everyone needs them. Hit reply, I read every response.

See you in the next one.

— Vivek

P.S. Know someone who follows the AV or tech industry? Forward this their way. They can subscribe at https://savvymonk.beehiiv.com/