Uber and Nebius to Invest Up to $375 Million in Avride
Waymo rides are forecast to rise dramatically, Tesla is trying to deceive investors into thinking it has San Francisco Robotaxis, and Nexar Unveils BADAS.
Top Stories of the Week
Uber, Nebius to Invest Up to $375 Million in Robotaxi Firm Avride (link, no paywall). The planned deal includes strategic funding and commercial commitments aimed at helping Avride scale its driverless fleet to as many as 500 vehicles, according to a company spokesperson. If certain milestones are reached, both Uber and Nebius could invest additional capital — bringing the total potential funding to $375 million.
This deal builds on a multi-year strategic agreement between Uber and Avride inked in 2024. And while Avride does not yet have a fully driverless fleet, they plan to launch a robotaxi service on Uber’s platform in Dallas by the end of 2025. In the meantime, Avride is already operating delivery robots via Uber Eats in Jersey City, Austin and Dallas.
Avride is quietly becoming one of the top AV players to watch in the United States. But while the headline number — up to $375 million — sounds impressive, it would have been nice to see more clarity around the milestones Avride needs to hit before that money actually lands. Still, AV hype is generally good for Uber’s stock price, and it is in Uber’s best interest not just to partner with promising AV startups, but to also invest in them directly. A fragmented AV market benefits Uber — keeping multiple operators on its platform instead of relying too heavily on one. And with the Waymo-Uber relationship cooling off (Waymo just announced its seventh new market without Uber), deals like this are more important than ever.
Related: Grubhub and self-driving startup Avride launch robot food delivery pilot in Jersey City (link, no paywall).
Waymo weekly ride projections per Bloomberg Intelligence (link). This chart was making the rounds this week on social media, but I wasn’t able to find any methodology on how they arrived at their predictions. So I did some additional calculations / fact checking:
If I put my investor hat on for a minute, Waymo is still in the early stages of rapid growth. They’ve found PMF, and are starting to scale their main distribution channels (Seed-Series A equivalent), so 10-20% MOM growth would be strong. From 11/29/24 (175,000 rides) to 4/24/25 (250,000 rides), they were at an 8.57% MOM growth rate. So the methodology in this chart appears reasonable. I’m not sure why there’s a jump in growth rate in 2027 but I do think 8-10% MOM growth over the next few years is doable. And then 4-8% in the following years as the business matures/growth slows.
Vehicle supply seems to be the real limiting factor to Waymo’s growth at the moment though. The Jaguar I-Pace is out of production and the mounting tariffs on Chinese EVs are bad news for the Zeekr. Waymo doesn’t have any other vehicles production ready so I’m not sure what the plan is here.
Based on my calculations1, I estimate that Waymo would have about ⅓ market share compared to Uber and Lyft in their top 20 US markets by 2030 at this growth rate. On the one hand, that’s significant, on the other hand, it’s not even a plurality after 7 years of operating. Of course, there are a number of competitors that could be in play by then too.
Other Stuff
Waymo to manually test autonomous vehicles at Newark airport (link, no paywall).
Nuro Appoints Co-Founder Dave Ferguson as Co-CEO to Accelerate Commercial Growth (link).
Grab invests in May Mobility as robotaxi startup expands to Southeast Asia (link, no paywall).
Tesla is trying to deceive investors into thinking it has San Francisco Robotaxis (link). I don’t normally like to pile on Tesla but this article raises a few interesting points. Tesla stated that it “launched ride-hailing service in the Bay Area using Robotaxi technology” but CFO Vaibhav Taneja also said “in the Bay Area, where we still have someone in the driver’s seat due to regulations, we crossed more than a million miles”. The former may technically be true but in practice, the current product is the same as having Uber drivers with Tesla FSD ferrying passengers around. Not exactly something you want to tell investors though.
‘Research report diving into why AVs are not scaling faster on their own (i.e. outside Uber’s network). Fixed supply providers suffer from a loss in utilization as they scale beyond a certain level - starkly in contrast to flexible supply ridersharing networks. So the choice is to scale to a small level of multiple cities or lose a lot of money per city to scale higher’ (link). I know Uber likes to make the case that ridehail networks (both Uber and Lyft really) are the best avenue for AV companies to increase their utilization but one idea I’ve been thinking about is building reduced utilization into your business model. Waymo has shown that there is lots of demand for their services during peak times. But instead of leasing expensive real estate in the city and building out ultra-expensive L3 charging stations, it might make more sense to travel 30-45 minutes outside of the city, where real estate is cheaper and charge at L2 speeds (30-40 mph). You can still easily get a full charge on 8-10 hours and the costs savings could outweigh the increase in utilization and the revenue sharing that would come from having to partner with an Uber or Lyft.
Nexar Unveils BADAS: Breakthrough AI Model Sets New Safety Benchmark for AV and Driver Assistance Systems (link). This is neat and I love the name!
Regulators probe Waymo after its robotaxi drove around a stopped school bus (link).
Uber and WeRide launch robotaxi service in Riyadh (link).
Zoox begins offering robotaxi rides to select San Francisco passengers (link). Passengers being taken off the company’s service waitlist are friends and family of Zoox employees, who will be asked to share feedback on their robotaxi experience. The rides are currently free.
Kodiak Self-Driving Tech Compared To Humans In Nauto Safety Study (link, no paywall).
Why D.C. is stalling on self-driving cars (link).
The Electric: A Chinese AI Expert On What It Will Take For Cars To Go Completely Autonomous (link, no paywall).
Skit: If Waymos could talk (link). Lol, this is pretty good.
What else we’re reading/listening to
Inside Self-Driving: The AI-Driven Evolution of Autonomous Vehicles (link). Great webinar hosted by Business Insider’s Steve Russolillo, featuring speakers like Rivian’s Head of Autonomy, James Philbin, and Johann Jungwirth, EVP of Autonomous Vehicles at Mobileye. They dig into how recent advances in AI and computer vision are helping autonomy gain real-world traction, among many other insights.
The Pragmatic Engineer by
: Google’s engineering culture (link).Autonomous System Safety by
: Robotaxi Scaling Is Just Beginning (link).Dwarkesh Podcast by
: Andrej Karpathy — AGI is still a decade away (link). You can jump to the 01:42:55 timestamp where they discuss the topic “Why self driving took so long”.ICYMI, we’ve released a couple of recordings from our Urban Autonomy Summit — presented by Nexar at Newlab in Brooklyn on October 8 — now available on The Driverless Digest podcast.
AVs’ Coming Impact on Existing Fleets & Driver Earnings (link).
Preparing Fleets and Streets for Responsible Operations (link).
Cool Rides
‘Using a Waymo is very liberating. As a woman, late at night I feel safer with a car than with a human behind the wheel’ (link).
‘I finally had the chance to put Waymo to the test in San Francisco, and it was an absolute incredible experience’ (link).
‘During my week in the US, I had the chance to ride Waymo several times. I’d tried it before, but it still amazes me every single time’ (link).
AVs/Humans behaving badly
Waymo gets confused in construction zone, goes into oncoming lane and immediately corrects (link).
Zoox blocking traffic in SF (link). May have been due to the cardboard box on the ground?
When we ask Waymo to be more aggressive, this is not what we mean (link). I kind of like this move tbh, maybe Waymo is testing my ‘haul ass mode’?
Neat Jobs
Senior Product Manager, Fleet Optimization at Waymo (link) via Vishay Nihalani.
Director, Business Development - Autonomous + Rideshare at Terawatt (link) via Peter Cohen.
Compute Workload Specialist at Waymo (link) via Xiangyu Dong.
Senior/Staff Technical Product Manager, Mapping Tools at Zoox (link) via Clem Besson.
2026 Intern, PhD, Machine Learning Engineer, Simulation at Waymo (link) via Sixu Piao.
AV/Mobility Events
TaskUs Forward ’25 (Oct 29, San Francisco) — “The Next Turn for AV” panel will feature executives from Waymo, Uber, and Serve Robotics (link).
Women in Automotive Technology (Nov 5, San Francisco) — The Road Ahead: Navigating the New Frontiers of Autonomous Transportation with Insurance and Risk Solutions (link).
Until next week.
-Harry
I used actual ridehail data from NYC and Chicago, and extrapolated that using population data to the top 20 cities in the US in order to come up with Uber and Lyft’s current market share.
The bit about the undefined milestones Avride needs to hit for the full funding really caught my attention. How do you rekon they'll define 'fully driverless' for those targets by 2025? It's always great to read your insights on these complex tech investments.
The milestone-based funding structure here is quite clever - it protects both Uber and Nebius while giving Avride room to prove their techology. What stands out is how Nebius is positioning itself not just as cloud infrastructure provider, but as a strategic partner in the AV ecosystem. The 500-vehicle target is ambitious but achivable given the multi-year timeline. Interesting to see how this compares to Waymo's current deployment strategy.