Walt, while you were busy dodging snowstorms — and I heard they were pretty big, they were huge up north — you were relaxing in the sun. And while you were doing that, the autonomy markets kept on autonomously moving ahead. We had big news from Waymo, Wave, Nvidia, and Uber's identity crisis continues. And I wrote in this week in the Autonomy Economy, 'Will the real Uber please stand up? Please stand up.' See, I'm channeling my inner Eminem there. Let's kick things off with Uber and their autonomous solutions. What are they trying to be now?
Well, I think the drumbeat is strong because the stock is now kind of languishing in the low 70s, and clearly the overhang is the risk of autonomy in the future. So they're really trying to push harder on changing that narrative and addressing those concerns of investors. Let's start with this tome of a piece that they put out called announcing, I guess, Uber Autonomous Solutions. A lot of stuff that they have talked about before, and kind of the value that they believe that they provide, whether it's insurance programs and mapping data, fleet financing, whatever. I mean, I think we should go through quickly as we can this point by point. But yeah, it's, you know, just more kind of messaging to say that they do have a plan for the future autonomous world. And by the way, haven't heard them talk about autonomy taking multiple decades to scale. Looks like that little message point has been tucked away, you know, recently.
That message point went away like the McRib because reading this, it reads like a McDonald's menu. You want to start in breaking it down here one by one? I mean, the first thing they talk about is training data. Later in this tome of a piece, Wave, an autonomous company, gave a quote saying, 'Uber's multi-sensor and dash cam data across global markets adds critical diversity and complexity to our training, accelerating commercialization.' So we know Wave uses training. You know, how much do you think they're now relying on Uber? And it said, you know, adds critical diversity. Obviously, Wave has their own set of data. How important do you think training data is to autonomous companies?
It's not as important today as you would think it would be. You asked me 5, 10 years ago, it was very important. The data that these vehicles are gathering is so specialized. And the problem is when you're gathering dash cam data, I'm just going to say it, you have a lot of noise. You have a lot of noise and a lot of things that you don't necessarily need. So I don't think it's necessarily as important as it is today. The way these simulation models have grown and evolved, and if you look at Wave, the tours they've done, they've gathered quite a tremendous amount of data.
We'll have to do more work on Uber in terms of like, is it just dash cam data, or are they seeing, is there, you know, in some cases some of these drivers have specific data so we see how they're actually reacting, as opposed to Tesla, which I think has a lot of dash cam data themselves. I'm not sure how much they can pull from the individual drivers, but if this is their number one bullet point on what Uber Autonomous Solutions means, then I think, you know, you can't help but not ignore a lot of the dash cam data that maybe Tesla has.
Well, don't forget Tesla has — it's not really necessarily driver behavior. Tesla has all the miles, behavior of how the vehicles drive. They have more data than anybody. And we don't have any public stats on this, but it'd be fascinating to know what percentage of the data that Tesla gathers that they use to train the model and what percent they throw away. And I'm going to go out on a limb here. I'm going to say the number that they throw away is probably larger than most people think.
All right, let's try and get through these points. The next point that Uber talks about in terms of this Uber Autonomous Solutions is data-enriched mapping, basically saying that their dynamic geospatial data helps AV partners refine pickups. And I got to say that this is, I think, legit. We'll get later to some of the research notes that were put out there. I think Bellagi was pushing — the CFO of Uber was pushing a Jefferies note. Very small sample size, but talking about, you know, how the pickups of a human pickup versus a Waymo pickup is different. Although I will say, Grayson, when we were in Austin and we got a human pickup from the Uber driver, the guy picked us up in the middle of the street, and the Waymos, you know what? I always forget the name of that rib joint, but Terry Luke's or whatever it's called, but they picked us up in the parking lot. So in our situation, the pickup was better. With that said, I get it. You know, the pickups are going to have to be something that's improved, and Uber's claiming that for their autonomous partners, they can help out here.
The restaurant's Terry Black's, and it became even more famous this week. Secretary Kennedy was there cooking barbecue. So even more famous there is a video on X there of the secretary. I like the drop-off and pickups. When I co-chaired the autonomous vehicle task force city of Beverly Hills, one thing that we focused on was drop-off and pickups. And for companies that are trying to solve that, I will tell you there's a lot of publicly available free data in the GIS database, mappings of curbs. And it's really going to come down to the debate of the curb because at the end of the day, the city controls the curb, and it's going to come down to local city ordinances of how they zone the curb. If you take a city, say San Francisco for example or Miami for example, if they start designating drop-off and pickup zones, that becomes irrelevant. But today it is a very, very good point because the drop-offs and pickups are something that is going to have to be figured out long term.
I mean, the challenge here is for a Waymo pickup, you know, it's kind of like the stop signs with Tesla FSD. It's literally coming to a complete stop where it really shouldn't have to do that. You know, but in terms of the pickups, like a human driver, an Uber driver is obviously, even if it's an illegal spot, if they see the rider there, they're going to actually go and pick up the rider there, where the Waymo is going to go to a pre-approved spot. So I think, you know, that could just be an enforcement issue where it's going to be harder to enforce these Uber drivers as opposed to what Waymo can do. So maybe even the mapping is not going to be able to help that. And that will be, you know, if this is something that really bothers people, maybe that is an advantage that the human driver will have for some period of time until they can figure this stuff out.
Because here's a little incentive. When a city has a budget shortfall, oh, let's paint the curb all this color purple. Every time the Waymo pulls in, then the city gets a check. You watch that get enforced really quickly.
Yeah. The problem with that is like that would be great if Waymo was willing to pay it, but you know how people are, and the Uber driver will stop on that purple and there's no one there to enforce unless they put a camera and like bill the license plate. But then we're getting to true nanny state. The next bullet was venue and event management. I mean, I guess this is an issue, you know, dealing with the scale of cars, sending enough cars. Like if you're at a venue, how many Waymos can you send? So that's data that they're going to have regulatory support. I'm not sure how you pay for this when you're saying you're providing this as support. Is it like for every autonomous company that buys data from Uber, are they basically going to send their DC people in to make sure that they get approval, or is this just the, hey, we're doing this for the industry? And if that's the case, if let's say Uber's working on Governor Hochul to allow autonomy to work there, that's not going to just benefit a partner of Uber, that's going to benefit anyone in autonomy. It's not like they're going to cut individual deals, at least I would hope not, with these states. It's just going to open it up for autonomy in general.
It's a good thing on the surface. It's a very good thing that Uber wants to put dollars behind lobbying, dollars behind advocacy. That is a very, very good thing. The question is, it comes down to the devil's in the details. What is it? Is it representing the Uber corporation? Is it representing the interest of partner A? Is it representing partner B? Who is it representing when they go and do this? On the surface, very good. The question is all the details because you can't let one partner run away with an agenda while it hurts the rest of the industry. If it's truly bipartisan and it's truly, yes, we're all for the industry, it's great. But only time will tell because, as you know, sometimes individual companies have policy agendas and they want to push their particular policy agenda sometimes to the detriment of the industry, to the benefit of their company. So it's something that we're going to have to watch and see.
Yeah. And like, I mean, like it wouldn't shock me to be honest if Uber joined labor in preventing autonomy in some markets if this ultimately became a risk. So I don't know about the regulatory support, how they get monetized for it, and how that's unique for, you know, for autonomous solutions.
That line is what makes Walt Walt. That is a very interesting line, and in some ways, if Uber decides that autonomy is a risk, yeah, it could happen. Don't ever rule it out. I like that contrarian view there. Well done, Walt. You know what you're going to get for that? You're going to get a clappy hat for thinking outside the box there. Well done, Walt.
I think it's just my cynicism of spending way too much time in DC and seeing how companies act. Let's move on to the next one. It's high up on their bullet points here. Fleet financing. This is part of Uber Autonomous Solutions that they're pitching as the thing. This is just using the balance sheet. And like again, we all know that they're saying like long-term they're going to, you know, financialize REITs. This is stuff that we've talked about, I think even before the company has. It's because it's not rocket science, right? In the earliest days. We've talked about this, but you know, in the meantime, they're going to use the balance sheet, right? Yeah. And we agree with Uber that there will be a market for it. As there will be a market to finance a lot of the fleet management, but you know, basically saying you're going to pay for cars as part of Uber Autonomous Solutions. Like, okay, I'm not sure how that's truly differentiated.
Well, it goes back to Slim Shady. Do they want to be in the asset business or do they want to be in the asset-light business?
The next part of the Uber Autonomous Solutions is in-car experience. And you know, there's T-Mobile basically purchased an iPad that goes into the back of cars. That's part of in-car experience. It's also like having the music on, all the stuff that we know Uber and Lyft can do. But let me talk about in-car experience for a second. Grayson, there is no better in-car experience today in rideshare than Waymo. You get in a car, it drives exactly the same every single time, and there's no risk or safety or different type of driver. And I don't even need to mention it. Everyone knows the issues that go along with this. So like this one, like honestly, in-car experience. Now look, Nuro and WeRide both gave a quote that when they're talking about the value of Uber Autonomous Solutions, Nuro's quote was, 'Uber has done an excellent work designing an in-vehicle experience that will deliver the intuitive in-car interface riders expect.' That's from Chapin. I don't know. I have no idea what that means. I love the great, great team. No idea what that quote means. They've done an excellent work designing an in-vehicle experience. What is your in-vehicle experience in Uber? Like that you're getting ads served to you in your app, that you're, you know, the music can go on, it's a certain — you're telling the driver not to talk to you. Guess what? When there's no driver, you don't have to tell them not to talk to you. WeRide's quote was also, 'Uber Autonomous Solutions complements these capabilities by supporting demand generation, rider experience, and customer operations.' Again, very generic thing. So explain to me, Grayson, how as part of Uber Autonomous Solutions the in-car experience is improved with Uber.
It's not what you describe to me as buzzword alley. Insert your favorite buzzword to go. It's not the experience today. Waymo, you said they nailed it. But here's the other thing that's not talked a lot about. I mean, you guys at Lightshed talk about this. When you get into the Waymo, your Spotify is already synced. It knows where you picked off, and you get in there and you can blast some music and have this really great consistent experience. That doesn't take rocket science to figure out. I don't know what it truly unlocks. It's not like they're putting a team of Walt Disney Imagineers in there to figure something out. I don't know what it unlocks. Every consumer that you and I know, you want to get in the vehicle and you want your music to be synced or your audiobooks to be synced. You don't want to have to think. That's what it is at the end of the day.
Okay. So let's next bullet point again. What is Uber Autonomous Solutions? Complex use cases. I don't really know how they help here or how is this different than the other points, like complex use cases. I'm just going to move on to the next one and you can comment on the next two. End-to-end support, basically someone to answer the call button. Like, okay, we just determined — well, actually we didn't determine, we know that there's 70 remote drivers. That's different than when you call support at Waymo if you're in the car and then there's someone there. I mean, okay, hello AI, and this is exactly the market that AI is going to, which is customer support. So what do you have on complex use cases or end-to-end support?
The only thing I could think of — speechless. Pretty much. I was going to say John Legere, call John Legere, the former T-Mobile CEO. Let him figure it out. That's the only thing I could think of with that. I don't get it.
I mean, no, you don't call John Legere. You call Google. You call the, you know, because Google is who's helping the telcos with their AI and how to deal with end-to-end support, and it's not going to really include humans until that's — I mean, that's the last resort is really ever talking to a human for any of these companies. And frankly, you just, you don't want to have to deal with that end support. You want a consistent enough ride that you limit the number of calls altogether. Okay, next again, part of Uber Autonomous Solutions is AV mission control, and I think this is really the meat of it. I don't know why they buried it so deep in their points on this press release. It's fleet management. But this is like, this is the element of it. Like they want a fragmented market of people that own separate fleets and fleet management. So I can see how this is part of Uber Autonomous Solutions. The question is, do Uber partners actually want this?
Well, not only do the partners want it, but then where's Aomo? Where's Move? Where's Autocar? Where's all the other companies that were listed in the Q1 2026 supplement? Where are they? Is Uber now saying, 'Wait, we invested in you, but now we're going to do this in-house'?
I don't know if they're doing it in-house, and maybe they're tying it up, but they're basically saying if you run your own fleet, you know, and you want cars to process through, like, we're going to be the ones to do it. Look, we've seen it, like, you know, we've gone, I think, to more fleet management centers than anybody. And like, if they're working on mission control, you know, is Aomo using this mission control in terms of how they're managing their cars? I'm not sure how that works, but I also just don't think if, let's say, I'm a fleet owner that I necessarily want that additional lock-in with Uber. Maybe you do. Maybe it's turnkey. It's going to depend on cost. I don't know if that limits your ability to service vehicles on other platforms or, you know, let's say Waymo wants to — you want to service Waymo as well as Uber and you're the fleet owner that's plugging it into the Waymo One network or whatever, you know, unclear. But to me, with that said, this seems like the meat of what, you know, if I'm a fragmented autonomy partner, what could really Uber could really provide some benefit with.
Could. So you raised the vendor lock-in, that's a real deal. The question is what are the terms? And then looking at it from an industry standpoint, to me it says, okay, Uber wants to compete with Lyft's FlexDrive. They want to compete with Move. That's kind of how I'm seeing this. I think maybe Uber is starting to realize the value of fleet management. But I said this earlier, I'm going to go back to it again. Will the real Uber please stand up? Because I thought you want to be asset-light. Owning and operating a depot is not asset-light. You have to buy chargers, equipment. It's not asset-light.
But I think you're missing the point here. They aren't — this is asset-light. They are saying that the value is AV mission control. They're providing the technology to allow someone else to do it. So I think that is consistent with what they're saying in terms of capital intensity. They just want to be the software. So some other company invests the dollars in getting the power and cleaning the cars, maybe even owning the fleet. They're just providing the software on how to get those cars through the system. That's the point.
That software is available today. You and I or any individual want to — you can call Transave, you can call other companies and get off-the-shelf software with no vendor lock, and it's readily available.
I don't think that's true. I think those companies that you mentioned and others have, yes, bought off-the-shelf software, have developed it themselves. I think in some cases these companies are looking for a way to then sell it to other fleet managers, right? So they're really perfecting it because they live and breathe it every day. So the question is, as this market proliferates, operates, is someone going to rely on Uber for their fleet management software for this AV mission control, as they call it, a third party, or just build it on their own?
If you look at AI, look at Anthropic Opus 4.6 or Gemini 3.1 Pro. Why not build your own dashboard? You go on X, you can see the dashboards people are building. It's not hard to build a dashboard. So I don't really see the big competitive moat there.
You may be right, but again, when I think about what they've kind of laid out here on Uber Autonomous Solutions, this has the potential for having the most meat. Let's move on to the next bullet point in this release, which is remote assistance. I don't even know why this is a separate point. Like, isn't that the same as end-to-end support? I mean, I almost feel like I'm writing a research paper when I'm in fifth grade and I'm like adding additional paragraphs, which is just a replication of — like, how's that not end-to-end support? Remote assistance. We already talked about that. Let's move on to the last one. AV insurance, you know, as a thing. I don't know. I guess what they're saying is like Uber has negotiated all these insurance things for their drivers. I think already today some of their fleet partners have done a better job at securing their own insurance for their own drivers and their own fleets. So I'm not sure they need help with this. And I know that you've spent a lot of time talking to the insurance industry. Where do you rank this in terms of one of the opportunities that Uber has to kind of get lock-in through Uber Autonomous Solutions?
Of all the things we discussed, it's the most important one hands down because depending — we don't have any publicly available data, so this is all based on assumptions. Depending on the way the policy or if you want to say that the captive is built, Uber, because of their balance sheet and their financial resources, could go out, hire an Aon or hire a Marsh to go out to the market. Star, for example, Star Insurance has Waymo, and build a program where those rates can go dramatically down and then reinsure some of that risk and allow, let's call it — I'm going to call it Walt autonomy robo-taxis — to get the insurance that you wouldn't be able to get for a price that you can afford to do it. So depending on how it's structured, it's potentially a very, very big, very, very big disruptive move.
So after going through every one of these points, I am going to summarize this and say what is Uber Autonomous Solutions. It's okay, we have this training data to help your actual product work better. Seems like it's supplementary, not reliant. We got mapping to make the car actually drop you off at a pretty important thing. I think they got, you know, the in-car experience. Honestly, that's a throwaway in my view. The mission control, I think there's meat on this. You're a little bit more skeptical. How do you, as a fleet manager or as an autonomous company, manage all these things — AV insurance, and then lastly, I think financing. I think it's basically those are the major pie pieces in what autonomy is. This is stuff we assume that they were already giving to Nuro and WeRide anyway, right? They're just effectively formalizing this in the name. Do you think they're formalizing and putting out this release, then going on the podcast circuit again to try and alleviate some of the fears that have been emerging on Wall Street against their autonomy strategy?
I mean, obviously that's what we led with. The stock's languishing in the 70s. This is, you know, the primary concern. People aren't asking about how Uber Eats is going to perform next quarter. It's the overhang of autonomy. And in terms of going out on the circuit, you know, that brings us to our next segment, which is the new CFO, Bellagi, who we know, we've known for some period of time, took a group of analysts back out to Abu Dhabi, met with the CEO of WeRide, their partner there. They have a fleet partner in Tawasul, I think is how you pronounce it. I guess they're going to see BYD, you know, BYD there. So a lot of Chinese companies clearly, you know, not shy about pushing Chinese technology. Let's just start there again, going with WeRide and BYD. Do you think there's any risk that they're going to catch some flak back at home, you know, for again partnering with these companies in some of these markets and not focusing on how you're bringing autonomy to the United States and Western America?
I think there's going to be a tremendous blowback. I'm looking to see what happens next week. If all the speculation holds true and DeepSeek releases the R2 model, which was apparently — I used to use apparently not trade on Nvidia chips — that's going to obviously get a statement from the White House and possibly the Commerce Department. Then if you just look at politics, they're going to keep digging. And to me, it's a risk because it takes one tweet from — or sorry, one Truth post from the president to blow up their entire model about this. It's a very big risk, especially when you have a president and an administration that is very active on social media. And as you and I have seen, a post from the president or cabinet secretary, as we had today from the secretary of war, he hugged about Anthropic, it can move markets and redefine businesses. So it has the potential to light up like a powder keg. I mean, let's just look at the Senate hearing with Waymo and, you know, effectively admitting that some of these remote driver assistants, you know, were in the Philippines. And like, you know, and it's just you want to make things as easy as possible. There still are, like, we have the governor of New York that's keeping autonomy out of New York. Let's keep our hurdles to a minimum. It's obviously fine to have these relationships. That's an important part of the building blocks of their international strategy, but to be profiling it, I don't know. I just think there's some risk there.
Now, let's look at the meat of those meetings because Bellagi was tweeting out research reports. Hopefully no one ever tweets research reports out. These are meant for clients only. Anyway, they talked about WeRide, talked about tens of cars autonomously and a hundred, I think, with human assistance included. I don't know what that means with human. Does that mean there was someone actually driving the car? Is that in the safety seat? Does that mean that that was the remote driver? Like, and are we supposed to be like fired up about tens of cars? Because all the feedback we get about Tesla in Austin with tens of cars has not been what I would call positive from investors in general. But here tens of cars is we're supposed to be thrilled about, you know, for WeRide that I think last week we heard, you know, on some podcast is allegedly equal to where Waymo is in the market today.
I put the inspector hat on here because when we saw the multitude of press releases coming out of the UAE and Saudi Arabia, I said, 'Okay, let's take the inspector hat and let's go over to Saudi Arabia and to the UAE and let's find out what's going on there.' And so, prepping for my trip, which is not happening, and I'll get to the reason why it's not happening, I went through all the SEC filings. I went through the press releases. And you got to love stuff because when the inspector gets the magnifying glass out, I find stuff all the way at the bottom of the press in a light gray font: vehicles with safety drivers, both with the BYD RT6 and with the WeRides. And then so I dive further into it and I find out the zones that they're operating in. They're operating primarily in tourist zones. The interesting thing, and the reason why I'm not going, that part of the world is currently celebrating Ramadan and there is lighter traffic. So I find it very interesting that you're hosting analysts at a time in that part of the Middle East — to that part of the world, in the UAE — where there is less traffic at this time just because of their religion and their holidays. I find it very interesting.
I think that's completely coincidental. It's just you have to get earnings. It's when people have free time on their schedule. So I would put that aside. I don't think there's any issue there. But let's go on to the next thing, which is, you know, one thing I didn't see in any of these reports. Doesn't mean it wasn't addressed, but I think a critical thing on the heels of those great numbers that Waymo put out last week — 3,000 cars, I think it was 70 safety attendants. So we just talked about 100 with human assistance. In the tens of cars that are driven autonomously, what's the ratio of remote driver assistance and what is the role of those driver assistants? Do they have — are they driving, you know, with a wheel, or is it the way Waymo has it where it's just telling the car what to do and the car is under control? You have the CEO of WeRide. Those are probably the questions you want to ask them. Maybe they weren't answered. Didn't see it addressed. I obviously responded on Twitter. You know, maybe we'll get a response from Bellagi or someone else.
