Can Uber’s Robotaxis Scale?

Uber has a Waymo hangover as its stock slides and near-term autonomous vehicle scale remains limited, with the $10 billion AV commitment expected to take years to materialize. Zoox launches paid rides in Las Vegas but faces NHTSA speed restrictions above 45 mph, limited stops, long wait times, and only 105 vehicles across up to seven cities.

The Uber-Nuro-Lucid robotaxi program is unlikely to achieve paid commercial service in California by year-end due to regulatory hurdles, though a pivot to Texas remains a possibility. Internationally, Wayve secures a TFL for-hire license in London and OMEGA data signals potential new Waymo market filings, while California Teamsters sue the DMV to block autonomous trucking regulations.

Key Autonomy Markets Episode Questions Answered

Autonomy Markets Topics & Timestamps

[0:00] Uber’s Waymo Hangover

Uber’s Q2 revenue comes in light as management finally owns the $10 billion autonomy spend, and the stock sells back into the sixties under an autonomy overhang.

[4:52] Zoox Prepares to Go Paid in Vegas

After 12 years, Zoox begins charging on August 10, but an NHTSA filing surfaced by OMEGA caps the vehicles at 45 mph while traffic out of Harry Reid International averages 55.

[12:19] Waymo Returns to Freeways

Waymo resumes freeway driving after a two-month pause, a giant airport unlock as OMEGA picks up new Waymo corporate filings around the world.

[14:13] Uber, Nuro, Lucid Robotaxi Timeline

Lucid burns over a billion dollars in the quarter but declares the Uber-Nuro robotaxi program a must-win project, with regulatory hurdles pointing to free rides in California and a potential Houston pivot.

[20:30] DoorDash and the Market for Autonomous Food Delivery

Autonomous delivery cuts deadhead miles, a quiet regulatory advantage, while tipping fatigue and big box retail shape the demand side.

[26:08] NVIDIA Alpamayo 2

The open model hits 500,000 Hugging Face downloads in a brilliant CUDA lock-in play, but no SAE Level 4 robotaxi runs full stack NVIDIA, leaving Qualcomm as the sleeper.

[32:22] The Teamsters sue the California DMV over Driverless Trucking Rules

The lawsuit in Alameda Superior Court signals California is closed for business as a federal preemption bill moves through the House.

[35:50] Tesla’s China Business Question

Reports of executives preparing to separate the China business ahead of a SpaceX merger draw a fake news label from Elon, but the government business logic is hard to ignore.

[39:00] Foreign Autonomy Desk

Wayve lands a TfL license for 15 vehicles with safety drivers, but the APS permit controls every relationship in London, and Baidu holding one creates a clear off-ramp from Uber and FreeNow by Lyft.

[45:35] Next Week

Zoox first paid rides land August 10, with WeRide, Lyft, and Kodiak earnings on deck ahead of field trips to Dallas, London, and Vegas.

Full Episode Transcript

Uber’s Q2 Earnings and the $10 Billion AV Commitment

Grayson Brulte: Walt, we got a busy show today, but like any good party, you get a hangover. And Uber has a Waymo hangover. While Uber tries to get over their Waymo hangover, Zoox is getting paid. And while Zoox get paid, the teams are trying to body slam autonomous trucking in the Golden State. Let’s start with Uber’s Waymo hangover. What did we learn on the call this week?

Walter Piecyk: Well, you know, they did talk about, and there was a lot of questions about autonomy as you would expect. I mean, there was– On the revenue side, things were a little light, not because of bookings. It’s, you know, kind of a take rate issue, so not- nothing that we could say that, “Hey, they were light because autonomy’s having an impact.” Remember, still early days, not a lot of cars in their markets to have financial impact on Uber now. But certainly if you’re, if if you’re a little light on revenue now before autonomy’s really hit, it’s not, not necessarily ideal. What was interesting is they, they did comment on this $10 billion number. You remember that was the Financial Times piece. So I know that created a hubbub at the time, but I think they’re finally just owning this and saying, “Yeah, obviously we’re gonna be spending a lot of money.” Remember, listeners, this is a com- this 10 billion is a combination partly of investments in these companies, and I think some of them might need some more investments in the coming year, but also just in the commitment in the number of cars that they’re gonna purchase. And so that, and that, that car purchase thing is gonna take time, right? You’re not– They’re not buying thousands of cars soon, right? There, there still has to be technology development and, and OEM things. So I don’t know if– I thought it was an interesting call. The stock obviously sold off back down into the sixties. and really the primary issue here is, is this autonomy overhang that’s on the stock

Grayson Brulte: It’s a hangover, Walt. It’s, it’s a hangover. And the reason I’m calling it a hangover, ’cause if you go on X, all the Uber bulls are out there t- trying to give you a Pedialyte and Gatorade to, to get over the hangover. It’s amazing what’s happening to try and cure this Waymo hangover

Walter Piecyk: Yeah, I don’t know what the cure is, ’cause again, this gets back to the, the report we put out last week and, and, and you just look at like what is the market gonna look like a year from now, and it’s not gonna look that much better, right? Waymo is gonna just extend that lead in terms of number of cars that are on the road. and again, if you’re having success with some of these early partners, they still not amounting to a lot of, a lot of cars. This doesn’t mean, by the way, that I’m saying that Uber’s, you know, out of the running, out of the race, whatever. It’s, it’s still early days in terms of autonomy. It, it’s just pointing out, like if you’re looking at the stock that, you know, the number of catalysts or the, or the information flow there and where they can get to in a year’s time isn’t necessarily, you know, a great situation. That doesn’t imply that in ’28, ’29, or ’30 when they hope to get to these 100,000 cars or whatever it is that their fortunes can’t turn

Grayson Brulte: It could. I mean, there’s, there’s still a lot of variables in play in the robotaxi market. And obviously you and I know it’s going global, and Uber’s made a, if you wanna call, a mission to expand autonomy global. They’re setting all initiatives in place, but those are big numbers. It takes a lot of capital. It’s gonna take more than $10 billion to, to reach those numbers in, in my opinion. And at that time, where is Waymo’s gonna be? Or i- will a new incumbent outmaneuver them and ex- and accelerate tech, and then does Way- does Uber potentially have another divorce on the horizon? I say that because at some point, the brand will get stronger, some of these AV companies than Uber, and they’ll realize that they don’t need Uber because it’s not hard to build an app. Yes, the pushback is, well, the demand, but consumers like brands

Walter Piecyk: Da- Dara, first of all, a couple things. Dara on the call indicated that he thought that that relationship would continue next year. Obviously, they, it contractually will continue. You and I have, have talked about that there might be language within those contracts that enables Waymo to get out. you can just refer to last week’s podcast on going through that. I think the other one you might be hinting towards is, is Zoox, who I think Zoox has always wanted to have their own app. Now, you and I are both somewhat skeptical about how many cars Zoox can actually get on the road over the next year. But I, I’ll give Dara credit. I think when asked about Zoox, he made it clear that, look, Zoox will exist on the Uber app, but Zoox will also, you know, you, you’ll also be able to get it on the, on the Zoox app itself. And I guess the real question is, like, if you really want to take a Zoox in any of these markets, are you really gonna bother going through Uber or are you gonna go directly to that Zoox app?

Grayson Brulte: You’re gonna go directly to the Z- the Zoox app and I, I gotta give Zoox credit. When you land at the LAS, La- Las Vegas International Airport, now known as the Harry Reid International Airport, you can’t turn left, you can’t turn right, you can’t even look up without seeing a Zoox ad. They pretty much, they, they painted the airport Zoox green. And so that says, “Okay, download the app.” It’s not saying, “Hi, go to Uber and get us.” It’s pretty clear what their ambitions are

Zoox Approved for Paid Rides in Las Vegas

Walter Piecyk: That’s a good segue to another, our next segment, which is that Zoox is, has gotten approval for paid rides as of next week. And I’m gonna fast-forward, because you bring up the airport, to a point that the CEO Itcha said on Bloomberg when she was getting interviewed when asked about getting to the airport, because, you know, one of the criticisms that we’ve had is, like, there’s a limited number of stops. I had to go to my famous, you know, Caesar chicken, greatest Caesar chicken at the Topgolf in Vegas, check it out. But I had to walk just to try the Zoox. It’s, it’s limited locations. she was asked like, “What about the airport?” And indicated that it’s coming soon. I think Rob had some, you know, your partner had some, some good feedback on what maybe why that might not be as soon as, as Zoox is leading us on, on, on again, a Zoox vehicle going to the airport in Las Vegas

Grayson Brulte: Rob did uncover some really great data. He, he used OMEGA, pulled the NHTSA filing, and the NHTSA filing says, this is a N- this is public NHTSA data, Zoox cannot operate per NHTSA restrictions above 45 miles an hour. So why is that an issue? Well, the average traffic speed coming out of the Las Vegas Airport is 55 miles an hour, and this is AAA data, I’ll cite there. That does not include the freeway. That is the tunnel, where the maximum speed of the tunnel is 35, average speed is 55, and access roads 45, and sometimes they go 60. So it’s gonna be very difficult to keep there because you can go cite Consumer Reports, AAA, any Edmunds, any data you want. When you’re going below the speed limit, it’s potentially a lot dangerous. And I, and I say below the speed limit, you say, “Oh, yes, well that’s the speed limit,” but you have to go with the flow of traffic. And if the ZooxMobile, it’s a mobile, is going below traffic, it potentially risk a safety hazard, which Rob uncovered in that NHTSA filing. There was a lot of safety requirements there. Now that I said that, I have to say this. Topgolf, why don’t you sponsor the podcast? Walt can eat a chicken salad here in person and tell the world how great it is.

Walter Piecyk: The chicken Caesar. The chicken Caesar. Try it out next time you’re in Vegas. It’s very tasty. and look, I’ll just, I’ll throw a good advertisement for you on OMEGA. You know, that’s just a small taste of, I think, what you can get from, from what OMEGA offers in terms of info on the autonomy segment. let’s go back though to this pricing for Zoox. you know, it, they’re– on, when it’s on the Uber app, it’s comfort to your pricing. I’m not sure if it’s gonna match what Zoox themselves are charging. So that’s above, I guess, Uber X. it’s a base fare plus time and distance. If it reroutes for whatever reason, there’re not, there’s not a penalty. What, what’s your, what’s your kind of view on where they were priced in the market?

Zoox Pricing, Wait Times, and Fleet Size

Grayson Brulte: They’re priced as a luxury option. There were several interviews that representatives from Zoox gave in the media where they said on the record they’re gonna price it 20% to 40% above your average rideshare fee. That is clearly positioned as a luxury product. What we know as of June 2026, according to NHTSA data, there’s 105 Zooxmobiles. And we had a friend of the show who was in Las Vegas yesterday, tried to get a Zoox, service unavailable. So if you’re gonna price it as a premium luxury product, you cannot pick your own destination and you have to wait long periods of time, the pricing power is going to evaporate

Walter Piecyk: So I think it’s not a comfort or it’s not a luxury pricing. I think it’s a tourism pricing. These things, as everyone knows, are kind of funny looking things that you might see at Disney World. I know Grayson doesn’t like them. I like the design. It kind of reminds me of a, of a cab in London, a London taxi, ’cause you can. You have plenty of leg room. I’m a tall guy. You know, you can talk to, face someone and, and talk to them. So I think the pricing’s there. And then the other thing I’d say about pricing is, like, I think a lot of times people overanalyze the pricing, these initial pricing, whether it’s Tesla or Waymo or Uber. Again, early stage here, they don’t have to lean on pricing yet. They don’t have to play with pricing. They might do some testing here and there. Wait till they have, like, thousands of cars in these markets and they’re really competing with rideshare and that- that’s when some of these pricing decisions probably really matter. To your point, waiting 15 minutes for whatever the price is, you’re really just paying for, you know, a tourist ride in Vegas.

Grayson Brulte: Well, not 15 minutes, hour and 15 minutes. You forgot 60 minutes there, so that’s 75 minutes. I just, I have to point that out there. So here’s the question. We know Waymo has put out a blog post. They are getting ready to launch service in Las Vegas. Okay? Motional is there. You can already get that on the Uber app. And if you read Elon’s X account, the robotaxi is going to Vegas. What do you think, from your perspective, that’s gonna do to the pricing power? Does that diminish or does Zoox continue to stay, as you called it, a Disney World attraction?

Walter Piecyk: I think it stays as a Disney World attraction. There’s just not enough cars. It doesn’t go to enough places. They’re just gonna maintain that pri- they should just maintain that price. I don’t think they’re gonna have an issue you know, with, with getting enough demand to, to fill whatever supply that they’re, that they’re planning on there

Grayson Brulte: So that raises the next question. In this NHTSA document that OMEGA found and Rob analyzed, five cities of service publicly approved, two blacked out in the NHTSA document. So we know that they have approval to operate in seven cities. 105 vehicles for seven cities, wait times are not gonna get up. How do you think that Zoox is going to deploy the, the vehicles? Will they just deploy 90-plus percent of them in Vegas just to try and get these wait times down? Because it’s gonna have a, a pretty negative impact on their brand if these wait times continue. And Aisha even admitted the wait times are long in a Bloomberg interview this week, so we’re not the only ones talking about it

Walter Piecyk: Yeah, but on– in that Bloomberg interview, she also talked about how production is ramping up, US production by the way. and, you know, she’s counting it by number of vehicles per day versus per week. Wasn’t specific with the actual numbers, but just implying that, you know, this 100 is just a first start and that things are really gonna start kicking in in terms of, of getting these vehicles off the line.

Grayson Brulte: Yes, but I have to be me. There’s a risk. I’m very thankful they’re made in the United States, but they’re made in California, and there is a risk of unionization, which will de- delay the timeline and push up the cost. But progress

Walter Piecyk: Before we move on, just one last thing. Are we exist ignoring the fact that this thing only has a limited number of stops in, in Las Vegas? I mean, is this true autonomy if it doesn’t go to every hotel that’s on the strip at a minimum?

Grayson Brulte: I gotta say, you had a really great back and forth with Jason. You said, “Can we go to the Cosmo?” And Jason goes, “Why Cosmo?” And then you had a response, and I have to vouch for you, the greatest party at CES is the LightShed party, and thank you for inviting me. I was very honored to be there, met a lot of great folks. But Zoox, you gotta go to the Cosmo because you gotta go to Walt’s LightShed party. It’s where it’s at. And then maybe we’ll do an Autonomy Markets party there as well

Walter Piecyk: I think that’s a good idea. Zoox, if you’d like to be a sponsor of the Autonomy Markets Party, be our guest, but it might be at the Cosmo, so you gotta get– We, we’d have to book it somewhere where it actually Zoox would actually drop off our guests.

Grayson Brulte: Valid point

Waymo Returns to Freeways After Two-Month Pause

Walter Piecyk: Th- so there are cars that do go on highways and are at least licensed to go on highways, I know other than what you know, Rob uncovered about Zoox and some of their limitations, and that’s our friends at WeMo, Wa- WeMo. Our friends at Waymo. By the way, last week there was an edit that didn’t get edited that someone pointed out when I misspoke. Okay, leave this one in. We’re not editing it. We’ll do it live. Okay, so Waymo freeways are back, and they’re returning to freeways after a two-month pause ’cause of the, all the construction zone thing. these are the airport unlock, right? And this is the same thing that Rob was pointing out to in terms of Zoox. You’re gonna have to get on a road that’s got 45 miles or more to get to a lot of these airports. There’s not many airports where you can just take back roads to get into. So what do you think about this in terms of a progress point?

Grayson Brulte: When you said Waymo, and next thing I thought, you’re gonna say Oscar Mayer Wienermobile. I mean, I didn’t know where you were going with that. It’s a positive. It is a, it is a positive step. And for any of our listeners and viewers who have taken the Waymo service from SFO to go down to the Valley or to go into the city, you know that it requires highway. And if you’ve punished yourself, which I have, and you’ve done surface roads, it’s not a lot of fun. So this is a very positive step for, for Waymo, and they admitted that there was an issue. They corrected it, nothing happened, and then they’re moving on. So I view it as a huge positive. And you’re right, it is a giant airport unlock. If you cannot access freeways in the m- m- majority of international cities, you’re not gonna be able to unlock the true benefits and revenue that come with airports

Walter Piecyk: This is kind of a quiet week for Waymo in general. Should we expect a, a new market next week? It seems like they’ve been quiet. I have to go back to my little scoreboard here to see which Midwest city we should expect them in next

Grayson Brulte: Waymo’s been quiet, but I will tease this. OMEGA’s been picking up quite a few new corporate filings around the world. So when and if Waymo decides to expand there and make those public, we’ll see. But Waymo, OMEGA thinks something’s cooking

Lucid-Nuro Robotaxi Timeline and California Regulatory Risk

Walter Piecyk: I love it that there’s, there’s some more reasons to check out a subscription to Waymo. Meanwhile, on the reporting front we had Lucid reporting. Why do we care about Lucid? It’s the partner to Nuro. and Nuro is one of the key partners for Uber to execute on getting some of these autonomous cars in the market. So when they. And remember like a week or two ago, there was this, you know, false report about them looking to restructure whatever. So look, they reported, the stock got crushed. Q2 didn’t really hit. Gross margin was negative by like 100%. They’re burning, I think, over a billion dollars I think in the quarter. The deliveries weren’t great. But at least they’ve kind of recognized that autonomy is, is the sto- top strategic par- priority, and that the Uber Nuro robotaxi project is a, quote, “must-win project.” And they put this all now in Lucid technologies. Hopefully, this isn’t part of the, you know, a corporate structure where they’re putting the robotaxi in a protected entity while other stuff happens somewhere else. I don’t know what the full you know, color on that is. But I think it’s, it’s at least good and, and obvious that these guys are, are focused on working things out with Nuro. I guess my question to you is, it is August early August. The timeline is, is, is clicking quickly for all three of these companies, Uber, Nuro, and Lucid, to deliver on some type of, of autonomy robotaxi in California, in the Waymo market you know, by year-end. Do you think they make it?

Grayson Brulte: No, No

Walter Piecyk: And, and what do you think the, what do you think the, the holdup is?

Grayson Brulte: The holdup, in my opinion, will be the regulatory holdup. And, and I want to clarify, there is a high likelihood and a high probability that they’ll be able to start service that is not paid. ‘Cause you can go talk to Rob and he can break down all the permits you have to get. That’s, that’s his role, that’s his job, and he’s great at it. There’s a high probability they, they, they can offer free service but not paid service by the end of the year. They must deliver, but if you’re looking at this and saying, ” Okay, you said you’re gonna deliver a robotaxi service and it’s free,” it’s like, “Huh? You put all this money in it. Huh? You can’t charge?” It’s just, I said this going on two decades now. Your first launch market for a robotaxi service should never be California. And I have to give a lot of credit to Alphabet and when they made the decision to go to Chandler. It was the right decision and, and, and look what happened when they made that decision

Walter Piecyk: Well, that big– the biggest problem then is, is really for Uber because you can’t. To call something a market launched, launched for Uber means like it’s on their app and you’re paying for it, even if there’s a safety driver in there. Like AVride in, in Dallas is launched and, you know, we know some of the, the, the issues that they’ve had there. So for Nuro it’s not a big deal ’cause like, okay, you have 100 vehicles, you can at least be out there showing the world that your tech works in these Lucid cars, even if it’s probably gonna be employees or whatever it is taking it. and then hopefully I guess they, they can get through some of these regulatory hurdles and, and, and have a true launch in, in early 2027. So I don’t think it’s the end of the world, obviously, if that’s, if what you’re saying is true. But, you know, you and I follow this pretty closely, right? It has been talked about a, as an end of the year type of launch. you know, and it’s one of the more important near term partners for Uber. So, you know, I guess we’ll have to continue to monitor that closely and I guess, you know, lean on your, your buddy OMEGA again

Grayson Brulte: This is a theoretical hy- hypothetical scenario. Let’s say that we are correct, the regulatory environment stops them from launching a paid service in San Francisco. And then you and I are sitting here, we’re getting over our turkey coma, getting ready hopefully to go to London, and then all suddenly across the terminal, Uber Nuro launch service in Houston. How does– What is the reaction from that if there was a pivot because of the regulatory environment to Texas, where there’s a lot clearer path to doing this? They could launch in Texas by the end of the year in Houston if the technology’s ready. So I wanna put that out there. What, what would the Reaction be?

Walter Piecyk: I think the reaction would be very positive for all three of these companies. And you just say, “Look, you know, we don’t need to deal with California, and we can get things going.” And it doesn’t have to be a ton of, a ton of cars. Like, if you can get something launched in any market, 25, 30 cars launched, you know, ideally with not a driver in there, we’ll see by the end of the year. I don’t care whether it’s, it’s California, Houston, or, you know, wherever, you know, Peoria. get something, get something on the tape, and then you kind of move forward to, you know, is Lucid getting to that point where they can start this program to get to the th- 35,000 units, which would be huge for Lucid as a company, obviously huge for Nuro and, and getting to their next you know, kind of wave of, of expansion. And obviously a big positive for, for Uber as well, where there’s, like I talked about before, like not a lot of, you know, scale progress that, that is likely over the next 12 months

Grayson Brulte: You’re right, because if Lucid can figure out the business, figure out the manufacturing, and hit the manufacturing targets, Uber has a very clear line of sight to vehicles. And as you and I know, that’s one of the most difficult things that happens in this industry

Walter Piecyk: Let’s put this in context one more time. 35,000 car program, the deliveries that Lucid had in this quarter was 4,000, right? So this is, you know, this is a big deal for Lucid. I think they get that. So we wish them the best and, and obviously as that production line gets going, we’d love to, to, w- to to visit it

Grayson Brulte: Love to visit it, and then I’ll ask this question here for the record. At some point, does Lucid just become a contract manufacturer for robotaxis and shut down the sedan and Gravity business? That’s something to watch

Walter Piecyk: We’ll have to see. let’s move on to another. I wanna touch briefly on this. DoorDash, which doesn’t move people around, it moves products, food reported. Now here’s a company, very smart management team you know, announced like last year about investing in technology and autonomy. The stock I think was challenged by that ’cause obviously investment, which is what you should do sometimes, some in- some type of investors don’t, don’t like that. But I think, you know, you and I are, are certainly believe in the future of autonomy. I was, you know, the, I’ve t- I mentioned this before on this call that we do these calls. There’s the earnings call and then they do a call with analysts afterwards and, and what was. And I get on these calls and what was surprising is just the lack of, of optimism or potential that I saw for autonomy for food. And I’m just curious, Grayson, from your standpoint you know, do you think food let’s just sp- talk about food, not the other things, the other products that they deliver and, you know, let’s go beyond serve and beyond the DoorDots or whatever the robots are that, that DoorDash has, ’cause they also have a partnership with Waymo, right? Where they’re, people are putting cars in the trunks of Waymo’s. is this a market, a realistic market over the next five years?

Grayson Brulte: Yes. Absolutely. And I’ll, and I’ll tell you the reason, and I’m not dunking on anybody, but I’ll tell you the reason, and it hasn’t been s- discussed outside of wonky policy circles. Rob pointed this out. When you start doing delivery, you’re starting to, to cut down on deadhead miles, and that helps you from a regulatory standpoint in certain localities. That is the big thing to watch there. So when you’re delivering food, you’re cutting down on deadhead miles. That’s a positive regulatory. Now that we have that said, from a consumer standpoint, I love it. I don’t have to talk to anybody. I can just go out, I can hit the, let’s get the open button on my phone, do-doop, get the food out of the burrito out of the back, go in the house. This is great. To me, yes, it makes sense. I don’t wanna have to talk to anybody, and then you don’t wanna feel bad. I don’t have any cash. How do I tip the person? ‘Cause I got news for you. America has tipping fatigue

Walter Piecyk: I mean, the cost fatigue as well. I mean, you look at what Uber Eats and DoorDash are charging you for delivering of some of this food, the costs are high. Now you can amortize what could potentially be a purpose-driven or purpose-built vehicle. And I’m not talking about like what Nuro did, where it’s kind of like not really a car and it’s on the road doing 20 miles an hour. No, like an actual car, but purpose-built, you know, for transmitting goods and services. Or let’s like really look out to the future and say you have your personally owned private vehicle, and you’re just sending your own vehicle, you know, to, you know, the local restaurant and, you know, to pick up the food for you. I mean, these are things that I think are, are on the horizon and, and are just a natural incremental revenue offshoot from the mass of the billions of dollars of, of investment that are already being made. So props to DoorDash for having the foresight to, to, to invest there. And, and also note that whether it’s DoorDash or Uber, this is probably an area unlike moving people around, where Waymo and Tesla can effectively, w- and even Zoox, if they’re successful, cut them out of the equation altogether. In this world where they have these relationships with the restaurants and the retailers, it’s more likely that autonomy can only help DoorDash and Uber as opposed to b- be a potential threat

Grayson Brulte: Yes and no. Let’s give you an example. If I could send my Tesla to Whole Foods, they put the groceries and sends it back, that’s a game changer for me. That’s awesome. That benefits Amazon. That’s great. But w- the the likelihood of me sending it to a restaurant to pick something up, and if it doesn’t have a dedicated spot to go, you have a problem there. Where f- Uber, with U- Uber and DoorDash already had that relationship, that’s a benefit to them. So perhaps your, your Teslas of the world, they take the, the grocery store, the Costco run stuff, and that your DoorDash or Uber Eats continue the restaurant stuff until you get some of these large restaurant chains that invest in this and probably start to say, “No, we, we don’t wanna pay your fees.” Because if that does happen, and all suddenly I could send my Tesla, you could send your Tesla your mom could send her Tesla, all suddenly those margins start to compress ’cause the restaurants aren’t gonna pay the fees anymore. So that’s something to watch there.

Walter Piecyk: I agree with that. But personal ownership is like kind of the nth thing down the road. I think there is a long period of time where the rideshare version of autonomy and that ownership and, and we know all the complications in terms of insurance and who own- like who’s taking responsibility and, and what people are willing to do. On the infrastructure side of it, you and I both know there are very smart people out there already thinking about ways, planting the seeds today for, you know, the autonomy wave that is to come, whether that’s for, you know, humans or for some of these delivery robots by ground, by air whatever it is. So it, it’ll be interesting to see how this all plays out. But I guess we’re in agreement that that food market is, is also, I think, a pretty sizable one and not something that’s, that’s not you know, kind of out of the reach of all of these companies.

Grayson Brulte: And when you look at food, don’t discount big box retail. I’ll leave it at that. It’s not just restaurants, it’s not just grocery stores. Do not discount big box retail

NVIDIA Alpamayo 2 Open Model and the CUDA Ecosystem Play

Walter Piecyk: So moving on, we had Jensen of NVIDIA on August 4th, I think the day before maybe Uber reported basically releasing this Alpamayo 2 Super. So they’re kind of delivering– This is the open model for autonomy. This is what they’re saying like, “Hey, go download this on Hugging Face,” of which I think there’s been 500,000 downloads on Hugging Face for companies to look at and say, “How do we use this to help us in our autonomy development?” Now, that’s a little different than what they’re trying to do with Mercedes, which is like, “Okay, here’s the open model, but then we’re gonna do our full stack, and then we’re gonna have something sit between it and say, well, the full stack or the traditional model is saying this is the trajectory your car should take, and the open model is saying this is the trajectory, and then deciding between the two.” And it’s, and it’s in, in that combination, which is what Mercedes is trying to do with NVIDIA, and I think that what would be great for Uber is if more OEMs, you know, went down that route. that’s the bigger thing, I think, than the open model. But what was your takeaways from this open model and the 500,000 downloads and like what role Alpamayo in a silo you know, will play in, in helping this industry?

Grayson Brulte: CUDA, CUDA, CUDA. And if you’re wondering why I’m saying CUDA, it is the developer platform that all the NVIDIA products are built on from a developer standpoint. NVIDIA very strategically, I give Jensen a lot of credit, very strategically is getting companies into the NVIDIA ecosystem. I believe it is a way to grow that automotive business and to continue to grow, as he’s calling it, that physical AI business. To me, it is a brilliant strategic move to get younger companies, and potentially some older established companies you said, into the NVIDIA ecosystem to start building on their platform, ’cause once you build on that platform, it’s very hard then to go swap into a TI, to a, to a Qualcomm. It, it’s a brilliant strategic move that essentially, I’ll just say it, is locking companies in for the long term into the NVIDIA architecture platform

Walter Piecyk: I don’t disagree that it’s smart intellectually and that it should open, open, you know, pun intended, open things up, open the playing field. But then the concept is like, forget about new companies. Let’s say existing companies with a lot of money, GM, Ford. Ford gets Alpamayo, and then they still have to sit there and develop their own full stack on their own, right? With their own people. I certainly don’t have the confidence that an OEM has the cultural DNA to do that. No offense, Sterling, but like you’re one guy at GM, like within a massive infrastructure of people. And then in addition to that, I don’t think that, I’m sorry to say that NVIDIA’s full stack, you know, per- version of this that would help those companies even further and really lock them in to the NVIDIA ecosystem is, you know, is close to where the existing players are in the market. And I think case in point is, is Mercedes. Like, you know, we’ll see if they deliver the cars that they say they’re gonna deliver, which is, by the way, the commitment, just to be clear, is first half of ’27 to Uber with safety drivers in, right? In the first half of ’27. And I don’t think they’re really getting driver-out until I believe it’s, it’s 2028. I’m just skeptical that NVIDIA, on that portion of the business, can deliver. Now, will Alpamayo help other like new startups like you’re saying, like a stripped-down car that wants to do, that has some smart people or help Rivian or whatever it is if, if Rivian wants to go? I mean, sure. I don’t– I’m just less, less optimistic that, you know, NVIDIA is gonna be a full stack competitor in the market. What are your views?

Grayson Brulte: Full stack? No, not necessarily. Service provider? Yes. Large service provider? Yes. Critical player? Yes. But there is a lot of things that still have to be proven from a full stack perspective. To my knowledge, there is not a SAE Level 4 robotaxi operating anywhere in the world that is full stack Nvidia. If it is, please send us a video. Invite us out. Invite Walt and I to test it. I am not aware of one. That is, to me, is the milestone that has to be seen, because there is a very big difference between supervised miles and unsupervised miles. Completely different

Walter Piecyk: So if that’s the, if that’s the conclusion that you and I come to, and if, and if you’re like a, you know, a CEO at one of these, these OEMs, and you’re deciding then between, okay, maybe I don’t care about the full stack, at least Alpamayo’s gonna help me with getting to an ADAS solution, like wouldn’t Qualcomm have a pretty good bid on that business given the, the relative potential price? Obviously, they have a deal, I think, with BMW already. and in that case, lower cost, maybe not as much lock-in, maybe not as much competition for a company in Nvidia that’s, that’s, you know, supplying compute across a number of different you know, use cases.

Grayson Brulte: Qualcomm, in my humble opinion, is a sleeper in all this. They have the Qualcomm digital chassis architecture. I would encourage the listeners and viewers, read the developer docs. Go, go through the developer documentation that Qualcomm has public- published publicly. It’s impressive. And you and I know, Walt, that Qualcomm has a very healthy L2 business, L2 plus business, and I’ve said this on the record and I’ll say this again. First step for Qualcomm, I think they knock out a big chunk of mobilized business. Eventually over time, I could see Qualcomm, depending on the development of the Snapdragon digital chassis, becoming a player in licensing an L4 system

California Teamsters Sue DMV Over Autonomous Trucking Rules

Walter Piecyk: Well, we look forward to hearing developments from BMW with Qualcomm and, and more so from Mercedes. Again, on the clock time is moving quickly. I’d like to see– I’d like to get in one of those cars and, and see it. I know it’s, I know Dara has done some rides and Jensen’s done some rides, I think, in these things, so would love to try that out. Let’s, let’s move on to some regulatory stuff in trucking. You know, Aurora’s been, you know, we’ve been making some noise of, you know, they’re, they’re trying to get to driver-out and, and signing up some customers. And then most of this stuff is, by the way, in Texas, so let’s make that clear. So back in California the Teamsters of California now sued the DMV in the Alameda Superior Court to, to repeal what were some, you know, new regulations that were proposed. And they were threatening all kinds of stuff. I think specifically Aurora might have been cited in the complaint. maybe not specifically by name, but talking about a thousand-mile driverless run. I mean, obviously this is all about protecting jobs relative to the safety, given the type of drivers that, that some of these companies are now approving, especially in California, to get in these trucks and putting, you know, people’s lives at risk on the road. By the way, even if this doesn’t happen, isn’t, if Becerra wins this election isn’t he gonna reverse these, these DMV rules anyway? So, like, what do you make. You, you’re an ex-California you know, what, what would I call you? Politico? Tell us what all this means. Is it all noise or is this a true risk for autonomous trucking?

Grayson Brulte: Wanna know what it really means? California’s closed for business. And on the LightShed side of things, your partner, Rich Greenfield, has done a brilliant job covering the Paramount-Warner Brothers merger, the Netflix drama, and now the attorney general drama at the state level with the lawsuits. And now you’re seeing the Teamsters are suing in California to stop driverless trucks. Why the fuck would anybody want to do business in the state of California? The state might as well hang a closed for business sign on there, ’cause that’s what this says. There is yes, you’re giving the pushback on the size, and I’m sorry, Mrs. Piecyk, for swearing, but that’s how I feel. The, the pushback of the freight market, is this worth the headache? No. This is a, a, a, a, a, a, a, a tactic, and I’m not gonna say the reason why this tactic is being used. You can reach out to me and I’ll explain all the nuances to you. But this is par for the course in California, and California is not the home of the innovation economy. California is the home of the nanny state

Walter Piecyk: Now, there was something that showed up in my replies, someone that showed up in my replies asking Grok to explain a commerce clause, and basically was indicating that there’s some federal kind of preemption that could occur here. Did you read that and, and is that possible?

Grayson Brulte: There is a, a bill currently going through, through the House that would preempt the nanny state. Oh, here goes the attorney general. He’s gearing up taking notes. Yes, you’re gonna be preempted. And oh, by the way, while I’m at it, Ellison should move to Texas. That’s what he should do,

Walter Piecyk: I pressed your, your politics button and you’re going bananas. So the answer I think sounds like there’d be years of litigation, and this is a frustration on certain states. I think my mom would probably agree with you on that topic, but maybe not in the same words

Grayson Brulte: Mrs. Piecyk, I’m very sorry for the language. I get heated up on this because this economy is built on free market principles, not regulatory capture in, in, in, in doing favors for political donors. That’s not how things work

Walter Piecyk: Let’s move on to one of my mom’s favorite topics, Tesla. She loves, you know, hearing about and, and knowing and, and is really desperate to get her own Tesla to drive around ’cause she’s no longer eli- enabled to drive, so she really wants full autonomy to happen for herself. But there was a story, I think it was The Wall Street Journal talking about executives preparing for a separation of the China business ahead of the SpaceX merger. Obviously Elon immediately attacked them, called it fake news. but I think what’s interesting though is it kind of makes a little sense though, doesn’t it? I mean, the SpaceX has a lot of government business and the Giga Shanghai which builds I think about half of all Teslas. So I don’t know, like, does it- everyone seem to think we’re headed to a merger, obviously given SpaceX’s, at least in the near term, a struggle. We’ll see if this really holds. Raising capital is gonna be a very important part of this business, seeing these two companies combine and realizing some of those truly vertical synergies would be a positive. do you think that merger can happen with the China business, or would they have to get rid of it?

Grayson Brulte: I’m gonna defer to you because this is, to me is, it is a telco corner, and I’m gonna defer to you based on the comments that were made on the SpaceX earnings call around telcos. So I’m gonna defer to you on whether it has to be spun out. If there is to be a merger, I’m not gonna give a prediction one way or another on a merger. I’m, I’m unsure. All I’ll state for the record is that I own a Tesla and it works, I own a Starlink and it works, and I’m a happy customer. With that said, I’m turning it back over to you

Walter Piecyk: Well, I’m gonna say this. sometimes Elon may call something fake news even though it might not be fake news, but then it then becomes fake news ’cause he’s like, “Well, screw this. I’m just not gonna do it,” as a result. but you have to take, I think to a certain extent, the CEO at their words. I just think it’s, it’s a string worth pulling even if it doesn’t happen. Like is this, does this become a complication for that transaction? ‘Cause, you know, defenses, that’s a, a huge revenue. But there’s a lot of things that I picked up on this quarter. I’m not gonna go through it on Autonomy Markets on the SpaceX quarter, the positive things that are happening. Grok AI and, and the space, and especially in the connectivity side. You have to tune into Lightshed podcast for that one. But it’s a, it’s an interesting concept if, if this kind of belief of this ultimate combination happens. By the way, meanwhile, they just announced, again, getting back to where jobs are welcome, Texas open for business, that the Terra fab is gonna be built in Texas, not, not surprising. the plans for it look pretty amazing, and let’s– it, it should look, you know, like it’s gonna be pretty good. So these companies are gonna obviously be working together in some shape or form, whether they merge or not

Grayson Brulte: Absolutely. Texas, open for business. David Ellison, move to Texas. Call Governor Abbott. Bring the business there

Walter Piecyk: That’s good, good stuff. Okay, let’s go, let’s, let’s leave the US. This is not the foreign autonomy desk, but let’s talk about London, which we, you know, are hoping to, to visit, to hop in some cars there before the end of the year if, if things work out. But our friends at Wayve just got some clearance. They have a TFL license. What is this? Tell us what happened in London. You’re the, you know, you’re kind of on top of this. Once again with OMEGA

Wayve Secures TFL For-Hire License in London

Grayson Brulte: Oh, thank you. So the, the TFL for hire license, this TFL transport for, for London. What that means is that Wayve can legally move paying passengers with a driver or a safety driver. FT reported, and also the BBC reported, The Telegraph reported, that the permit is for 15 vehicles. In order for Wayve to remove the driver and, and launch a true robotaxi service, whether on Uber or not, I said it, it comes down to something called an APS permit. It stands for the Automated Passenger Service permit. That is the key in every relationship in London is the APS permit. The APS permit, in my opinion, is an asset ’cause that controls who controls the service. This is hypothetical. If Wayve is the APS permit holder, they can launch a service without Uber. If Baidu, for example, is the APS, they can launch the service without Lyft or Uber. Watch the market for who owns the APS permit, and if you want a deep dive in this, last week in The Road to Autonomy, This Week in the Autonomy Economy newsletter, we gave a deep dive on this. Or if you want more, reach out to Rob and I and we can walk you through every layer. It is complex, but watch who owns the APS layer

Walter Piecyk: That’s really good color. but let’s, let’s, let’s get into the guess mode. Obviously Waymo’s gonna operate on their own. my guess would be that, that Wayve would not wanna do something on their own, that they’re gonna be beholden to Uber. Baidu though, like, you know, I would, you know, Baidu would seems to me like a company that, that would obviously wanna do something on their own like Waymo. What are your, what are your views on those three? On, on who kind of goes direct versus sticks with their partnerships?

Grayson Brulte: My view is 100% Waymo gets the APS, operates fully, solely on their own. Wayve, based on all the communication that they put publicly into the market, and Julianne’s doing a really good job with the marketing there now, they want to become a pure licensing company. So I don’t necessarily see them going for the permit. I see Uber going for it. Or if Wayve does get the permit, they will stick with Uber because they want to be a licensor. Baidu’s the kicker here. Baidu gets that permit, and then when their relationship with FreeNow, by Lyft, ’cause you have, the FreeNow’s the one who holds the license, not Lyft, and Uber, Baidu can say, “See you later, we’re launching Apollo Go in this market,” and leave. That gives them the flexibility. And just so for the audience knows, the APS permit lasts half a decade. Six months before expiration, you can start renewal. How long the Uber and FreeNow relationships are, we don’t know. But if Baidu holds it, there’s a clear path for an off-ramp. And then Uber maybe had three martinis or four martinis, and the hangover goes to London

Walter Piecyk: So that was, let’s just consider that part of the autonomy desk. Let’s continue now to Pony.ai, who I think gave us some information in this briefing earlier in the week on robotruck. what do, what do we know about robotruck? Is this another autonomous trucking you know, path that we need to follow here?

Pony.ai Robotruck Ambitions and Tesla Semi as a Sleeper

Grayson Brulte: It, it’s funny past couple weeks I’ve been getting emails for the Road Autonomy, Autonomous Trucking Index, “Why are you not covering Pony’s robo trucks?” And then all suddenly, we get this announcement from Pony. Pony’s been working on robo trucks, if you wanna call them autonomous trucks, for years. When I was at their facility two or three years ago in Fremont, they showed me videos of the, of the trucks operating in China. So we know they have a robo truck program, and now Pony put some numbers on it. They’re trying to, targeting 500 to 1,000 Gen4 trucks in two to three years. We know CATL’s the partner, but we have no clear insight outside of that they’re gonna be operating on mainland China. We don’t have miles. We don’t know whether supervised or unsupervised, and that’s very important in autonomous trucking, whether they’re supervised or unsupervised and that, miles, and that includes somebody in the cabin. We don’t have that detail. All we basically know is that Pony.ai and James Wu and the team are looking to ramp up their autonomous trucking efforts on the mainland. But there was a little fine thing in there, foreign ports. So I’ll go on the record and say perhaps Rotterdam

Walter Piecyk: Well, we’ll see. I mean, if the target is 500 to 1,000 heavy trucks in two to three years, I think Aurora’s gonna be ahead of them. I think Kodiak’s gonna be ahead of them. you know, so I don’t know if that’s kind of leading edge. And then if it’s China, like labor costs there and, and insurance costs and things, not exactly the most attractive market to operate in. So I would hope that your Rotterdam or whatever you just said would be happening. But, you know, I’m sure they’re on the radar, but I think they’re kind of far down on the radar based on the other companies that are in the mix here. I think I’d be more optimistic seeing like Tesla’s Semi than Pony.ai delivering 1,000 trucks in the next three years.

Grayson Brulte: I agree. and we, and th- as the market knows, there’s o- opaque data in China. Tesla Semi is the very interesting one to watch. Matt McClellan from Covenant Logistics has tested it, ridden in it. Our friend David Moss has ridden in it. They both love it. Matt put it- said for Covenant it worked great, so I say this for the record, I think Tesla Semi is a sleeper. The Tesla folks, Walt, myself, and Rob, we’d love to go visit the Semi factory in Reno. We, we, Giga was cool. Would love to cap it off visiting Semi

Walter Piecyk: And I’ve never been to Reno. I like to travel to a lot of places, so just give me a reason to go to Reno. I- I’ve heard positive things. Have you ever been there?

Grayson Brulte: I have been there. It’s the biggest smallest city in the United States. I have been there.

Walter Piecyk: Okay. Well, let’s get there. All right, so what do we have for next week? We have Zoox hopefully with their paid rides on August 10th, and I’m sure our listeners will give us some videos and some feedback on, on how that’s going. are you flying out to Vegas or are you gonna do a quick field trip out there to be the first fare? Or maybe our buddy David Moss will head to Vegas and be the first fare in Zoox. David, I know you listen. Can you get to Vegas and be the first fare on a Zoox?

Grayson Brulte: David is currently in Vegas and he was posting videos in, in Zoox last night. But I thought about it, I s- I said to Rob, I said, “Do you wanna go to Vegas?” And so we did what any, any, any good person would do. We read the terms of service, and there are some very interesting clauses in there about videotaping in there, and I’m af- and it becomes Zoox property. So I wanna be able do a field report according to their terms of service the way I interpret it. And for the anybody that’s gonna ride, and if you see anything wrong with that vehicle I would suggest you read the terms of service, and you know exactly what I mean. I would suggest you read the terms of service before you even ride in it. That’s all I’m saying. That’s Not advice

Walter Piecyk: I highlighted. I– On Twitter, I highlighted one of the terms and services that Claude helped me understand ’cause it was a very long sheet that they emailed. Also next week we have DBS who owns AVride. they have their earnings call. Who knows whether they actually mention what’s going on at AVride. It seems that they’re more focused on the NeoCloud or whatever you wanna call it. WeRide a company that, you know, we’ve said very positive things on, on this ear- on this call, on this podcast, has their earnings call. Lyft obviously wanna get some update on what’s going on at Nashville. They report tonight actually, so this is occurring this week, but you, we won’t talk about it until next week. and then, you know, what else? I think just getting ready for our, for our big Dallas trip. We got a lot of field trips, Dallas, London, California, Vegas, and hopefully Reno on the list. Get a. If you’re out there in the autonomy industry and you’re in a city we haven’t listed, let us know. We wanna come see your, see your operations, Especially anyone in Dallas

Grayson Brulte: Yes. We wanna come s- we wanna come see you and do field reports, inspect, and just, just have open conversations about what you’re building. And also, Walt, tonight, after we record Kodiak’s reporting, so we’re gonna have a lot of stuff to go through next week. And each and every week, Walt, I gotta tell you, when we first started this, you and I were debating is there enough stuff to talk about every week? Should we do every two weeks? And now we made the decision to go every week, and lo and behold, we’re rewarded because autonomy keeps on chugging along. Autonomy is becoming an economy. The future is bright. The future is autonomous. The future is unsupervised. Walt, until next week