Waymo's Lead and Autonomous Trucking's Inflection Point

Well, first and foremost, the death of San Francisco is greatly exaggerated. We spent some time in downtown and I was in most of the time in the valley. San Francisco is, you know, much changed. So, congrats there. I don't know if our friends in Texas would reconsider moving back to San Fran. You know, because there's other issues that obviously California has. But really positive experience. And obviously still a ton of great energy and optimism within the tech folks that are there. Unfortunately, this means a ton of traffic that we had to deal with. Sorry, Mom. And thank you for all the industry listeners who reached out and we were able to kind of meet for the first time. So, that was great to meet some new people within the industry. But yeah, Grayson, I did feel like you at one point we were on whatever road that is that connects San Fran. I forget the names. Always in an Uber or Waymo. And I saw this gigantic Waymo depot. I posted it on Twitter, now known as X. Had some Zeekrs in there, the Geelys, that were there. I saw one on the highway as well. What did you think of that sighting? I was channeling your inner Clouseau on that one.

The hat was not there. That was one of the most massive and busier ones that I think I've seen and actually upon observation looked like the most effective in terms of other ones are I don't know weren't operating that efficiently. That one looked super efficient. I don't even know who owned it. We got to do some more work on that. We'll get back to our listeners and we have some more of those details. We also met obviously with a bunch of executives across the industry Aurora, Uber, whatever rode in, you know, multiple cars, Nuro, WeRide, Waymo, all of these were solid trips. I'm going to prefer to withhold ultimate judgment. I mean, again, you and I think I've ridden in more of these than most people. I was joking with Brandon on the LightShed podcast even he now has been in a lot of these autonomous cars more than most people that I think cover or at least claim to cover the space. But again I'm going to withhold judgment because you know a lot of times you visit the companies they put you in the cars you're kind of on a set route. I'd rather just be able to pick at will my location that I want to go because I've had, you know, experiences in the past whether it's AV Ride or May Mobility where you have kind of the set ride from HQ and then you know there's some differences when it's actually deployed out in the field and you're picking a location. With that said, you know, again all of these rides, you know, were very high quality. There were, you know, some short stopping on one of them or two of them. They all kind of people were complaining maybe that they were following the speed limit. I think that's kind of early stages unlike Waymo which was obviously more aggressive but I think, you know, generally a very positive experience and again just once again Grayson increases my optimism kind of about with the opportunity for this industry and the number of competitors that are still in the mix to deliver product.

Well thank you for noticing that one. I was kind of whipping through a lot of the companies because we saw so many companies. Yes, good job. Clappy hat for you for not that I was trying to sneak it through there. I mean, there was no one in the driver's seat. So, I was in a WeRide, you know, and there was literally no one in the driver's seat. We also got a great update from their team on the tech evolution, the market opportunity, kind of how they view the different competing technological solutions out there. A very thoughtful, I think, approach, you know, to the global market and clearly a or not clearly, but like I said before, you know, a smooth ride. Again, I'm going to caveat that by saying like, you know, it's their selected route. It's not necessarily something that we picked. That's fine. I could fly to Riyadh and try it there, I suppose.

Good hire for Nuro. Also an Eagle Scout fan. So, that's a positive. I feel bad for them out there in the land of the Niners and having to deal with those fans. You know, again, good ride. I will say that, you know, we in that ride, we drove out of the lot as opposed to basically like hitting a button. I think that's kind of with the process of how their AI is working where they're putting in these kind of drop-off and pick-ups. Drop-off and pick-ups are still certainly a challenge. It's the topic that came up at every one of our meetings. I think also, you know, having to deal with Lucid is an incremental, you know, process, right? When you change cars like, you know, whether I was in a we had a Toyota and there was a Lexus that we were in. Lucid's obviously going to much more comfortable car to Lexus. Sorry to Lexus owners, but it's just not a great car. So, but I think like the Lucid thing is going to take time. We weren't able to get into a Lucid car. And I think this is a challenge that any company and something we should think about when we talk about some of these startups and like how many OEMs can you truly partner with? You need to dedicate resources to get a market commercial, to get a 100 cars commercial in a market to prove out to your partners, to the market, to investors to get forward. So, I think, you know, that it's kind of a critical decision, the partners that you choose. In the case of Lucid, like they're needing to a certain extent Uber's help in terms of financing and, you know, I don't know. It's not I don't think ideal, but I think Nuro certainly's got other opportunities. And I think we'll continue to develop with Nuro. The other interesting thing about Nuro, I think they have actually an opportunity with their tech when they launch and they go commercial at the end of the year to not be like AV Ride and some of these others where there's a safety driver there. I think if they when they launch commercially, I think they're gearing up to have a driver out in that situation. We'll see. I mean, I don't know, no promises there. That's just my kind of sense on, you know, the development path of that tech. And I think they've got some other interesting things going on there.

I don't know. I think that's a great question. And maybe for my picture back here, I need to start putting some Eagle stuff there. But, you know, the Eagles fan base can be can also drive there for every Eagles fan club, there's an Eagles hater club. That's for sure. Maybe 10. Look, the bigger takeaway, I think, from this trip though was the use of, you know, real data. Like, you know, Waymo, you saw like fleets of these Mach-Es going out to collect data. Interesting, by the way, which we'll get into a little later, in that it looked like they were training without lidar. I mean, with lidar to basically help like a camera-based what appeared to be like a camera-based solution. You see these things, obviously, on the road as they kind of exit where they're at. But collecting data is important. And I actually, Grayson, you know, this age-old question of like, "Why doesn't Elon use lidar?" And I think I got finally one of the more interesting takes on this, which was, well, if all of his data is from camera only. So, like, if he decided, "Hey, we're going to slap in a $500 lidar into our windshield, you know, would that whole process have to restart because they don't have the training data that includes the lidar?" That has nothing to do with all the other stuff that we've talked out before. But, I mean, that could be a hurdle and a reason why they would resist and continue to try and get the camera-only solution to work rather than just like saying, you know, doing a why not and sticking the lidar. Now, look, I know for those that follow Tesla closely, which so do we. The explanation that they've given is, you know, if you know, the excess data and the conflicting data between the lidar and the camera, you know, can cause additional latency. And let me assure my listeners that the multiple engineers that we talked to, I constantly push back on those that, you know, that have the lidar solutions when they say, "No, no, I can handle." I'm like, "Okay, but if this, that, and the other thing, will it add latency?" And then there's like this admission of like, "Well, yeah, I guess there is latency, but is it really that much that it's going to make a difference?" Then I'm saying like, "Well, yeah, maybe it does make a difference if you're hitting a kid at 6 mph versus 1 mph. Maybe that fraction of a millisecond." So, I think, you know, the jury to me, I think it's still out on the original excuse, or not the original, maybe the second excuse that Tesla has provided on this kind of lidar decision, which is again these having these multi-sensor inputs. Obviously, most of these people are like, "Oh, it's going to be much safer to have lidar." And then we get into the debate of like, "Okay, does it have to be 100 times safer, or can it be 50 times safer?" Like, at the end of the day, like you get into safe cars sometimes and less safe cars on other times. There was no to me, my takeaway, I didn't hear anything from the smartest people in this industry that has convinced me that definitively like you have to have lidar in order for this solution to work. But obviously, getting data is still fundamental. Even going to autonomy 2.0, having training data, whether it's real use or the simulated data, which a lot of people were talking about as well, is still a critical part of this development process.

Yeah, well, I hear you, but I mean the one takeaway was again about getting data. And obviously everyone talks up simulated data, which is basically some verification of what Raquel said at Waabi plus this selling data plus being a trucking autonomy company. I will say this though, like and by the way, and Uber is investing in their own fleet of cars and I think they're going to be paying customers in there, but primarily to generate a data set. But what's also interesting about this data, like I said earlier before with Lucid and Nuro, is like there's like you have to retrain on whatever the new car is. Like if a car is a different size and these sensors are in different locations, that's going to require its own incremental training. That's where I think a lot of this simulated data comes into play. So I think there's certainly value that Uber is going to provide for its autonomous partners. There's dollars that can be made there as plus I think has demonstrated in the millions of tens of millions I think of dollars. I think your buddies over at Applied Intuition are creating, you know, some of this data as well. So I think, you know, there's value there and I think Uber wants to help their partners. Because obviously the ones generating the data out there are Waymo and Tesla, which don't seem to have the greatest relationship with Uber or its partners.

One more thing before we move on to some of the news that hit Waymo this week that we should talk about it, which is, you know, the potential size of this ADAS market. Again, this is not what we want, right? Grayson, you and I want level four, you know, autonomous robotaxis ultimately like private ownership, which I guess is level five, six, seven. I don't know. I don't keep up with all these levels. Being able to have go to sleep in your car, whatever level that is. I mean, this can generate revenue, right? This is a sizable market that OEMs want. Survivability is key for these startups. I'm already worried about someone like, you know, Lucid and like how does that hamper, you know, Nuro? Or like are they going to be a strong enough partner in the interim? Like so, ADAS I think can provide an opportunity. So, I know that there's a seeming pivot at Waymo, which we've, you know, I would say criticized in the past, but I think, you know, is embracing the ADAS opportunity a positive or negative in your mind?

Yeah, I mean, I don't you know, the demand, I think people would criticize like why is Tesla's FSD demand or penetration not like 50%, but if you think about the size of the market, if you even hit what Tesla with very little advertising, and really only getting super good in the last year or so. You know, that's still a decent number and a revenue opportunity for these OEMs. And again, something that the autonomy guys can use to kind of bootstrap them because, you know, as we progress here, we've talked about before like, you know, there's going to be consolidation. We want them all to win. We will call it out when someone's kind of moving ahead versus falling behind on autonomy markets. I also don't think that we've seen the end of the big dollars from Uber. Like I once had this kind of thought that like, okay, what if Nuro delivers 150 cars to the market in the middle of next year, and AV Ride and others were still like having struggling with the driver behind the wheel? Like would they get cut off from Uber and obviously not have the ability to fund? I think Uber probably wants to extend the competitive landscape there for some period of time. So, I don't think we've seen the end of Uber's dollars there. And obviously like on the flip side, as this gets successful, you know, there's not that market to finance these cars. So, there's going to be a lot of dollars spent I think in the billions, right, to purchase the cars as things ramp. Again, that's we're more bullish than I think most on seeing that ramp, but it's clearly in my mind at least not the end. Now, Uber will argue that like, look, they're still going to stick to their mantra of half of our free cash flow, which is a $10 billion-ish number, is still going to be used for share repurchase. That still leaves a lot of dollars to be invested in autonomy.

Yeah, I don't know. Like this market's wide open. I still think there's a good possibility that Uber ends up owning one of them. You know, and there still could be others that they partner with that would need the demand that their platform can deliver.

First of all, I agree with you that they should. This is all like the proper choice of words. I think they are far afield from going down that route. You never know. Like one guy could emerge, you know, and take the leadership. It's fragmented now. No one really I don't think does a good job quite yet. We'll obviously be tracking Avis, Hertz, the Uber partners, and other independents that are emerging. Speaking of Waymo, we had some news from them this week. Grayson, what's up with that?

Was the highways what does the highways have to do with the flooding or are these two separate issues?

I mean, to me this is all noise. Look, bottom line Waymo is the undisputed autonomy leader globally, in this country, everywhere, period, end of story. I mean, and my rides in the car are getting even more aggressive, more human-like. They're even developing the UI. A lot of times these companies will develop the tech. I'm still getting in the cars and it's like, you know, iPads that are kind of, you know, stapled on blocking the airflow that's in there. Uber, in my settings, the temperature is exactly what's in my settings in the app beforehand. The seat is set exactly the way I want it. It's clean. I know there's people out there that have suggested because maybe there's a one-off like soda bottle in there. I was in cars that had 80,000 mi, 100,000 mi and these things were, you know, a quality clean experience, right? So I mean, undisputed leader. So all of this stuff to me is noise unless there's like major I was personally, by the way, to our Waymo listeners, I was disappointed. I did not You did not turn my highway on and at the airport that rental car location didn't work. I guess I should have tried the Hyatt in order to try and get in and out of the airport, but I was kind of in a rush and not wanting to deal with those extra walks. Sorry, I didn't have the same Inspector Clouseau diligence that our friend Grayson here.

It was because of a time crunch, not because of, you know, of any laziness. By the way, the other takeaway from a lot of these conversations is and I think this is kind of logical when you think about it. The value of Waymo using their limited growth of supply to get a beachhead in existing markets and establish themselves as like, "Hey, I'm the first guy in, I develop the rules." So, like slap a couple hundred cars in a market, you're the person that's there first. You design the local rules the way you want them to, right? And then keep expanding markets before you worry about necessarily scaling up. At some point though, you get that critical mass and then all of a sudden 300 cars in a market turn into 1,000 or 1,500. And obviously, there's other issues that they're working on in terms of service depots to do that. So, it's kind of this like, you know, this spring that's getting kind of like wound up, right? In all these different markets and then all of a sudden Hyundai is going to come in with lower price cars, with a cheaper sensor stack and then bam! Like there's going to be all these cars in the market and those shares that some people have alleged are moderating. Well, we'll see what happens then.

Well, you should only look at my Twitter account at Walt's LightShed. And I tweeted a picture of a young lady who was in a bus in New York City and she was trying to hop out of the bus and the water just swooped her way down the street into the street drain. Yeah, obviously humans do this as well. So, if this is our biggest problems with AVs, I think we're going to do okay.

Yeah, I can't wait. We should get back there, you know. Now that I've been in a Waymo in California, I'd like to get in a Waymo on the streets of London. My understanding is the same software load wherever you are. So, but you know, good for London. I don't know, is this in any way related to some of the political changes that are occurring in the UK overall, do you think?