Walt, we're back from CES. I took the red eye. And of course, our listeners and viewers know you took the gentleman's flight. And when I got back from CES, look what I got. I got the clapper hat. So now I'm going to harass you with the clapper hat. Also, if you say something nice, I'm going to clap for you. So thanks to Hugh at KPMG for the hat. Really appreciate it. But now on to autonomy. We saw a lot of interesting things. Had a really good intimate meetings with folks. And people came up to you, asked you for selfies. You're pretty famous, Walt. Then there was the chatter of CES was the Zooks debacle. What's that going to lead to? Waymo officially unveiled Ojai, which is the Zeekr. Mobileye made some news. So, let's break down CES to start off the show. What were your takeaways?
First of all, to your point, you know, which is always the case with CES, it's just meeting great people. We had obviously a lot of time to spend, you know, some quieter time with some of the key leaders in the industry, both trucking and in the autonomy side of things, car side of things, OEMs as well as tech stack, you know, chips, whatever. That's why CES is perfect. It's just a bunch of people, you know, coming together. And again, to reiterate what Grayson said, thanks for everyone that kind of came up to us and gave us the feedback in person, which obviously tends to be more humanized than sometimes some of the other feedback that we get. But the main takeaway I think is just that I'm even more bullish on the future of autonomy and the opportunities that exist for multiple companies. This is not a one or two player race. There's a lot of great technology that's getting developed, a lot of people that are very smart, have great backgrounds, that are putting a lot of hard work and bringing their technology to market. And regardless of what's in the market today, which we're all very bullish about, I think the opportunity going forward is pretty significant. There was a darker side though, Grayson, and maybe you can outline some of the things that happened that were disappointing, you know, in terms of CES and at least one of the members of the autonomy community.
It was a darker side and it didn't involve going to an establishment. It happened right there on the strip. And I'll say this bluntly and I have already said this privately to Amazon. It was a disaster for Zooks. Vehicles were breaking down consistently. I didn't have one trip where I went from one end of the strip to the other where I didn't see a vehicle broken down. And the night that we had our little shindig, and I have a video which I will share with us, a Zooks vehicle was blocking the entrance of the Venetian causing all sorts of traffic. So I don't know. I got to ask Walt what we experienced. Is this normal for Zooks operations? There's a lot of questions. A lot of questions.
I mean, we've both been in Zooks at their headquarters. Myself, I can't remember. It's probably like 18 months ago now. I actually went, I left my hotel. I walked to a location which was a pickup spot. So, just to be clear, like the ODD, I think those are the right acronyms. I'm sure I botched that in the past, which we've heard from our listeners about. The geofenced area of the Zooks. It's not like you just use the Zooks throughout that. You have to go to like one of, I think it was five or six pickup spots. So, anyone that's been to Vegas knows it's not easy to get around. So, I was walking over to this pickup spot. As I'm walking, I'm saying the app doesn't even open until 11:00 a.m. So, I sketch, you know, I time it. So, I just walk up there 11:00 a.m. There's like five other people who had walked there that was trying to test it. As a result, I have a 45 minute wait to get in one of these Zooks. So, I go in to Topgolf, grab something to bite, get in the Zooks, and the ride hadn't improved in two years. And you know, you and I, as our listeners know, get in all of these cars. That's one of the best ways to get a sense of what's going on. And it was just a very herky-jerky ride. Like it was not smooth during the course. Did you even try to get into one of these Zooks racing?
I tried multiple times and I was like, 'Oh, since you told me the 45 minute wait, there was one post that went viral on X, an hour and a half wait, so let me order one before I walk over there.' You can't even order one before you walk over there. It was inconvenient for me. So, no. And I did not try one.
My other complaint, while we're in the complaint section, is now granted I like, I can't play... You're clapping. You're using your clapper hat for my complaints or was that an accidental clap? I can't play first-person shooter games with my kids. Obviously some parents think no one should play first-person shooter games with their kids, but in my case, I get nauseous. But by the way, side note, you know, I saw an incredible Samsung 3D TV where for the first time ever, I was in like a first-person shooter and I did not get nauseous. That's just a side note. When I got in the Zooks, like you could, you have the side windows, but the front and back windows are effectively covered by some mesh. So, you can't see out the front of the window while this thing is driving, which is just not for me. Maybe everyone's not like this, a positive riding experience on top of the fact, you know, that it was herky-jerky. So, I think that just the experience itself, given how long this product has been around, was disappointing to say the least.
I would say that's an understatement. And the Uber drivers I spoke to said that this is a common occurrence of how these vehicles do not operate in a well manner, which raises the question, Amazon's got to make a change. This was not a good reflection. What do you think they're going to do? I mean, you know, for our Light Shed top 20 predictions for 2026, I think I specifically said they need a big shakeup there and maybe whether that's an acquisition of a technology company and as you know in the process of doing that to kind of maintain their space, maybe there's some benefit of having Lyft as part of the family, who knows? But something has to change at Zooks. This is not suitable for where Amazon, you know, a company of that stature should be in the market.
So, do you think it's an acquisition? Is it a management change? I'll just tell you point blank. They got to do a technology overhaul.
All of the above, Walt. You know, we're going to be right if they do even one of these things, but it's an all of the above. And maybe how they can justify it and say like, look, we'll have this parallel path of continuing to develop this purpose-driven, you know, box car and then at the same time maybe buy one of the existing technology guys in the market as a way to kind of partner with OEMs and have a parallel path. You know, maybe that's the situation. I don't know.
Or could we see a situation where Amazon just pulls the plug and said it was a mistake?
I mean they certainly could do that. Apple had an autonomous car program which they pulled. I think that was a mistake by them given the size of this market opportunity and the multiple products that you can sell. I mean just to jump forward, JAC, which is a Chinese auto manufacturer, was showing, you know, their car. This is similar to what they showed last year and how you can have this very comfortable seat, 50-inch television. I mean, you know, this speaks to the bullishness that we have and what autonomy can enable. But you would think that Amazon for a variety of reasons, that what, you know, their presence in media, the presence, you know, in terms of AI and the investments that they're making there, as well as just delivery, right, and the crossover of it makes sense for them to have real product. Zooks, sorry, it just isn't it. And it just, you know, to be at CES and be known for the car that everyone was seeing breaking down around CES. I mean, you know, maybe they'll provide stats to refute what we're saying, but like I saw myself multiple times. It was just sitting there. Same thing. Taxi drivers telling us the performance of these things have been problematic. And like even when I was in the Zooks itself, there's no button to say press it to like pull over on the side. Like if you're in a Waymo, you can press a button and it will pull over wherever you're going because it's all in these point-to-point locations. So it's just, I don't know. That's the dark side. Let's get more positive and go back to the positive.
Before we get to the positive, I have to say it's a poor reflection on the Amazon brand since Amazon's innovating on so many other things. So hopefully Amazon can right the ship. But we do know, or I'll say we think that change is probably coming at Zooks. The big news obviously of CES as always is whenever Jensen of Nvidia speaks then everyone listens. There's a couple different things he talked about as it relates to autonomy. They introduced this, I guess you call it Alpamayo. Alpamayo, I forget how the pronunciation is, but it's open source reason-based vision language for those of our, you know, people that love to debate the lidar versus vision language model looking at long-tail autonomous driving edge cases. They also put in, I think, simulation tools in there. They tied it to the Halo safety system. I mean, it's new, but is it really new? I mean, Nvidia already released Open Reasoning on, I think, early December. They had this Alpamayo R1 thing out there and that was already framed as open source. We know, we've talked about on this podcast before how, you know, we think that they're trying to go OpenStack. There's been my long thesis of as your big companies, you need to be vertically integrated. So I mean the only nuance I guess is that Nvidia is explicitly positioning it as a teacher model rather than running like directly in the vehicle meant to be, you know, distilled into like these smaller stat runtime stacks. So for the bulls out there, if you look at how the stocks reacted, like the concept is, oh, open source. Oh, Nvidia, like this is like the Android for autonomy. So this is like Uber's dream because you've kind of democratized autonomy because Nvidia is just going to enable all these, you know, different groups of people or different companies, you know, and challenge and create the fragmentation that Uber needs and that the industry needs to get autonomy to the next level. What are your thoughts on what this really means for the industry?
What this really means is Nvidia's looking to vertically integrate automotive, but at the end of the day, they're looking to sell more GPUs. Nvidia has open-source models in the past for autonomous driving stacks. The major, major, major, major difference in this model, this model was put under an Apache 2.0 license that you can use commercially. The kicker on this is on the previous model that they open sourced, you had to sign a legally binding contract where you could only use Nvidia GPUs and you were subject to audits by Nvidia. So, it's a very restrictive Nvidia license. This has an Apache 2 license. So, you can do a lot more with it outside of the model weights and it's getting traction. It's been downloaded 15,720 times on Hugging Face. So, I'll give them a clap, Nvidia, for that. So, there's clearly traction there. Where this ultimately goes, I could see more restrictive terms coming back, but at the end of the day, Nvidia wants to sell more GPUs as Qualcomm, ZF, ARM look to enter the autonomous driving space. I mean, so my question to a lot of the companies that were there that are using Nvidia is like, what does this mean for you? You know, if the market thinks that this is democratizing it and Androidizing it, you know, creating competition that theoretically is not great for, let's say, a Nuro that's using their compute. Now, the answer is, I think for many of these companies is like, look, and I think even though they didn't say this, like it was good for a lot of companies. Everyone wants to like get on stage with Jensen and announce a product because it's that nice halo effect. Obviously, the compute is great product. Not many, if any, of these companies have bought more than the compute and some of the software that they're already trying to sell. Obviously, Nvidia can go directly to the OEMs with this type of stuff. And some of the OEMs are trying to develop this stuff on their own, but that's like a moilized type approach. And I think there's a lot of criticism in the industry saying that these companies that start with like, hey, we're going to go level two ADAS with the OEMs directly is a very different thing than going to level four. Right, this is the criticism of, I think, Tesla's approach and, you know, where maybe Mobileye is and even Wayve, who by the way had like no presence at CES, nothing to be seen of Wayve, given the importance of autonomy right now. So I mean I think that is the kind of the pushback. The challenge that the industry is going to have is like, you know, Nvidia, like Tesla, if they're entering a market, everyone, it becomes an overhang, which is a positive obviously for Uber, when the reality is I'm not sure, like what are we supposed to believe that Mercedes-Benz is going to all of a sudden become like the leader in autonomy or like GM is going to come out of nowhere and partner with Nvidia and have Sterling Anderson like have an OEM develop autonomy possible and we'll follow those things as they take place. But in the meantime, the companies that are in the market today are moving towards delivering cars this year, not in a couple of years.
Yeah, we had the announcement also from Ford where they're developing new advanced autonomous driving systems. So, the OEMs are coming back and there's a lot of partners to dance and what role Mobileye plays, who knows, what role Wayve plays, who knows? And at the end of the day, is Nvidia going to be a customer or are they going to be a competitor? And I think that's something that we're going to have to continue to watch because we don't know yet.
Yeah. At the end of the day, I think they're going to be fine. Let's say Nuro is wildly successful, that they sell a ton of, you know, a ton of product. They're still going to be using a bunch of the Nvidia compute in there and that's still going to benefit from those companies. So maybe they're just, I don't know, whatever the reason is. Look, they also overshadowed Mobileye was trying to hype people up for the announcements that they had there. And basically the Mobileye announcement overshadowed that, right? Because now you're saying like, okay, Mobileye is effectively trying to help the OEMs enable level two and now like Nvidia is going to say, well, we're going to open source this and, you know, this is going to be the way to level two. But then the Mobileye's message boiled down to this, that we're going to win in the pragmatic lane. Cheap, consolidated L2 plus that OEMs can standard fit, robo taxi timelines are still alive. And basically just get it in the cheap cars as opposed to just everything. So the two threats that exist obviously is the OEMs like, you know, Sterling Anderson and GM that want to do it in house. I don't know whether they're using Mobileye or not and obviously Nvidia pushing a more complete stack. So, you know, they had an announcement with a big US car maker to put this in the driver assist system in every version of the vehicle. Again, underscoring it's not just for the high-end cars, but basically just everything standard fit, right? Trying to move it this typical Mobileye style. But again, like, you know, we'll see how many auto manufacturers actually adopt that and use that and generate revenue for that company.
But why give these vague statements a major US automaker? Why not, if the deal's signed, why not publicly identify who that is? Because that can help us analyze that market.
Yeah. I don't know. And look, at the end of the day, like the dominant player in level two plus right now is Tesla. You know, you and I use it every day. It works great. You know, we believe that it can get to level four. Obviously, there are those that believe that won't happen. We'll know, I think, soon enough. But this Mobileye approach or the Wayve approach or even the Nvidia approach are seemingly not taken seriously by the rest of the industry. A lot of people like say just this is a very different thing than perfecting level two versus getting to level four and Wayve is in that bucket and so is Mobileye and obviously so is Nvidia at this point.
We're going to see, we'll have data points on Wayve next year when the Nissan ProPilot starts coming off the line.
And when is that expected and do you think that they're going to deliver on those timelines?
Based on reporting from the Nikkei out of Japan, yes, I do believe they will. That'll be in 2027 in the ProPilots.
Based out of reports, you're relying on the Nikkei to gauge your confidence level on whether Wayve can deliver. And what are they delivering? Wayve, you know, level two or level four?
Great. Well, so let's move on to level four players because that's what autonomy is all about. And Waymo obviously the existing leader in the market with several thousand cars. You know, the criticism that we've had here on the Autonomy Markets podcast is the volume and so they showed this Zeekr, this Chinese-made car. They basically just rebranded. I have a video. I think I put it up on my Twitter account. They literally put a decal on it. So maybe this was like a very recent brand name that they came with. Oh, I think it's pronounced Ojai. The car itself looked a little bit better. You still have these gigantic sensors. I believe it's still a fit. I don't know. It doesn't seem like it's coming off the manufacturing line. You were on the trade show floor as well. What were your impressions of the Ojai?
Big, bulky, and ugly. That's the first thing that I looked at. If you look at it from the rear, you'll notice if you want to call them the sensor pods, they're all very bulky. It's not a good looking vehicle. And it's Chinese at the end of the day. And where geopolitics are going, no matter sticker or no sticker, it's still a Chinese vehicle and I don't find it to be very attractive. I know there's a lot of engineering that went into it, but still it's not very attractive.
I don't care how it looks, Walt. I mean, if it gets me somewhere and the ride is comfortable, unlike the Zooks, which has a similar boxy look, it has a front window that you can actually look out of. It was a little odd that Waymo didn't allow us to look inside the car because, you know, a couple of booths over was the Zeekr, which I think I mentioned before, opened up. They had these great comfortable seats and, you know, it showed the very comfortable interior. If you have the interior that Zeekr was showing, which makes this thing, then who cares how it looks on the outside? This is all about your comfort and safety on the inside.
It's a very valid point. It's still ugly, but very valid point from you. It was also great. The Waymo had a great booth.
Yeah. Well, we can switch that then. The Hyundai also looked good. So, but let's weave the Hyundai in to or Lucid. Well, yeah, we can talk about the Hyundai now. It looked good. It was the, you know, what the production model should look like. Again, it underscores, you know, what we're hopeful with Waymo. Just to, you know, for our podcast listeners now on the light side of things, I predicted that Waymo will hit 6,000 cars by the end of 2026. That's the prediction that I'm on the tape for now. As well as getting to production, but I think they're going to compete with Lucid and who can get to production faster because the Lucid car looked awesome, like better than anything out there. I mean, the quality of it. And by the way, they let us sit in the Lucid car, unlike Waymo. High quality. They had representatives from Uber, Lucid and obviously Nuro there. It sounds like they can get to production by the end of the year. So like here we are, you know, Waymo who's been in the market considered the dominant player and like maybe Lucid gets to a production level car before Waymo. Do you think that's possible?
Well, this is a clap moment to the Lucid Nuro team. Beautifully designed car. The integration was flawless. It was well thought out. And here's something that a lot of folks don't talk about. This is massively inside baseball. Lucid has one of the best engineering cultures out there. And what we saw with that product was a clear example of what Eric Bach and the engineering team are doing at Lucid. So well done Lucid. And depending on capacity, and I don't have the exact numbers, capacity that they can do in the Arizona plant and they can also do with the Riyadh plant. There is a very clear path for Uber, Lucid, Nuro to scale this pretty significantly.
I mean, the prediction or the commitment, I guess, is 20,000 over what is it, five or six years. I suspect it's going to be faster than that based on some of the conversations that we had. Lucid, you know, obviously a smaller player. I'll also kind of acknowledge that, look, what you show on the trade show floor isn't necessarily what ends up getting made. So we can't speak too much about like a trade show car, but look, it looked good. I think the price point is good. I think the roadmap to getting lower price Lucid is good in Nuro's case. I think they probably got line of sight on a second OEM beyond Lucid. This is all good news for Uber, the Uber team in terms of pushing forward a technology and it just makes me think that like, look, I know how Uber has been fairly restrictive of their capital to this point and it still makes sense for them to not kind of own the stack, but you know, over time as you know, maybe like a Lucid-Nuro partnership succeeds and others like Wayve maybe fall by the wayside or just stick to level two ADAS that it makes sense I think for Uber to ultimately buy one of these things. We're talking multiple years down the road, a lot of things can happen. But if like the Waymo relationship, you know, continues to sour with Uber, you know, Tesla obviously doesn't do anything with Uber. Obviously, there's people out there that think that Tesla will never do anything that basically, you know, the winners of the others, you know, could end up basically being owned by Uber. Thoughts?
