Voiceover:
Welcome to TD Cowen Insights, a space that brings leading thinkers together to share insights and ideas shaping the world around us. Join us as we converse with the top minds who are influencing our global sectors.
John Blackledge:
We're live from TD Cowen's 54th Annual TMT Conference. Really happy to have Alex Roy, a general partner of New Industry Venture Capital. And I'm with my colleague, Itay Michaeli, our autos analyst. I'm John Blackledge, internet analyst here at TD Cowen. This is going to be a great podcast. Maybe I'll kick it off, Alex. Thanks for coming. Just current thoughts on the state of autonomous industry?
Alex Roy:
Couldn't be more bullish technologically, but the political dimension or maybe the sociocultural dimension is really coming into play recently. I live in Arizona, and there's Waymos all over my neighborhood. I've got a Tesla. It is getting better despite my criticism in the past. But wow, I'm amazed at how aggressive some political voices have been about AV. It's in primetime and how suboptimal, would be a nice word, for how Silicon Valley and the AV companies are handling the messaging around their quite obvious technical success.
Itay Michaeli:
Alex, I know you ... Welcome to the program. Great to get to see you. You went coast to coast with Tesla FSD, and you've of course tracked this system for years. Talk to us about the rate of improvement you've seen in your personal vehicles. And from based on that, what do you think the next couple of years look like for FSD and Tesla Robotaxi?
Alex Roy:
If you're referring to my zero-intervention cannonball drive from a few months ago, I was surprised that it worked as well as it did given that we went at the very end of winter through a snowstorm. It was really good. It was our fourth attempt in a year. And there was a lot of progress, obviously. And since I've gone, someone else has already gone and done it seven or eight hours faster, and that's primarily because they went in clear weather. And so there's obvious progress on the Tesla FSD side. There's no debating it, but it's also very obvious that the performance is highly ODD dependent, and that's why Waymo is driverless in ODDs where it's optimal, and Tesla is not driverless anywhere except in a very constrained domain in a couple places Waymo is currently operating. So, Tesla will get there. Waymo is already there.
The only real question is, who can afford to scale operations while simultaneously driving down the costs? I'm not referring to the sensors. We'll get to that. The rate of improvement has to correlate to the ratio of people in operations to driverless vehicles. If that doesn't move at the same rate as your cost as you scale out, you may not be in business in five years, and it's a very dangerous place to be.
John Blackledge:
On that note, just curious your thoughts on the competitive set. You mentioned Waymo obviously in the lead. You mentioned Tesla, and we also have the Chinese players. I don't know if you have a tiering of Waymo up top, and then how would you rank maybe some of the other players?
Alex Roy:
If one's just talking about the US market, one can't rate any of the Chinese players because, politically, I don't think it would matter if the Chinese companies were perfect and free. The political dimension of the United States may prevent them from deploying here for a long time, if ever. So, in the US, that leaves Waymo, Tesla, Zoox, Motional, May Mobility. It's hard to classify May Mobility because we haven't seen them operate at scale driverlessly anywhere.
But, Zoox suffers from maybe a surplus of engineering talent and being very late to deploy at scale. I mean, they have, in theory, unlimited resources. And they have some wonderful leaders. The engineering leaders are great. And Aicha Evans is a wonderful person, but we hadn't seen from her the level of aggression that you see from Elon, not in her statements and not in the pace of rollout. So, Zoox currently, it's impossible to rank. They might be nowhere. We just don't know.
Then, they have a second dimension issue which is the vehicle itself, which is super optimized for low-speed urban environments, not really optimal for highway driving. And the configuration, as cool as it looks, and I've been in it, it's not the first thing you'd want to get. If you had a choice of a Waymo or a Tesla driverless or a Zoox driverless take you from San Francisco Urban to the airport, SFO, at 65 miles an hour, you probably want to sit facing forward. It's just cultural. So, Zoox, can't say.
The Tesla Waymo. Waymo needs to keep hitting milestones to keep raising to get to IPO. That's it. That's the entire game for them. I don't know what their milestones are, but they're clearly scaling on operations. So, the clear leader in capability and the UX. I mean, you ride in a Waymo ... I'm an Arizona, John. You're in Arizona. It's great. It's a great experience. And the progress is clear.
Tesla. They're hiring operations people in dozens of cities, so they clearly have a roadmap. And so the only real question is, can Waymo IPO get to profitability or meaningful revenue that they can keep raising before Tesla can go driverless with some level of perceived safety in the realm of Waymo? Waymo's got to do that before Tesla gets there. Because if Tesla gets there, that's a dangerous position for Waymo. They need to have IPO before that happens, I think.
Now, Motional, like Zoox, is just unknowable. Hyundai backs them. I know what their risk tolerance is, their financial patience is. It's unknowable. And Motional and Zoox are both ... In terms of public awareness, they're almost nowhere right now. There is one more thing, and I think I talked about this last year. The wildcards remain Mobileye and Wayve, the UK company. If these companies mobilize the majority of vehicles sold outside China for driver assistance, and they're whispering in the ears of the CEOs of every OEM on Earth, "We will give you L4 on some timeline," if they could deliver that before Waymo can miniaturize a hardware form factor and license those OEMs, Waymo's boxed into Robotaxi. If Wayve, which has dominated in leadership, dominated by former Mobileye executives, pretty much all of them except the founder, Alex Kendall, if Wayve gets there before Waymo can miniaturize the hardware form factor, it's another problem.
And so these are the quiet, potentially wildly dangerous, powerful players that are out there because if they can flood the market with factory vehicles, conformal sensors, that are L4 capable in three to five years, Tesla's not going to be licensing FSD to anyone, and then Waymo has to overcome that. And then you're at a market equilibrium. But instead of Boeing and Airbus, you got five players of roughly comparable capability fighting it out on costs and price. You don't want to be there. That's not a good place to be.
Itay Michaeli:
Alex, you mentioned earlier ODDs and scaling. And I'm curious your thoughts on other AV verticals outside of Robotaxi. We're seeing momentum in trucking, freight logistics. How do you kind of rate some of the other verticals out there in terms of scalability, the ability to get cost down, and ultimately the business models within those verticals?
Alex Roy:
All right. I mean, the ones that come to mind, as an investor, trucking and construction, obvious ones. I'm interested in defense, although I wouldn't say the defense autonomy ship has sailed, but there are already some big players there. So, in trucking, the slow surprise has been that we have not yet seen a meaningful deployment of driverless trucks and that the players from whom you would expect to see that deployment haven't done it.
And so the obvious one's Aurora. He's Chris Urmson. He's a celebrity in AV. Aurora should be the Waymo of trucking. It's not there yet. And that leaves five or six other players, some of whom have declared dual-use capability and done some defense work. And so still waiting to see it. It's also inevitable, but that's a surprise. I mean, I would've expected to have seen it by now and not just here but in China. I mean, a ton of announcements with China about Robotaxi. What do you hear about China and AV trucks? Nothing. So, that's interesting.
And the other one is then construction. Boris Sofman, ex-Waymo, he's got Bedrock Robotics has raised a mountain of money for AV construction, and there's a bunch of others. So, that's going to be huge. It's happening, but it's not really something that people are really taking some notice of correlated to the work that's being done. And I'm thrilled to see it because if the United States is going to be exporting technology for the next 50 years, I'd like it to be something that generally improves the world. We need to build stuff.
John Blackledge:
I'm also curious on your maybe updated view on the cost of the AV. Just take Waymo, for example, with them pivoting to Zeekr and then also the Hyundai and going from driver five to driver six, if they're on some path to materially lowering all-in cost.
Alex Roy:
I mean, Waymo's definitely on the path to lowering cost. And behind the scenes, I'm quite sure that every weekly executive-level meeting is, "What's the rate of progress in cutting these costs?" It has to be. It's the sensor cost and then the cost of operations. There's a point beyond how much more you could do to reduce the cost of sensors, but the operations you'll always be able to cut somewhere, and it's totally dependent on the software performance. And so it's a three-legged stool. And the legs all have to be roughly the same. And so you can cut all the costs you want one dimension, but the others don't go in the same direction, you're in a bad spot. So, I'm an optimist for Waymo because the pressure on them is so enormous. I don't think Waymo's going away. I think they're a forever company. But whether they are dominant or at equilibrium depends on the rate of those cost cuts.
And again, unknowable. Absolutely unknowable. I mean, the wildcard for Waymo isn't even how well they do their job. It's whether Tesla can get to a perceived level of safety that's perceptibly similar to Waymo. If the public thinks they're equally safe, Waymo's got a problem. That's not great. And that might be associated only with time. So, my daughter's seven. She rides in Waymos. She calls it Mr. Robotaxi. She thinks it's completely normal to be in a Waymo. I'm like, "Oh, do you think it's safe?" She's like, "Well, of course. We're in it, right?" That's how she measures safety. So, by the time she's of driving age, she's already aged into believing Waymo's safe. I have a Tesla too. She asked me why I'm in the front seat. Because I routinely don't touch the wheel. She thinks it's also driverless, but that I'm dumb for sitting in the front. She's seven. We're four or five years from a generation of people who will be paying out of their allowance or their parents' credit card to ride in these things who think it's already fine.
So, that's the clock for Waymo and Tesla to figure out, to determine if one's going to be dominant, or if they're going to fall to market equilibrium.
Itay Michaeli:
Curious on your thoughts 10 years from now, how far the Robotaxi rideshare market can go in terms of people owning cars. Maybe they're owning autonomous vehicles, level four vehicles, that can do a lot of things for them as opposed to using an app to access rides. What do you think? Is it a split between cities and suburban areas where cities are much more dominated by rideshare, and suburbans are ownership? Or where's that line kind of draw for you in terms of usage of all this autonomous technology?
Alex Roy:
I'm going to make a lot of enemies with this statement. So, in major urban centers that are center that are purple or red in this country, there's growth. And anyone who's going to the airport I think will enjoy it enormously in any city, anywhere, always.
However, in the United States, in a purple state or red state, the suburban and rural bands, I don't believe that the Robotaxi market will make any difference or make any inroads and replace cars at all. I mean, as someone who grew up in New York and moved to Arizona in the last five years, it wouldn't matter if Waymo is free. I'm going to own a car. It's baked into the American consciousness, and the further right you go in that political consciousness, the more you will, you must, own a car, even if you don't need it.
The SUV and pickup market exists. And as enormous as it is, it's not merely a function of tax breaks on the weight of a vehicle. It has to do with the American requirement for ownership of potentiality. I don't need to own a truck, but I might need it someday. I don't need to own a gun, but better to have it and not need it than need it and not have it. And one doesn't have to like that logic. It's baked into the American consciousness. Now, that's radically different in Europe. And I've lived in Europe, where the transit system is awesome. Even in small towns and cities, it's awesome. And so it's entirely possible that someone who owns a car now will be on the fence and then go use the driverless serves, but that is also dependent on the price. And there's no evidence that the savings from the removal of a driver will correlate to a massive reduction in price. Those profits are going to get taken. They have to be because so much R&D's got to be amortized.
I'm so bullish on autonomy but bearish on its replacement of vehicles outside some markets. And my last statement would be here is any replacement of vehicles goes out the window if the second phase of autonomy is it works, and Robotaxi scale's great, but eventually Wayve, Mobileye, Waymo are licensing their tech to go and personally owned vehicles. It might increase the number of vehicles sold because everyone who doesn't really want to drive or really can't drive who's older, they can still keep a car. So, if my mother was still alive today, and driverless was possible, I'd buy her two.
So, that's a fantasy.
John Blackledge:
That's great. Well, I think we're going to wrap it up. Thank you for the podcast and for participating in the conference.
Alex Roy:
Thanks so much for having me, guys.
Voiceover:
Thanks for joining us. Stay tuned for the next episode of TD Cowen Insights.
Ajouter un Signet
