Travis Hoium: Really depends on how the business models play out. If we have somebody building their own fleet and they're also the person or the company that is connecting with the customer, so you're calling a Waymo, then Waymo is going to win the entire market. I think there's going to be a lot of value that accrues to building scale. The question is, can a company like Uber be the company that connects dozens of different autonomous companies? Because then that may distribute the value a little bit more or maybe it just all accrues to Uber. My answer is probably, we just don't know exactly who that's going to be quite yet.

Ricky Mulvey: This is a space where a lot of cash has been burned. There's been a lot of promises that have not come true on the timeline that folks were hoping for them to come true. Why are you interested in investing in this space? Why in 2024?

Travis Hoium: I think we have a turning point here where autonomous vehicles have been on the road for a number of years now. Waymo and Cruise are the two biggest in California that actually had commercial operations. Cruise has been shut down for a few months now. But this is actually real, and it's happening, and it's not something in science fiction anymore. When you look at the potential opportunity here, it's just absolutely enormous. In the US alone, we drive 3.2 trillion miles per year. Uber had 10.4 billion trips and took in $150 billion in total bookings. That's just Uber. Think about if you replace one of your vehicles in your garage, if you have more than one, or maybe all of your vehicles with autonomous driving. I always think about this, I have young children, are my kids ever going to need to learn how to drive? Well, if there's autonomous vehicles available, the answer may be no. Now you're talking about a trillion dollar market or a multi trillion dollar market. That's something at least worth paying attention to.

Ricky Mulvey: Now that we've gotten some of the excitement there, why folks might want to pay attention, let's get to the groundwork. There are different levels of autonomous driving. Can you walk through what they are, and where the big players are standing?

Travis Hoium: There are Level 0, which is basically no autonomy, all the way to Level 5, which is basically a vehicle can drive anywhere. Level 1 is the features that you may have in your vehicle today. Basically, the vehicle will take over with minor things, will adjust you in a lane, it will do smart following, so adjust your speed. Level 2 is going to be more, the vehicle will drive itself, but you are in charge. Actually, Tesla's FSD, is a Level 2 autonomous system. The big distinction comes between Level 2 and Level 3, because that's when the liability of an accident moves from the driver, the human in the car, to the actual auto maker. That's why Tesla is a Level 2 because any accident that happens in a Tesla, the driver is technically at fault. Level 3, now you have to be available to take over, but you can basically not pay attention to the road anymore. Level 4, now you get to full autonomy, but it's limited, typically geographically limited. They call this geofencing. Waymo, and Cruise, and Zoox, these are all geofenced technologies. You can't take a Waymo out of the San Francisco area, for example, if you're driving around in that area. Level 5 is what Tesla is trying to get to, and that is going to be basically, I can get in a vehicle in Los Angeles, and I can say, drive me to New York, and it will just do it for me. It can go anywhere and can basically learn on the fly, it doesn't rely on that geofence. That's the range there, and at least a few of the companies and where they fit.

Ricky Mulvey: I'm surprised that Tesla is only at Level 2. That sounds like you're saying that Tesla is behind a lot of these other companies maybe with more quieter efforts. This is a company that has millions of hours of driver footage. It has a program called Full Self-Driving.

Travis Hoium: Right. But they recently renamed it Full Self-Driving (Supervised), because it has to be supervised, and that's where we are today now. Ultimately, the goal, Elon Musk and a lot of investors think that they will eventually get to Level 5. That is not where they are today. I think we just need to make that distinction. Now, we'll see what they announce. Supposedly, this event is going to happen on October 10th, or robo-taxis. We don't know exactly what that looks like. But right now the data that we have, and it's mostly third party data, is that, there's still about 300 miles between each disengagement in a Tesla FSD vehicle. You can't have a disengagement that often and let a vehicle go on and drive by itself.

Ricky Mulvey: A disengagement every 300 miles? What do you mean by that?

Travis Hoium: That would mean you have to adjust the steering or you have to break. Basically, the system doesn't sense things the way that they should. Like a Waymo or a Cruise vehicle, for example, you wouldn't be on the road if you had that many disengagements. This is actually why Apple shut down. It was in the neighborhood of a disengagement about every 50 miles. Waymo and Cruise, you're now talking about hundreds of thousands of miles that they need to go without disengagement. Just to put some context to this, Mobileye is one of the companies that's providing services to other companies. Their goal is to not have a disengagement for one million hours. That is about 114 years, just to put context to where Tesla is today and where some of its competitors say that they're going.

Ricky Mulvey: We haven't talked about the Chinese companies like Baidu. They are also in this self-driving race, perhaps more internationally. There's a lot of regulatory concerns with having Chinese self-driving cars in the United States. But where do they stand? How do they stack up compared to the many of the American companies that we've just described?

Travis Hoium: I would say that they're all in the same boat moving in this direction toward typically Level 4 autonomy. You're right, Mobileye's the company that is impacted because a lot of these companies are either putting pieces together to build their own autonomous driving system or developing their own autonomous driving system. Some of these are a little bit opaque as a US investor getting information about exactly what's going on in China. I do think it's pretty apparent that there's going to be a split between the Chinese market and the US and probably European market, the Western markets, if you will. That may come down to regulatory concerns. We've seen that Tesla vehicles were not allowed in certain areas in China.

We have also seen that the US, now, I think, just this week, said that, hey, maybe we don't want to have any of these autonomous driving or software features in the US if they're coming from China. It seems like they're splitting, but on a parallel path. I would say that there are companies that are testing similar technology to Cruise and Waymo in China. But the features that are actually on the road are more like Tesla's FSD.

Ricky Mulvey: Let's focus on Cruise for a second, and specifically, General Motors. They have Level 4 autonomy. They have put robo-taxis on the road with mixed success. I think they just got booted out of California after some accidents. But they at least got robo-taxis on the road. General Motors also trades at five times earnings. It's getting absolutely no premium for this self-driving technology. What's going on here? Why don't you think GM is getting any love for its self-driving tech?

Travis Hoium: That's a good question, and I don't have a great answer for you. I think Tesla is the one company that is getting that premium, and it's not really clear why when you look at, like I said, the facts on the ground, they don't have a Level 3, Level 4 or Level 5 feature, they're not even testing any of them. We can see that testing data in states like California and Texas, and those would be the two places that Tesla was headquartered. I think that the question that investors are asking about a company like GM is, is this going to be a cash burning business for the foreseeable future, or is it going to be a cash-generating business anytime in the near future?

If you're skeptical of GM being a autonomous vehicle company and you say, what I'm seeing in the financial statements is they're burning a billion and a half dollars per year funding Cruise, I don't like that as an investor. Just give me the cash, buy back stock, that will be a better return. I would rather put that autonomous money into a company like Tesla. I think that's what investors are thinking now. If you are looking at the optionality with a business like Cruise or technology like Cruise, I think that provides a really good risk reward profile. Like you said, you're getting GM, the company, for a relatively cheap multiple, and you essentially get Cruise for free.

Ricky Mulvey: Mary Barra is buying back stock at the same time that GM is investing in Cruise. The next rollout, so they shut down the Chevy Bolt for a little bit, which is their electric car, this is going to be the next big rollout for the Cruise technology. That's coming in 2025. Getting some hype in the GM earnings calls. What are you expecting from the next iteration of the Chevy Bolt?

Travis Hoium: I'm expecting this to at least be something that is scalable. If you look at most of the autonomous vehicles today, the Cruise, Chevy Bolts that they were modifying, the Waymo vehicles today, it looks like they just took a regular car and they stuck a whole bunch of equipment on it. That's not going to be what we're going to be doing in the future. GM has put off their Cruise Origin vehicle, that's that little miniature bus. Actually Zoox has a vehicle that looks very similar, Zoox is owned by Amazon, which is like this miniature bus with doors that open to the side. Seems like a really cool vehicle. You could drive a wheelchair into it, but it doesn't have a steering wheel, and it doesn't have a brake pedal. They did not have clarity from regulators that they would actually be able to put those on the road and actually build them at scale, because if they're going to build the Cruise Origin, they don't want to build 50 of them, they want to build 10,000 or 100,000 of them. They can build 100,000 Chevy Bolts, and this will be the same vehicle that you could be purchasing from a GM dealer, it's just going to have a little bit different technology on the inside, but it is going to drive itself just like the old Chevy Bolt. I'm looking for, can they hide those things a little bit more and build the technology, the LiDAR, the radar, the vision cameras into the car, so it's unnoticeable for most of us.

Ricky Mulvey: I'll be curious to see if they can expand Super Cruise. Right now, Super Cruise works pretty well on highways from what I've seen in the reviews. The question is, can they bring that to the cities? We've talked about the car builders, there's another side of this, which is the subscription service. Are you seeing all these self-driving technologies going to that subscription service model? Tesla, you can pay, I think it's like eight grand for the now supervised full self-driving. A lot of these are going to be subscription services. How do you see that business model working? How do you see the pricing there?

Travis Hoium: That's a really great question. If you look back on technology advancements like this, it's not often the best technology that wins, it's actually the best business model that wins. So this is something where if you're watching this space, watch what the business model is for these companies. I think that we have shown that the subscription model, at least for most features in vehicles, is not working very well. Was it BMW who tried to say you got to pay a subscription to get heated seats activated? People just went, absolutely not. I don't want to pay for stuff that's already in my car. This is not SaaS software, the way that we're used to on the Internet. I think that's a little bit tough. Even the adoption rate for FSD from Tesla is relatively low. I think it's about 5%. Well, now it's not fully autonomous. You're not able to turn your vehicle into a robo-taxi. If that was the case, maybe that adoption rate would go up a little bit, but they've had to bring their prices down too. I wouldn't be surprised if some of these things are either a premium that you would pay for vehicles. Let's take GM, for example. Maybe some of these Super Cruise features are included with a Cadillac, but they're not included with a similar vehicle that's a Chevy brand or something like that. Maybe there is a little bit of a subscription, but I don't think a $200 a month. That's another car payment. I don't think that's the value that people are looking for. What is more compelling to me is just replacing vehicle ownership and turning into transportation as a service, and I think that's where there's much more potential for disruption.

Ricky Mulvey: I'd love to own a self-driving car. I just don't know how much I want to pay for it. I hinted at this earlier. This is a technology that it is easy to be cynical about. It's a great local news story when a self-driving car gets in a crash. I was watching in Phoenix, Waymo, these cars slowly snaking down the oncoming lane, or there was someone who actually traveled to Phoenix because they were excited about trying out these self-driving cars, and the car goes into an alley, crashes into a pole. Waymo offers to send another car to get them to their destination. They send another car, and then that car gets stuck behind the crashed car. [laughs] How has the tech holistically changed over the past couple of years, though? We've seen this explosion in AI and large language models, and one of the things that Elon was excited at Tesla was changing the self-driving approach to a rules-based one to an LLM style. What that means is you train a vehicle to recognize, at a stop light, when a stop light turns green, you go, versus an LLM model, which is more inference-based, looking at driving holistically and rewarding, essentially behavior you like, and then penalizing behavior you don't. Long setup, how are you seeing self-driving tech change over the past few years?

Travis Hoium: All of these technology advancements are, I think, making it easier for all of these companies to advance what they're trying to do and improve the technology. You're having higher better cameras in vehicles, better LiDAR. Mobileye actually announced that they shut down some of the LiDAR research because regular LiDAR is getting so cheap. The information coming in to the system, the autonomous driving system is getting better. The compute inside the vehicle is getting better. The ability to read and understand what's going on is getting better. The question is going to be, what is the strategy that each of these companies is going to use, and what redundancies do we need to have in the system? You talked about Tesla, and I think the question fundamentally for Tesla, and this is really, what's their market cap today? Five hundred and $600 billion. The $500 billion question for them is, can you learn enough with artificial intelligence to only use vision for their entire system? Or do we need to have a level of redundancy in that?

I have an engineering background. You don't fly a plane that if one of the engines go down, you're just going to crash. You want to have failure modes. You want to be able to fail safely. I mentioned Mobileye earn a million miles earlier, that's not the system having a failure every million hours. It's actually, we have two systems. One could potentially fail every thousand hours, and the other could potentially fail every thousand hours. They have different failure modes, and so you basically overlap those two, and when you add them together, you have a much safer system. The other thing with artificial intelligence is there's what it can do today, and there's what it could maybe do in the future. I think that's a big question too.

The FSD is getting better, but it's not getting a bit better at a very rapid pace. Again, we're seeing the third party data, but it will go from 300 miles every disengagement to 315 miles. It's not 300-3,000-300,000. That's going to be the big change. Then the other question is regulators. We don't have really great answers for what these companies have to do from a regulatory perspective if they're actually going to deploy autonomous driving technology at scale. One of my questions is, if you're using AI, a lot of really smart people are willing to admit, we don't really know what goes in these AI models. If you have an accident and the regulator says, why did this accident happen? You say, I don't know. [laughs] That's not a very good answer.

Ricky Mulvey: There's the promise, which is, this will make driving tremendously safer, because computers, AI, don't get tired. You talked about the failure rates. There's the more uncomfortable question, which is, when one of these cars hits a pedestrian, who's at fault? [laughs] If one of these things kills someone, which has happened, as humans, we want reciprocity. We want to see someone go to jail and be punished for that. Let's get to a more comfortable conversation, which is how [laughs] these vehicles are going to deploy. Seems like Tesla wants to do both. Let's stay there for a second. They want to do robo-taxis. They want to have vehicles that people own. How are you seeing the Tesla business model working here?

Travis Hoium: We're going to hopefully learn more next month. But what they have indicated is that their ultimate goal is for you to buy a vehicle, pay $100 a month, $200 a month, deploy your vehicle when you're not using it into the Tesla robo-taxi network. That becomes an asset light business model for them. They're able to still sell you vehicles and sell this relatively high margin service and then ancillary revenue as well with doing the right chain part. To get there, they have also said that they need to build their own robo-taxi fleet. They will become Cruise or Waymo before they become the autonomous version of Uber. I think that's the way to think about it at least today. We don't know when these steps are going to happen or exactly how they're going to happen. But that's the vision for them. Then the other companies are doing the two things, but they're doing them individually.

Ricky Mulvey: We haven't talked much about Uber. Uber has had some significant updates with Waymo. What's happening with Uber and their self-driving efforts?

Travis Hoium: Uber wants to be the marketplace for self-driving companies. I think this is a really interesting place. If you go back in Uber's history, they were actually one of the earlier companies to try to go big into autonomous driving because Travis Kalanick, Uber's founder, really saw this as the future. He didn't want to be dependent on the supply that they're currently dependent on, which is human drivers. They made some mistakes in their technology. There's a whole long backstory there, but they basically got completely out of autonomous driving. Their current strategy is to basically white label all of these other autonomous driving companies.

Waymo and Cruise are the two that they have deals with currently. They're currently in operations with Waymo, and then they're expanding that. The interesting part about their recent expansion with Waymo is that Uber is actually going to be running the fleet. This is not going to be something where Waymo is building its own ride sharing service, and then when they're not busy, they'll offload some of that supply to Uber. It's actually all Uber supply. It's just a Waymo vehicle underneath. That's what's interesting, is can they get to the point where Waymo develops good enough technology to be fully autonomous, so does Cruise, so does Zoox, so does a dozen other companies, and then Uber can just say, we're going to be the touch point for customers and they're going to be interacting with us and we'll just figure out which vehicle to send them, and it takes that power out of the autonomous vehicle companies. That would be their strategy. But there again, does that work, or does somebody build scale in the autonomous technology and fleet first? That's the chicken and egg that we're in right now.

Ricky Mulvey: As we start to wrap up here, let's say someone's interested in this technology. There's not one easy winner to pick here, as we've talked about, but what stocks would you add to a self-driving autonomous vehicle basket?

Travis Hoium: The technology companies in the fleet owners, really, there's two main companies. That would be General Motors Cruise, and then Alphabet, which is Waymo. Now Waymo is a very small piece of Alphabet's business overall, but you get a great business in Search, and you get the optionality with Waymo, as well. The other company that I'm really intrigued by their potential is Mobileye. Mobileye is a company that's more of a horizontal business model, and they're going to be providing both hardware and software to automakers who are going to then implement that and then offer those services. They have deals with Volkswagen. In the future, I think they're going to start probably with Audi, you're going to be able to get fully autonomous features with Audi, be able to fall asleep in your car. That's not going to be Audi's technology, that's going to be Mobileye's technology under the hood. Again, another company that is beaten up by the market, a little bit like GM, but has a lot of potential to be a winner in this space.

Ricky Mulvey: Let's stay on Mobileye for a sec. What's going on with Mobileye lately?

Travis Hoium: They have a lot just going on in the background. They have messed up their inventory over the past couple of years with the ups and downs of the market, the changes in the Chinese market. Zeekr is one of their big partners. As that market has declined, the revenue has suffered, the overbuilt inventory in the process. A mess from a historical financial perspective, and then you add on the fact that their majority owner is still Intel, and everything that Intel is going through. The stock fell recently when it was rumored that, maybe Intel will have to dump this company. But I think you look out 3, 4, 5 years and what the potential business models are in autonomous driving. If you think that there's going to be an Uber fleet of autonomous vehicles, maybe some of those are going to be Volkswagen vehicles, some of those are going to be Zeekr vehicles. Mobileye is going to be a company that's going to be able to benefit from that. If you think you're going to be buying a vehicle that has Level 4 autonomy in it, and it's not a Tesla, more likely than not it's going to have some Mobileye technology in it. I think it's one of those companies, you may have to be really patient, but a decade from now, this is probably going to be one of the winners in the space.

Ricky Mulvey: Finally, a very unfair question. We talked about the robo-taxis. They're in warm climates. They like places like Phoenix. How far do you think we are from a former driver sitting in the back seat of a self-driving car going from, let's say, a bar to a house on a snowy night?

Travis Hoium: As somebody who lives in Minnesota, I don't think that we are probably one of the first places these are going to scale. We're probably a ways out from even testing that. They're going to scale these technologies in, like you said, nicer climates and make sure that they can figure all this out. But the nice thing with some of the technologies that these are building, in particular radar, is you can actually see better on a snowy night using radar than you could with vision alone. It is very possible that we will see those relatively soon in the next decade, and it will be a relatively easy and safe way to drive as opposed to as hard as it is to drive on a snowy night just as a regular driver. I'm optimistic, but this is absolutely not a location where these are going to scale first because there's plenty of other places in the world where they can build out their autonomous fleets.

Ricky Mulvey: Travis Hoium, thank you for your time and insight on this. Appreciate the work you've done looking into autonomous vehicles.

Travis Hoium: Thanks for having me.

Ricky Mulvey: As always, people on the program may have interests in the stocks they talk about, and the Motley Fool may have formal recommendations for or against, so don't buy or sell anything based solely on what you hear. I'm Ricky Mulvey. Thanks for listening. We'll be back tomorrow.

John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool's board of directors. Ricky Mulvey has no position in any of the stocks mentioned. Travis Hoium has positions in General Motors and Mobileye Global. The Motley Fool has positions in and recommends Amazon, Apple, Baidu, Tesla, Uber Technologies, and Volkswagen. The Motley Fool recommends Bayerische Motoren Werke Aktiengesellschaft, General Motors, Mobileye Global, and Volkswagen Ag and recommends the following options: long January 2025 $25 calls on General Motors. The Motley Fool has a disclosure policy.

The State of Autonomous Vehicle Development was originally published by The Motley Fool