Name: Your Autonomous Ride Has Arrived
Uploaded: 2025-12-09
Duration: 15 min 21 s

The dream of autonomy has become reality. With robotaxis now navigating real traffic, Zoox is showing how AI can master one of the world’s most complex systems: the city street.From breakthroughs in edge intelligence to safety engineering, the Amazon-backed company hopes to redefine how and where the world moves. Which problems are still to solve? And how will the company compete with more entrenched rivals?

Your Autonomous Ride Has Arrived

Conferences

Good morning. Thanks for being with us, Jesse. We don't have a ton of time, so I want to tell you from the jump. I'm going to try to get at least one question from the audience, so please start thinking of that now. Um, let's do a quick show of hands. How, how many folks in the room have driven in a or rode in a Waymo? Anyone? A lot of people? OK. Has anyone been fortunate to ride in a zoos yet? One? Anyone else? No? OK. So let's, let's just start there, um. How, how should we think about that disparity right now and how I understand it's the early days, but how you close that gap. Yeah, absolutely. So you know Waymo has been at this for quite a bit longer than Zoos, and we also took a little bit harder path of building our own vehicles from the ground up. Sounds a little bit harder, a little bit harder, not that, not that what Waymo did was easy. It's also incredibly hard, and they're doing an awesome job, um. But we did take that harder and longer path, and we actually are pretty proud because we think we've been able to close the gap quite a bit, but of course there's still a gap. So we just started taking the public off of our waitlist in San Francisco over the last couple of weeks, and we still have a very big waitlist. We are open to the public in Las Vegas and we have been since September with the public launch over there. It's still a preview service and it's free, but it's going really well. We've had over 100,000. and we're actually excited. Just just last Thursday we crossed 1 million miles in the Robotaxi. That's the actual robotaxi, you know, no driver on public roads just this year calendar year 2025. So it's coming, it's going super well, but obviously we have more work to do to expand our geo fence, add a bunch of vehicles, and then take everybody off the waitlist in San Francisco. When, when are you able to start charging for these rides and, and also when, you know, talk to me about the expansion. Both in other markets and sort of what you can do within your current markets going forward. Yeah, so it's a little complicated, so we need approval on both the federal level and the state levels. And so for example in Nevada we already have approval from the state, but in California it's a little bit more complicated because you have to work with the CPUC and there's more steps there so. Moment we are also waiting on the federal exemption, the Part 555 exemption from NHTSA, which we're working with them on as we speak. We expect to get that hopefully early next year and what that would give you is the ability to start charging. That gives us the ability to start charging from a federal perspective, and then we can quickly start charging in, for example, Las Vegas. Uh, and then in California there's a couple of additional steps with the CPUC. We're working with them as well, so charging in San Francisco would be coming later in 2026 in California as opposed to earlier in 2026 in Nevada. I was that's a good, that's a good point, uh, time to point out that you are a subsidiary of Amazon. Amazon typically likes making money, so I was going, I was going, you know, safe to say, and um. I was going to ask how confident you feel in sort of the long term runway you have, um, but it sounds like revenue coming in next year will be a start. Absolutely, and I want to be clear, I mean, the amount of revenue will make next year won't really pay for Zoos. I mean Zoos is pretty expensive and you know we're a few 1000 people and building a lot of robo taxis, but it's definitely a start and a proof point. And then you know over the next few years it'll start to be a really interesting business because the revenue you can generate from a robotaxi is quite a bit more than the expense to run a Robotaxi, including amortization of the bomb cost as well. So we're really excited about that, but it's still several more years before we can operate at a scale where you know that's financially interesting to Amazon. Um, let's talk about the technology because part of the reason, um, there is that gap we talked about for now with, with a company like Waymo is you've built your own vehicles, um. Can you give the the sort of queerest um Uh explanation on why you did that and why you're confident that's still the best path forward. You call them purpose built vehicles, right, versus retrofitting an existing car. Yeah, absolutely. So cars have been designed over the last 100 years for humans to drive them. And so all of the choices, their shape, their architecture, what components they have and don't have, they were all designed for human drivers, and you can modify them and add a little bit more redundancy and you know maybe remove the steering wheel, maybe not, but you're still ending up with an architecture that's far from optimal for a Robotaxi. So we're able to create a much nicer customer experience inside the vehicle with carriage seating and features you don't find in a car, much more room inside and a much more compact. Externally, which is great. We're also able to add a much bigger battery. This is really good environmentally and economically because you don't have to be charging the battery multiple times a day like you like you do on a retrofitted car. You're also able to have much better safety and redundancy. So not only can you be safer, but you can also ensure that if a hardware component fails, you can still take the customer to their destination, and these are all things you can't do on a retrofitted car. So in the future, and I'm going to come for a question in a second, anyone has one in the future. In your mind, someone would choose a zoos for a ride over Awamo or um we can talk about that company, Tesla, um. For what reason or reasons? Oh, it's just a much better experience, right? So, so your, your time in the vehicle is dramatically nicer. You have much more space. If you're with friends, it's dramatically more social because you're facing each other you're facing each other and, and you know, trying to put multiple people. In a regular car is not not very nice. And even if you are by yourself, I mean, you know, you have that, that front seat in front of you, it's just, it's just again it was very much designed for, for human drivers. So when you do not have that constraint, we can build something much, much nicer, and that is exactly what we've done. Um, do we have any questions? Otherwise I can Yeah, we have one here um just wait for a mic please and tell us who you are and. Where you're uh Where you work up. Oh yeah, right here they're coming. Hi, my name is Ashley. I'm with Microsoft. Um, my question for you is, since you're a subsidiary of Amazon, why focus on RoboTaxi and not just the last mile delivery component? Like, it seems a pretty crowded field, so why not just go into the space where we know Amazon has been a dominant player. Well, we don't really think that Robotaxis is a crowded field. There's really only one company out there right now in the US that's doing Robotaxis, and so that doesn't seem very crowded to us, and there are actually no companies doing a purpose-built roboTaxi anywhere in the world yet. So we think there's a really exciting market for us from Amazon's perspective. It's not that they wouldn't like to automate some aspects of package delivery. They're obviously working on that. But the economic opportunity and the opportunity for customers to create this whole new category of transportation is actually much more exciting and even more financially compelling than simply taking something they do today and saving a bit of money so um. We'd love to save money, obviously, but that's not as exciting as creating this whole new market, and the addressable market for moving people around cities is just profoundly huge, and the union economics of Robotaxis seem like they're going to be very compelling. So that's why Amazon and Zoos are really excited about that. Is there any um Delivery component on the roadmap, not explicitly. We do talk to Amazon last mile as they're kind of, you know, working on their autonomy plans, but we're quite separate, and we've we've been instructed all the way from the very top to stay laser focused on moving people around. It's harder to move people around than packages. In terms of what you have to do with the vehicle, but on the other hand, there are also challenges with automating packages, right? You have to get the packages in the vehicle and get them out of the vehicle, so it's actually not as straightforward as it might seem to do automated package delivery. Um, as an innovator, I'm curious about, you mentioned earlier, no one in the world is doing purpose-built, um, autonomous vehicles right now. Does that, that, that looks like a huge opportunity, but does that ever make your confidence waiver that it's the right approach? Not really, especially lately. I mean, I think it was. Very fair to be skeptical, you know, a decade ago when we were starting Zoos, but I think at this point we have now a lot of data from from the field from actual customers. I mentioned we've had over 100,000 people riding Zuxes, and these are not people we know. These are just people, and the reaction has been actually much better than we expected, which is great to see because we obviously thought people would love it, otherwise we wouldn't have built this thing, but seeing how much. People prefer being in the zoos over any other form of transportation that can get them around cities is super encouraging. And so our view is that people aren't doing this not because it's not a good idea, but because it's just really hard. It takes a lot of time. It's very cross functional. It's expensive, but I do think over time this is, this is going to be a much more popular way of doing urban transportation. I think I saw another hand. We have one up here if we can get a mic please on this side. Thanks for your patience. Just tell us who you are and uh who you're with. Hi, this is Ursh from Adeline. Um, I have like a two parter. How do you model the economic impact for zoos of zoos and What would you say Zooks is doing in 10 years from now? Is it still just Robotaxis? If not, how does the portfolio of products and services expand from kind of the foundational stuff that you've built today? Yeah, so the economics are super interesting. It's actually one of the reasons why we started Zoos in the first place is, uh, you know, cars are used 4% of the time. And that means that 96% of the time they're just sitting around taking up space and depreciating. Not only is that economically wasteful, it's environmentally wasteful as well because there's a lot of stuff that goes into a car. It's not just driving cars that's bad for the environment. It's making them as well. Making a robotaxi also bad for the environment. But if you have to make way fewer of them to serve the same number of people, that's a win. And so when we look at the economics, it's a, it's a really exciting win for customers because you know you only have to pay for the vehicle when you're using it. You don't have to pay for parking, insurance, depreciation, all that, and it's great for us because one vehicle can charge for itself all day and all night long. So those unit economics, obviously you don't have to pay a driver, are really compelling. Now it takes a lot. Of capital and a lot of work to get to that point, but we're very close to it and that's very exciting as far as 10 years from now, you know, look, I think you could hopefully assume we have multiple types of vehicles, different form factors, and I would say 10 years from now if things are going well, you know, we should also be in many countries, not just the United States, and we'll probably be in some other areas, you know, maybe we'll be working. On packages or other things like that, it's not that we're not interested in expanding. We're just trying to stay super focused on, you know, robotaxis for people because that's already pretty hard, and we believe that, you know, we can create a multi $100 billion company just moving people around if things go well. We have a lot of work to do to get to that point, but we do believe that's possible. Um, I do have one, question about Tesla. Um, you, you made a little news last year, maybe accidentally, um, you were asked about full self-driving in Tesla, and, um, and you said something along the lines and I'm paraphrasing, so please correct me of, well, it, it doesn't fully work and in that you meant, you know, you still need a driver in the car. Um, Elon has. Made some sort of promise which he's known to do or said that they will be removing drivers test their safety drivers from some of their rides that they're currently offering. I'm curious if your opinion has changed at all um based on. Um, what they're saying they're going to be doing soon in taking a safety driver out of the vehicle. Yeah, I mean they've been saying similar things for many, many years. I think their FSD is by far the most advanced. and safest driver assistance system in the world. I mean there's nothing that's even close, and it's a remarkable technical and product achievement. It's also pretty clear from the recent data that if you have an attentive human driver supervising it in many scenarios it's safer than, you know, just a regular human by themselves. So that's all wonderful. My point is that the. Difference between a supervised system and an unsupervised system, which means that it can be, you know, safer than a human with nobody watching. That's a, that's a huge difference. And so our view is that, you know, having the extra hardware, you know, active sensing and and really designing a system from a systems engineering perspective very rigorously is very important if you want to remove the driver, and we continue to believe that. Would you take a ride in a Tesla with no driver in it? the taxi. That's an interesting question, um. Maybe for a few miles, a few miles, OK, um, just to see what it's like. One final one for me and we'll try to squeeze in one more, uh, what. What AI advancement in the last year or couple of years had made has made the biggest sort of difference in Zu's technology? Yeah, I mean. It's incredible. It's happening in the AI field. When when we started Zoos, we did not anticipate, you know, AI broadly being this capable and you know, sort of so close to something AGI-like, and we don't think you necessarily need to have AGI to drive safely in cities. That's clearly the case. But since, since there is so much progress in that direction, we've been able to really automate a lot of our pipelines. One of the most important. And things we do is how quickly can we improve our software and validate that it's safe? We've been able to use these modern techniques to really speed up and improve the quality and precision of these offline pipelines that help us validate the software. We're also online using a lot of these modern transformer-based techniques, for example, to take in all of the sensor data and really just understand what's going on in the world rather than having to have separate pipelines. For cameras or radars or liars, it's also really improving the way we do prediction and motion planning. So we're really moving to kind of combine prediction and motion planning into some of these newer models still with some of the guardrails and costs associated with some of the traditional techniques we're really kind of doing a hybrid approach to really guarantee safety while leveraging a lot of the advantages of these newer techniques. So it's a very exciting time. And you finished right on time, Jesse. Wow, let's get a round of applause for Jesse Levinson. Thank you for having us here. Thanks so much.