Building the World's Most Trusted Driver

Steph: Hello, everyone. Welcome back to the A16Z podcast. This is Steph. Now, one of my favorite podcasts we've recorded since I joined the team was just about this time last year. That episode was on autonomous vehicles, but it was actually also in an autonomous vehicle. That was my first ride in a self-driving car. And over the last year, I've seen so many others have their firsts as Waymo has expanded to the public in Phoenix and San Francisco, while also placing its roots in Austin and LA. In 2015, Waymo tested its first fully driverless ride on public roads. It then opened to the public in Phoenix in 2020, but it wasn't until 2022 that autonomous drives were offered in San Francisco. And by the end of 2023, it clocked in over 7 million driverless miles. Slowly, then all at once. So with this space moving so quickly, we wanted to give you an update on where this industry is today. Passing the baton to properly introduce this episode, here is our very own AI Revolution host and A16Z general partner, Sarah Wang.

[Music transition]

Sarah Wang: Hey guys, I'm Sarah Wang, General Partner on the A16Z Growth Team. Welcome back to our AI Revolution Series. In this series, we talk to the Gen AI builders who are transforming our world to understand one, where we are, two, where we're going, and three, the big open questions in the field. Our guest this episode is Dmitry Dolgov, the co-CEO of Waymo. Dmitry has led Waymo to solve some of the biggest challenges in bringing AI to the real world. And after tens of millions of miles of testing, Waymo's vehicles have shown themselves to be safer and more reliable than human drivers, myself included. Dmitry has a unique perspective, given that his work has spanned multiple AI ML development cycles across decades. He was an early pioneer in self-driving cars, working with Toyota and Stanford on DARPA's Grand Challenge before joining Google's self-driving car project, which then evolved into Waymo. In this conversation from a closed-door event with A16Z general partner David George, Dmitry talks about the potential of embodied AI, the value of simulations and building training data, and his approach to leading a company focused on solving some of the world's hardest problems. Without further ado, here's Dmitry in conversation with David.

[Music transition]

David George: Maybe to start, take us back to Stanford, if you will. And that was when you first started working on the DARPA project. And maybe give us a little bit of your history of how you ended up from there to here.

Dmitry Dolgov: My introduction to autonomous vehicles was when I was doing a postdoc at Stanford that you just mentioned, David. This was during, I got pretty lucky with the timing of it. This was when the DARPA Grand Challenges were happening. DARPA is the Defense Advanced Research Project Agency that started these competitions with the goal of boosting this field of autonomous vehicles. And the one that I got involved in was in 2007. That was called the DARPA Urban Challenge. So the setup there was, it's kind of like a toy version of what we've been working on since then. It was kind of supposed to mimic the driving in urban environments. So they kind of created a fake city on an abandoned air base, and they populated it with a bunch of autonomous vehicles, a bunch of human drivers, and they had them do various tasks. So that was kind of my introduction to this whole field. And it was a bit of a, I think, you know, these challenges are often, by people in the industry, considered kind of a foundational, pivotal moment for this whole field. And it was definitely that for me. It was like a, you know, light bulb, light switch moment that really got me hooked.

David George: What was like the hardware and software that you guys had at that point? This is 2007.

Dmitry Dolgov: Yeah, yeah, I know it's, you know, at a very high level, not unlike what we talk about today. You know, a car that, you know, has some instrumentation so you can, you know, tell it what to do and you get some, you know, feedback back. Then you have kind of what's called a post system, a bunch of inertial measurement system, accelerometers, gyroscopes that kind of tell you, and GPS tells you how you're moving through space. And it has sensors, radars, lighters, and cameras, the same stuff we still use today. And then there's a computer that gets the sensor data in and then tells the car what to do, and a bunch of software. And software had perception components and decision-making planning components and some AI. But of course, everything that we had, like each one of those things over that, how long has it been? Almost 18 years more than that. It's changed drastically, right? So when we talk about AI today versus AI we had, you know, back in 2007, 2009, you know, nothing in common. And similarly, everything else has changed. You know, the sensors are not the same. Computers are not the same.

David George: So then, okay, so take us, so at that point, that was the pivotal, that was like the light bulb moment. And then at that point, you said, okay, I'm at Stanford. I want to make this my career, right? Is that, and then it was Toyota, and then where did it go from there?

Dmitry Dolgov: I don't know if I thought about it in those terms. I was like, this is it. This is pretty cool to work on. This is the future. I want to make it happen. I want to be building this thing. Career, okay, you know, they can wait. But it was, that was the next step. That was the next big step is a number of us from the DARPA challenge competitions started the Google self-driving project. It was about a dozen of us. Then in 2009, came together at Google with support and excitement from Larry and Sergey to see if we can take it to the next step. And we worked on it for a few years, and that project then became Waymo in 2016, and we've been on this path since then.

David George: Okay, so we have this new big breakthrough in generative AI. Some would say it's new. Some would say it's 70 years in the making. How do you think about layering advances that have come from generative AI to what many would describe as more traditional AI or machine learning techniques that were kind of the building blocks for self-driving technology up to that point?

Dmitry Dolgov: Great question. So maybe I can, you know, generative AI is kind of a broad term. So maybe take a little bit of a step back and talk about kind of the role that AI plays in autonomous vehicles and kind of how we saw the various breakthroughs in AI map to the space of our task. As I mentioned, AI has been part of self-driving autonomous vehicles from the earliest days. Back when we started, it was very different kind of AI, ML, classical decision trees, classical computer visions with kind of hand-engineered features, kernels and so forth. And then one of the first really important breakthroughs first really important breakthroughs that happened in AI and computer vision, but really was important for our task, was the advancement in convolutional neural networks right around 2012. Many of you are probably familiar with AlexNet and the ImageNet competition. This is where AlexNet blew away out of the water all other approaches. So that obviously had very strong implications for our domain, like how you do computer vision, and not just on cameras, right? How you run, you know, you can use ConvNets to interpret what's around you and do kind of object detection and classification from camera data, from LiDAR data, from your imaging radars. So that was kind of a big boost around that, you know, 2012, 2013 timeframe. And then we played with those approaches and, you know, tried to extend the use of ConvNets to other domains, you know, just be able to be on perception with, you know, some interesting but limited success. Then another big, very important breakthrough happened around 2017 when Transformers came around. It had a really huge impact on language, language understanding, language models, machine translation, so forth. And for us, it was a really important breakthrough that really allowed us to take ML and AI to new areas well beyond perception. And so if you think about transformers and the impact that they had on language, kind of intuition is that they're good at understanding and predicting and generating sequences of words, right? And in our case, we think about, in our domain, about the tasks of understanding and predicting what people will do, like other actors in the scene, or the task of decision-making and planning your own trajectories, or in simulation, kind of generating, generative AI, our version of General AI, kind of generating behaviors of how the world will evolve. That kind of these behavioral like these sequences are not unlike sentences right You kind of operating the state of objects And then there kind of local continuity but then the global context of the scene really matters So this is where we saw some really exciting breakthroughs in behavior prediction and decision-making and simulation. And then, you know, since then, we've been on this trend of models getting bigger. People started building foundational models for multi-tasks. And most recently, all the guys, the last couple of years, all the breakthroughs in large language models, modern state, modern-day generative AI, visual language models where you kind of align image understanding and language understanding. And there's been most recently one thing I'm pretty excited about is kind of the intersection or combination of the two. So that's what we've been very focused on at Waymo most recently is taking kind of the AI backbone and all of the AI, the Waymo AI that is over the years we've built up that is really proficient at this task of autonomous driving and combining it with kind of the general world knowledge and understanding of these, you know, VLMs.

David George: One of the things that you just mentioned is the role of simulation and how that has been. You guys have had major breakthroughs in the use of simulation. And this idea in the recent breakthroughs in generative AI around synthetic data and its usefulness is somewhat in question. I would say in your field, this idea of synthetic data and simulation is extremely useful, and you've proven that. So maybe you could just talk about the simulation technology you guys have built, how it's allowed you to scale, build that real-world understanding and maybe how it's changed in the last few years?

Dmitry Dolgov: Yeah, definitely. It is super important in our field. Largely, if you think about this question of evaluating the driver, is it good enough? How do you answer that? There's a lot of metrics and a lot of data sets you have to build up. And then how do you evaluate the latest version of your system? You can't just throw it on the physical world and then see what happens. You have to do it in simulation. But of course, the new system behaves differently from what might have happened in the world otherwise. So you have to have a realistic closed-loop simulation to give you confidence in that. So that is one of the most important needs for the simulation. You've also mentioned synthetic data. I guess that's another area where simulation allows you to have very high leverage. You just got to explore the long tail events. Maybe there's something interesting that you have seen in the physical world, but you want to modify that scenario. You want to turn one event into thousands or tens of thousands of variations of that scenario. How do you do that? This is where the simulation comes in. And then lastly, if you sometimes want to evaluate and train on things that you've never seen, even on a very vast experience. So this is where purely synthetic simulations come in that are not based on anything that you have seen in the physical world. So in terms of technologies that go into play, it's a lot. And that is like a huge generative AI problem. But what's really important is that that simulator is realistic. It has to be realistic in terms of your sensor or perception realism. It has to be realistic in terms of the behaviors that you see from other dynamic actors. If other actors are not behaving in a realistic way, like if pedestrians are not walking the way they do in the real world, You need to be able to quantify the scenarios that you create in simulation to the realism and the rate of occurrence in the physical world, right? It's very crazy to sample something, very easy to sample something totally crazy in a simulator, but then what do you do with that? So I think that that brings me to the third point of realism is that it has to be kind of realistic and quantifiable at the macro level, at the statistical level. So there's any can imagine there's a lot of work that goes into building a simulator that is large-scale and has You know that that level of realism across those categories and they've kind of intuitively think about it You know to build a good driver you need to have a very good simulator But to have a good simulator you actually have to build models of like realistic pedestrians and cyclists and drivers, right? So it's good, you know, you kind of do that iteratively. Yeah, of course And then by having this simulation software that is very good at mimicking real world and very usable in the sense that you can create variables in the scenes, you can actually give the driver multiples of the amount of experience that they have on the road. That's exactly right. In real miles. Is that right? That's exactly right. We've driven tens of millions of miles in the physical world. At this point, we've driven more than 15 million miles in full autonomy, what we call a rider-only mode. But we've driven tens of billions of miles in simulation, So you get orders of magnitude of an amplifier.

David George: Speaking of multiples of miles driven, one of the hotly debated topics in the AI world today is this concept of scaling laws. So how do you think about scaling laws as it relates to autonomous driving? Is it miles driven? Is it certain experience had? Is it compute? What are the ways that you think about that?

Dmitry Dolgov: So model size matters. So we're seeing scaling laws applied. A lot of typical old-school models are severely under-trained. And so if you have a bigger model, you have data that actually does help you. You just have more capacity, they generalize better. So we are seeing scaling laws apply there. Data, of course, hugely matters. But it's not just counting the miles or hours. It has to be the right kind of data that teaches the models or trains the models to be good at the rare cases that you care about. And then there is a bit of a wrinkle because then you can build those very large models. But in our space, it has to run on board the car, right? So you are somewhat completely constrained. So you have to distill it into your onboard system. but we do see a trend which is a common trend and we see that play out in our space where you're much better off training a huge model and then distilling it into a smaller model than just training small models.

David George: I'm going to shift gears a little bit and I'm going to do a sort of simplifying statement which is probably going to drive you crazy but the DARPA school of thought is there's sort of a rules-based approach a more traditional kind of AI-based approach with a massive amount of volume and you document edge cases and then the model then learns how to react to those. The more recent approaches from some other large players and startups would say, hey, we just have AI from the start, make all the decisions end to end. You don't need to have sort of all that pattern recognition and learning, like the end to end driving that is kind of a tagline out there. What is your interpretation of that approach? And what elements of that approach have you taken and applied inside of Waymo?

Dmitry Dolgov: I think it's kind of sometimes it's the way people talk about it as kind of this weird dichotomy. Is it this or that? Yeah, of course. But it's not. It's that and then some, right? So it is big models. It is end-to-end models. It is generative AI and combining these models with VLMs, right? But the problem is it's not enough. Right. So we all know the limitations of those models, right? And that's what we've seen through the years, a lot of these breakthroughs in AI, right? ConvNest, Transformers, big end-to-end foundation models, they're huge boosts to us. And what we've been doing at Waymo through the history of our project is constantly applying and pushing forward these state-of-the-art techniques ourselves in some cases, but then applying them to our domain. And what we've been learning is that they really give you a huge boost, but they're just not enough. So the theme has always been that you can take your latest and greatest technology of the day and it's fairly easy to get started. Like the curves always look like that. And the curves have been shaping, but the really hard problems in that remaining 0.0001%. And there it's not enough. So then you have to do stuff on top of that. So yes, you can take, nowadays, you can take an end-to-end model, go from sensor to trajectories or actuation. Typically, you don't build them in one stage, you build them in stages, but you can do backprop through the whole thing. So the concept is very, very valid. You can combine it with a VLM, and then you add closed loop simulations, some sort, and you're off to the races. You can have a great demo almost out of the box. You can have an ADESP or a driver assist system, but that's not enough to go all the way to full autonomy. So that's where really a lot of the hard work happens. So I guess the question is not is it this or that, it's this, and then what else do you need to take it all the way to have the confidence so that you can actually remove the driver and go for full autonomy And that a ton of work That a ton of work through the entire kind of life cycle of these models and the entire system right So it starts with training. Like, how do you train? How do you architect these models? How do you, you know, evaluate them? Then, you know, if you put in a bigger system, the models themselves are not enough. So you have to do things around them. You have to, you know, they have modern generative AI is great, but there are some issues with, you know, hallucinations. hallucinations, explainability. Exactly. They have some weaknesses in goal-oriented planning and policy-making and understanding this operating in this 3D spatial world. So you have to add something on top of that. We talked a little bit about the simulator. That's a really hard problem in and of itself. And then once you have something, once you deploy it and you learn how do you feed that back. So I guess this is where all of the really, really hard work happens. So it's not like end-to-end versus something else. It is end-to-end. and big foundation models, and then the hard work. And then all the hard work.

David George: Yeah, it totally makes sense. That is a great segue into all of the progress that you guys have made, right? Writing in the Waymo, for those who have done it, is an extraordinary experience. It's not to say that you have solved all of these complex tasks, but you've solved a lot of them. What are some of the biggest AI or data problems that you still feel like you're facing today?

Dmitry Dolgov: The short answer is going to be taking it to the next order of magnitude of scale, multiple orders of magnitude of scale. And with that come additional improvements that we need to make it a great service. But just to level set in terms of where we are today, we are driving in all kinds of conditions that we're driving 24-7 in San Francisco, in Phoenix, a little bit. Those are the most mature markets, but also in L.A. and in Austin. And all of the complexity that you see, go drive around the city, right? All kinds of weather conditions, whether it's fog or storms or dust storms or rainstorms down here, all of those are conditions that we do operate in. So then I think about what makes it a great customer experience. What does it take if you grow by next orders of magnitude? There's a lot of improvements that we want to make so that it becomes a better service for you to get from point A to point B. We ask for feedback from our writers. A lot of feedback we get has to do with the quality of your pickup and drop-off locations. So we're learning from users. We want to make it a magical, seamless, delightful experience from the time you kind of start the app on your phone to when you get a decision. So that's a lot of the work that we're doing right now.

David George: Yeah. Pick up and drop off, for what it's worth, is an extraordinarily hard problem, right? Like, do you kind of block a little bit of a driveway if you're in an urban location and then have a sensor that says, oh, actually, I just saw somebody opening a garage door. I need to get out of the way. You know, how far down the street is acceptable to go pull? Or if you're in a parking lot, where in the parking lot do you go? So this is an extraordinarily hard problem, but to your point, it's huge for user experience.

Dmitry Dolgov: That's exactly right. I think that's a good example of just one thing, one of the many things that we have to build in order for this to be an awesome product, not just a technology demonstrator. And I think you just hit exactly on a few things that make something that at the face of it might seem fairly straightforward. right? Okay, you know, I know there's a place on the map and need to pull over, like how hard can it be, right? But really, if it's a complicated, you know, a dense urban environment, there's a lot of these factors, right? Is there like, you know, another vehicle that you're going to be blocking? Is there a garage door that's opening, right? Like, you know, what, what is the most convenient place for the user to pick up? What is, you know, so it really gets into this, you know, the depth and the subtlety of understanding the, you know, the semantics and the dynamic nature of this driving task and, you know, doing things that are safe, comfortable, and predictable, and it leads to a nice, seamless, pleasant, delightful customer experience.

David George: Of course, okay, so you've mentioned this stat, but 15 million miles, I know the number's probably a little bit bigger than that, but you just released it Tuesday, yeah. It's growing by the day. 15 million autonomous miles driven. That's incredible. Even more impressive, and you didn't share this stat yet, it results in 3.5 times fewer accidents than human drivers. Is that right? And I think 3.5x is the reduction in injury. And that's about 2x reduction in the police reportable kind of lower severity incidents. This sort of comes to a question of both kind of regulatory and kind of business or ethical judgment. What is the right level that you want to get to? Obviously, you want to constantly get better. But is there a level at which you say, okay, we're good enough, and that's acceptable to regulators.

Dmitry Dolgov: Yeah, so there's no, you know, simple answer. Super simple short answer. Right. I think it starts with that. It starts with those statistics that you just mentioned. Yeah. Like, at the end of the day, what you care about is that roads are safer. So when you look at those numbers, yeah, when we operate today, and we have strong empirical evidence that our cars are, in those areas, safer than human drivers. So on balance, that means a reduction in, you know, collisions and harm. Then, actually on top of the numbers, we've probably been publishing this as you're quoting the latest numbers that we've shared, consistently sharing numbers as our service scales up and grows. If you can also bring in an additional lens of how much did you contribute to a collision? And we actually published, I think it was based on about 4 million miles, 3.8 million miles we published a joint study with Swiss RE, which is I think the largest global reinsurer in the world. And the way they look at it is, you know, who contributed to an event. And there we saw like the same theme, but the numbers were very strong. That new field was a 76% reduction in property damage collisions. And it was a 100% reduction in claims around bodily injury. So if So if you kind of bring in that lens, I think the story becomes even more compelling. That is extremely compelling. Right, but there are some collisions where we'd be, and that's the bulk of the events that we see, we'd be stopped at a red light, and then somebody just plows into you. Sure. But then, I think we do know it's a new technology, it's a new product, so it is held to a higher standard. So when we think about our safety and our redness, framing methodology, we don't stop at just the race. we've built over the years. That's one of the huge areas of investment and experience over the years. Like how, you know, what else do you need? So we have done, and we've done a number of the other different things. We've done, we've published some of our methodologies. We've shared our readiness framework. You know, we do other things like we actually, not just statistically, but on specific events, we build models of an attentive, very good human driver. Like not distracted human, it's a good question whether such a driver exists, right? That's kind of what we compare our driver to, right? And it's a model. In a particular scenario, we evaluate ourselves versus that model of a human driver, and we hold ourselves to the bar of doing well compared to that very high standard. And then you pursue other validation methodology. So that's my answer, is that it's the aggregate of all of those methodologies that we look at to decide that, yes, the system is ready enough to be deployed in scale.

David George: I'd love for you to talk about what you think maybe today and in the future about market structure, competition, and what kind of role you envision Waymo playing.

Dmitry Dolgov: The way we think about Waymo and our company is that we are building a generalizable driver. That's the core. And that's the core of the mission of making transportation safe and accessible. And we're talking about right-hailing today. That's our main, most mature primary application. But we envision a future where the Waymo driver will be deployed in other commercial applications. There's deliveries, there's trucking, there's personally owned vehicles. So in all of those, our guiding principle would be to think about the go-to-market strategy in a way that accelerates access to this technology and gets deployed as broadly, while of course doing it gradually and deliberately and safely, but as quickly and broadly as possible. So with that as our guiding principle, we're going to explore different commercial structures, different partnership structures. For example, in Phoenix today, we have a partnership with Uber and ride hailing, both in ride hailing and in Uber Eats. So in Phoenix, we have our own app. You can download the Waymo app and take a ride and our vehicle will show up and take you where we want to go. That's one way to experience our product. Another one is through the Uber app. We have a partnership where you can get through Uber app matched with our product, the Waymo driver, the Waymo vehicle and it's the same experience. But this is another way for us to accelerate and give more people to experience full autonomy and it gives us a chance to kind of think about the different go strategies right One is us having more of our own app The other one is more of a driver-as-a-service for somebody else's network. So we'll, you know, still early days, but we'll iterate and hold in service of that main principle.

David George: That's amazing. Yeah, that's going to be exciting. um maybe on back to the vehicle what about the hardware stack that you use you and i have talked a bunch about you know you said like hey going all the way back to darpa um you know it's kind of the same stuff right it's you know it's sensor they've advanced quite considerably but um you know you still use you know radars and lidar um do you think that remains the future path for autonomous driving, LiDAR specifically?

Dmitry Dolgov: Yeah, no, I mean, the sensors are physically different, right? They have each one brings their own benefits. Each one brings their own benefits, right? You know, cameras obviously give you color and they give you very high resolution. LiDARs kind of give you a direct 3D measurement of your environment, and they're an active sensor, right? So it kind of brings their own energy, pitch dark when there's no, you know, external light source, you know, you still get, you know, to see just as well as they do during the day, you know, and better in some cases. And then, you know, radar is, you know, very good at, like, punching through just, you know, physics, different wavelengths, right? So if you build an imaging radar, which we do ourselves, you know, it allows you to, you know, give you an additional redundancy layer and it has benefits also an active sensor it can directly measure you know its Doppler velocity of other objects and it can you know degrades differently and more gracefully in some weather conditions I can a very dense fog you know are very concerned so you know they all have their benefits so if you you know our approach has been to you know use all of them right and you know that's how you have redundancy and that's how you get an extra boost and and capability of the system. And we are on, today deployed in fifth and working to deploy the sixth generation of our sensors. And over those generations, we've improved reliability, we've improved capability and performance, and we've brought down the cost very significantly. So yeah, I think the trend for us that we're using all three modalities just makes a lot of sense. Again, you might make different trade-offs if you are building a driver assist system versus a fully autonomous vehicle where that last 0.001% really, really matters.

David George: One of the observations that we have from the very early days of this wave of LLMs is that there has been sort of already a massive race of cost reduction, and many would argue that it's sort of a process of commoditization already, even though it's very early days. I would say the observation from autonomous driving over many, many years now is kind of the opposite thing. There's been a thinning of the field. You know, it's proven to be much, much harder than expected. Can you just talk about maybe why that's the case?

Dmitry Dolgov: You know, they always have this property that it's very easy to get started, but it's very insanely difficult to get it, you know, all the way, you know, to full autonomy so that you can remove the driver. And there's maybe a few factors that contribute to that. One is compared to the LLMs and AI in the digital world. Video, yeah. You have to operate in the physical world. The physical world is messy. It is noisy. And it can be quite humbling. There's all kinds of uncertainty and noise that can pull you out of distribution, if you will. Right, sure. So that's one thing that makes this very difficult. And secondly, it's safety, right? Sure. These AI systems in some domain, this is creativity, and it's great. In our domain, the cost of mistakes, our lack of accuracy has very serious consequences, right? So that's just the bar very, very high. And then maybe the last thing is that it is, you have to operate in real time. You're putting these systems on fast moving vehicles and you have to, milliseconds matter, right? You have to make the decisions very quickly. So I think it's the combination of those factors that really together lead to the trend that you've been seeing is that it's an and, right? You have to be excellent in this and this and this and then, right? It's all of the above. The bar is very, very high for every component of the system and how you put them together. But there's big advances and they boost you and they propel the system forward, but there are no silver bullets, right? And there are no shortcuts if you're talking about full autonomy. And because of that lack of tolerance for errors, you have a very high bar for safety. You have a very high burden from regulators. You know, it's very costly to go through all those processes. And so it makes sense. And I'm very grateful that you guys have seen it through despite all the humbling experiences that you had along the way. So it's been a long journey, but it's, you know, for me and many people at Waymo, it is super exciting and very, very rewarding to finally see it become reality.

David George: Now we talk about safety and AI in many contexts, right? That's a big question, right? But, you know, here we are in this application of AI in the physical world. We have, you know, at this point, a pretty robust and increasing body of evidence that we are seeing tangible safety benefits. So that's very exciting.

Dmitry Dolgov: Yeah, I always say to people, it was a long journey and very costly and expensive along the way, but this is probably the most powerful manifestation of AI that we have available to us in the world today. I mean, you can get in a car without a driver and it's safer than having a human. And that's just remarkable.

David George: What were some of those humbling events along the way? In those early days?

Dmitry Dolgov: Oh, I'm sorry. I remember one, there was one route that we did that started, I think it started in Mountain View, then went through Palo Alto, then went, you know, through the mountains to Highway 1. Then took Highway 1 to San Francisco and I think, you know, went around the city a little bit and like actually finished for Lombard Street. So like in 2009, 10 people. That is really complicated. 100 miles from beginning to end, right? So, yeah. Human drivers would fail at that task, I think. So, yeah. Yeah, yeah. So, you know, we're doing it one day and then we're driving, we kind of made it through the Montenegro Palo Alto part. We're driving through the mountains and it's foggy, it's early morning. And then we're like seeing objects. And, you know, objects, like random stuff on the road in front of us. There's