 Good morning, and thank you for the privilege of the chance to talk with you today. In 1913, Henry Ford gave the world a gift. He figured out how to make the automobile accessible to the common man. And over the last 100 years, what we've seen is the transformation that's had across America and around the world. It's affected the way we live, the way we get around. It's affected the shape of the cities around us. We're now on the cusp of, I think, the next great democratization and transportation. And that'll be the delivery of self-driving vehicles into the ecosystem. At Aurora, this is our mission. We deliver the benefits of self-driving technology safely, quickly, and broadly. Safely because it's important that we do this right. Quickly because if it's worth doing, we should do it now. And broadly because if we build 10 of these things, who really cares? We want to have an impact at scale. Why do we think this is so important? So that gift that Henry Ford gave us has had profound negative consequences as well. Within the United States, 40,000 people die on our roads every year. Around the world, that number is more like 1.3 million. The good news is that 95% of those accidents are due to human error. So if we can bring technology to bear, we can interest them. And this isn't just a very painful social issue. This is an incredible economic issue. The drag on the US economy of traffic accidents is estimated somewhere in the order of $1 trillion. About $150 billion of that is direct cost. And the other $850 is indirect cost. So if this is a real, real, both social and economic impact, our cities and the world we move through could also be dramatically transformed. By making the way we get from point A to be more efficient, we can reduce the amount of carbon we put in the air. We can reduce the time wasted getting from one place to another. In San Francisco, something like 30% of the traffic at peak is people looking for parking. Imagine if we take that out of the mix. And then in cities like Los Angeles and Houston, something like 30% to 40% of the adversary of the city is dedicated to vehicles. If we can transform that space into something that is now green space, real estate, it would be just incredible. And then finally, this very human element to this, 3 and 1 half million Americans never leave their home. Over 1.9 million of those are due to the fact that they have some kind of disability and can't get from one place to another easily like you or I can. So if we can give them back the ability to take part in our community, that's important. And then for the rest of us, it turns out I'm not a car guy. But a couple of years ago, I bought a convertible. And I had that moment that people who become car guys car people do. I was on 280. I had the roof down on my convertible, driving along with my music on. And it was magical. And then I got to San Francisco. And the traffic stalled to a halt. And it took me the next hour to go 15 miles. And that was when it really hit me that even people who love cars hate commuting. And if we can give them back those 80 minutes a day that they waste, then this will mean something profoundly important to the quality of their lives over time as well. For me, I've had the, oh, I'm sorry. And then I forgot, we are building a company. And it turns out this is an incredible Greenfield economic opportunity. This is UBS's estimate that by 2030, the self-driving car industry is going to be about $2.3 trillion. So there's some real meat here to go after. So for me, the journey started about 15 years ago when the Defense Department announced the DARPA Grand Challenge. The idea was to drive 150 miles across the desert and not bump into too much stuff. This was the robot that we built at Carnegie Mellon. I think of this as kind of the missing Lincoln self-driving car space. A lot of vehicles before this had used lasers, radar, and cameras. But this was the first time we put that together with maps. And that turned out to be one of the key ingredients. So we built and developed this vehicle. We took it out to the desert, we tested it, didn't always go well. But we got this thing back together. We entered it, we qualified, we put it on the race course. The 150 miles we set it off in the morning, it was incredible. And it drove 7.2 miles. It got caught up on a berm here, almost literally burst into flames. And obviously, we were crucified in the press. But the Defense Department said, come back next year. And you know what? We'll give you $2 million if you win instead of one. And so we built a car. Again, this was one of my lessons about move fast and break things. Doesn't really work in this industry. So we got it back up on its wheels. We entered the competition. And what was incredible was that this year, this unwinnable challenge, this DARPA Grand Challenge, five teams finished. And so we got up on our soapboxes. The little blue vehicle here, by the way, was from Stanford. It beat the two red vehicles. At Carnegie Mellon, we were very sad. But this day, we got up on our soapbox and we said, this is exciting. We can go take men and women out of harm's way in the military. And the Defense Department said that's great, except where we drive, there's also other people on the road. So why don't you come back when you can drive around a course where there's other moving traffic and where you can stay on your side of the road and where you can deal with stop signs. And so they created the DARPA Urban Challenge. And this is a fun video from that. So that's MIT, the Black Vehicles from Cornell. And this was the, so as a Carnegie Mellon person, this is entertaining. As a Stanford audience, I'm sure this is fun. So I did not realize it was on loop. I apologize for that. So what's interesting about this, one is this is the first self-driving car on self-driving car accident. The other is that it's very representative of the key problems we face today. And that problem is that the vehicle here, the MIT vehicle, is trying to understand what the other person on the road is doing, in this case, a robot. And there's confusion. It thinks because the Cornell vehicle hasn't moved. It's good to go. The Cornell vehicle at some point makes up its mind to move. And then we have it coming together. This is the same kind of problem we face today. At the end of this challenge, though, they pride those two vehicles together, they apart, they finished. We had another couple of vehicles. The team that I happened to be the technical director for ended up winning that year. And so again, we got in our soapbox and said, we've got this. We can make self-driving cars. And the defense department said, you're right. Problem solved. Go figure out how to do it. And that was that for a couple of years. Until in 2009, Google picked up the torch. I had the opportunity to go there and help lead the self-driving car team for a number of years. And over that time, I think we kind of introduced the concept of what this could mean to the world and have actually pushed the industry forward to the point where now there's an incredibly broad set of talented people pushing this ball forward and trying to reap the benefits of this technology for society. And it's everyone from the tech companies to the automotive companies to startups like ours. Why is this happening now, I think, is one of the big questions many people ask. And there's two reasons for it. One is the configuration of the automotive industry today. Right now, the automotive industry is kind of under attack from three fronts. It's under attack from electrification, from connectivity, and from automation. And they recognize this. They recognize they're facing many of these challenges. And so they're looking to try and find ways to do innovative ways of doing business to move the ball forward. So that creates a moment. The other is that we're able to build off the giants that are out there. Cloud computing is now a thing. The advances in deep learning are incredibly important to this space. Just Moore's law and the amount of computation we can stick on a car today is profoundly different than it was even a decade ago. To build this, though, it takes an incredibly diverse set of people. At Aurora, we have folks with backgrounds across the industry. We have folks who've landed rockets, folks that have worked with the military, with agriculture. It's a very hard problem. A lot of what we think about is engineering, but we also have to engage on the social side of this. How do we communicate with regulators and help them understand the benefits and challenges of what we're building? When we think about what we're building, well, that's, again, broad. There's the hardware, so the combination of the computer, but also the sensors, the camera, the laser and the radar that allow the vehicle to see the world around it. And then there's the software that takes those signals, figures out what's happening in the world around the vehicle, and then how do we safely move through it. And then finally, there's the data systems and services that come along with this, things like the maps that allow the car to understand where it's going in the world. And this is what that looks like from the car's perspective. It starts out with a car understanding roughly how it's posed in the world. And then the red lines and the yellow lines here, the curbs and the lane markers, and they've come from a map, and they're now loaded into the vehicle. And then the cloud you can see come in is the laser data from the sensors on the vehicle. And then ultimately, the onboard perception system is identifying the bicyclists here with yellow boxes and other vehicles on the road with blue boxes. And at this point, we can then figure out how do we move safely through the world and hand it over to motion planning. In a more complicated environment, it gets more interesting. So here, we're applying interesting machine learning techniques and fusing those with Bayesian filtering on top to be able to drive through San Francisco. So we're again tracking the vehicles parked on the side of the road, the vehicles around us moving safely through the world. Driving, it turns out, is a very social activity. And expressing the cost function, expressing how the vehicle should behave is actually very subtle. And so in this video we're watching here, and it'll loop back at a moment, our vehicle is entering from an off-ramp or a non-ramp, I guess. And it looks to the left and it sees that big yellow truck and it sees a minivan. And at this moment, it has to figure out do I accelerate to get in front of the truck, slot between the minivan and the truck, and or fall behind the minivan? It turns out if you did the default thing here of just kind of get up to speed and hope, it becomes very uncomfortable very quickly. And so what we do is we take these vehicles out on the road, we have expert human drivers do behave in the way that they would to make their passengers feel comfortable, and then we're able to learn those cost functions over time. And that allows us to come up with a relatively efficient way to build a system like this that works well. And then finally, simulation is a key part of the way we build these systems. What you're looking at in this video is the black car is what happened on the road. Our test drivers were moving through a neighborhood they, by policy, disengaged because we were approaching a pedestrian. And then the red vehicle that you see separate from the black car is run in simulation. So offline, we can ask the question, what would have happened? And in this case, the vehicle would have yielded appropriately for the pedestrian crossing the road and then continued on its way. And this is a way that we can thoughtfully and safely test these systems out on the road, out on the real world without putting the public at undue risk. So where is this going and how do we kind of get there? My very deep belief is that this technology would come to market first through ride hailing and shared transportation services. And the reason why is fundamentally economic, that if we try to sell this as a feature on a vehicle, people pay $5,000, $10,000 for that, maybe. That's probably not enough to cover the cost of the technology we have to embed in the system. In contrast, if we can bring this technology to market providing a service of transportation and offering basically robotaxi capability, but at somewhere between half and a fifth the cost, then that will be clearly a clear economic winner, winner for the consumer and a winner for the people delivering that service. We also see an incredible opportunity to revitalize these cities, give people back their time, expand access to mobility, and increase safety in vehicles. And those are really easy things to talk about, but let me take you a moment to talk to you about what that means very concretely. So this is Steve Mann. He's a gentleman that I got to know at my time at Google. He's an artist. He's the director of the Santa Clara Institute for the Blind. He's incredible. It just turns out that he's blind. And that means every day he has a tax of a digital two hours to get from one place to another that you and I don't have. So by giving him just a simple moment of freedom like this, riding in a self-driving car, changes his life, allows him to engage in the community the way that he wants it on his terms. This gentleman is the brother of the best man in my wedding. His name's Steven Fletcher. He's a Canadian politician. When he was in his early 20s, he was a mining engineer. He was driving to a mine site in northern Manitoba on the way there. Moose stepped into the road. He swerved around it and hit a second moose and instantly became a quadriplegic. Now I'd like to think that if we have got our technology to the point where it's commercially viable, he would have never had that event. But much like Steve, if we could give him today the freedom to get around that you write for granted, that would transform his life and give him back joy that he doesn't otherwise have. He's had an incredible career. He was in the cabinet in Canada, but every day he's held back by his inability to move through the city. And then finally, this is my family. These are my two boys. Ethan here on the right is about half a year away from getting his driver's license. That is absolutely terrifying. And not because he's a great kid, he's smart, he's courteous, he's thoughtful, but if you look at the statistics on traffic fatalities, there's this terrible bathtub shape to it, and young drivers are particularly at risk. If we can deliver this technology the way we intend to, we can keep him and his brother and your families safer on the roadway. But I think that is something really worth working for. So this again is our mission. We hope you join us in it and we look forward to a transformed future together. Thank you very much.