Field more of that to come right here right now on stage we have somebody who has been in the news quite a bit lately his company has been in the news quite a bit lately. Actually. I had a great memory I looked back it's been just about a year I wrote an article that was a it's time to admit that self driving cars are never going to happen. This was in October 2022. Me and Kyle had a fun exchange on Twitter about it but there they've seen like a dozen or over a dozen cities their pilots so bad take I have a lot of bad takes on the website. Go read those. Check out my archive. You'll see how many times I've been wrong. It's a lot. That's why he has a lot of money. I have no money and I'm here. But he's co founder of Twitch before and then now founder of cruise as I just mentioned, which are running in San Francisco. So please welcome Kyle VOGT And your moderator Kirsten Korosec.
All right, here we are, again, five years later.
It's good to be here. But that was kind of fun. There's a lot of people against a visa for we're skeptical about them. So
yeah, I mean, do you have any last words to Darryl, you want to you want to public apology or?
No, no, no, no, I guess go for a ride. That's it.
All right. All right. Well, thank you for coming. And thank you, everyone for joining us today. It's great to see all of you here. Darrell did say something on just a moment ago when he talked about the number of cities, the pace of announcements that you have had actually since last October. It's kind of crazy. I think it's 14 cities now that you're there testing pilots and even fared rides. And I every time there was an announcement I have was one to ask you. Why? Because on the one insane
Well, I mean, you're right. So when we work on a technology like this, the goal is to get to massive scale of it anywhere for anyone who wants to use it. And so when we add a new capability to the system, we find that now it sort of checks the box, not just for one city, but maybe three cities, and we add another feature and maybe it's heavy rain, and now suddenly it's eligible for 10. And so the rollout becomes exponential because you know, each new capability unlocks more cities. But as for why so many cities, our view is the scale is going to be very rapid. We're going to build, you know, 1000s or maybe even more next year. And one thing we've learned especially being in San Francisco is it takes communities time to add to change. And so we want to you know, give those communities time to acclimate get familiar. You know, before they see a lot more vehicles, and so we're spreading that out. We also learn a lot, you know, every situation or encounter we have in a city like Houston or Miami. Some of those are very distinct from what we see in San Francisco. And as we adapt the system to handle those situations, we also find that works better in San Francisco. So this notion of fleet learning and every vehicle out there being able to help the cars in San Francisco is pretty cool. So you don't think
it's a distraction to you know, go wide right now when in San Francisco. There's there's been some controversy, to say the least.
Yeah, well, I mean, this comes down to how we run the company. And we made a big change in the last couple of years as we shifted from an r&d company, where the only thing that mattered really was was getting the product ready to deploy and reaching the level of safety. We needed to put the cars on the road. But since the last couple years, it's been about scaling and so we reorganize the company. So we have some teams who their sole focus is to be a launch team to show up in a new city, lay out the infrastructure and figure out how to operate those vehicles. And then there's an engineering team who's focused on the AV as a platform that can operate the same way across all the cities. So since we divided things up like that, it's actually not you know, a lot of extra work for engineering to focus on multiple cities. But certainly, we're influenced heavily by what happens in San Francisco. It's one of the most challenging markets in the country.
Yeah. Challenging to operate, but also challenging, perception wise and how Cruz has either been, I guess, embraced or repelled a little bit. You told me not too long ago that in the past AVS are here to stay, and the tech is working. And yet there are people in San Francisco including city officials who disagree and who are actively working to either limit cruise or revoke the permit. And I'm wondering if you think their concerns are valid?
Well, depends what your priority is like our mission. Our purpose is to get driverless transportation out there because we know it's safer. We know it's gonna be more accessible. We just announced a wheelchair accessible AV the first of its kind, gonna have a massive, positive, massive positive benefit on on society. So when I look at an individual city, the data is very clear like across the first million miles, we saw a three quarters reduction in the type of collisions that cause injury. And so as a society that that's probably the number one that's that's where we have to keep our focus. That said, we get it like in the early days of AV, sometimes they do things differently than human or that look a little awkward. And I think that gets a lot of attention. I understand that. I think it's it is newsworthy in many ways. And it's kind of fun as a society to poke at the differences between EVs and humans, but if we're serious about safety in our cities, we should be rolling out the red carpet for AVS there's there's really no way to look at it if safety is the top priority.
There are though some kind of funny incidents like going into the wet cement that was one that went viral and then the orange cones that disabled both crews and Waymo vehicles, but there were some more serious ones and specifically, you know, just days after the permit that light you finally after years of fighting for this to get full 24 hours, seven days a week was a collision with an emergency responding vehicle. So is there a way to what are you doing to decrease those incidents?
Yeah, so I'll say one thing to realize is that you know, crews out there we have more vehicles on the road. Actually, just last night, we crossed 5 million driverless miles so that's some news for us. And so because we're out there having the most miles the most vehicles, you're going to see these rare events that do happen, you know, at some frequency, with crews first and so I see this as the beginning of a conversation with regulators with city officials on the reality that these vehicles are here. They have a positive impact on the society. But we cannot expect perfection, there still will be collisions, those collisions will occur far less frequently. We hope and that's what we're seeing in the data so far. Than with human drivers, but expecting zero. Meanwhile, when we have you know, just a few weeks ago, someone running over and killing a four year old in San Francisco, just a couple of blocks from here. Like we have a serious problem in our cities. And if we expect perfection from this technology and demand nothing less we're gonna do a lot of damage.
Is it possible technologically speaking to actually eliminate all these problems? Or do you expect that there will always be something a potential incident that happens with your robo taxis?
Well, you know, we can't beat the laws of physics. So there's always going to be some collisions. But we what we need to know today is that we take this for granted. You know, almost a full 737 worth of people die every day on the road from car accidents, and we've come to accept that. I don't know why that is, other than there's no viable alternative. You still have to drive to work. You have to take your kids to school, we have to go to the grocery store. But this is the first time there's been an alternative presented that can actually do better. And I think we've got to embrace that. Do you
think it's because humans we accept humans like injuring or killing other humans and we just can't wrap our minds around a robot or a robo taxi, inflicting any sort of injury on a human that that's the issue or is there something else?
Well, I think like if we're being really rigorous about safety, and again, I think if you look at our federal revenue regulators, state regulators, they do take a lens on like, how do we have less harm on our roads, whether it's a new technology or road safety improvements, whatever it is, the goal is better road safety. And so from that standpoint, I think we do have to accept that perfection is not the goal. It's a reduction in the things we see every day. And certainly I think some people are going to feel you know more upset about when they read about a collision involving a robot versus a human because we want to forgive, you know, people who make mistakes. But again, the focus has got to be on safety.
Well, I'm glad you brought up safety because just yesterday, Speaker Pelosi and Representative Mullins sent a letter asking for more data to help determine safety and they sent this to federal regulators specifically NITSA. And in your perspective, do you do they need this data to determine that cruise or any Robo taxi is safe enough?
Well, so that safety data they're asking for we already do report, we're compliant with all regulations. And so both the state and the federal government have access to crash data. And it's that same data that tells the story showing AVS are beating humans on the most important metric, which is safety.
So is this just political theater or are they not understanding what's going on here?
I think this goes to show that there's a lot more conversations that need to happen between AV companies and government to help them understand not just how this technology is already performing, but the rate at which is improving. You know, again, I if you look, I've been on the stage a few years ago right at that point. ABS were always had people behind the wheel. They couldn't drive in cities without drivers, they couldn't operate during the daytime, all these limitations. It's improved exponentially since then, we're talking like, you know, 10x improvement in some of these metrics year over year. And so, you know, like betting against AVS right now is like looking back and saying Facebook is never going to make it because it's only on college campuses. It's a very sort of short term view on things. And if we take a step back and just look at the way this technology evolved, and where it's going to be a few years from now. I think there'll be a lot more people excited about this technology than there are and again, it's the opposition and or concerns or critiques by media and, you know, some people in government, a lot of these people haven't written in NAV or talk to the 10s of 1000s of people now in San Francisco who use this all the time, and I absolutely love it.
I'm glad you brought up Facebook because you're right, you know, back then it was like this college, you know, it was great and, but scale didn't necessarily make Facebook better. So how are you going to eliminate that issue, which is scale, but maybe not bad for society? I think we can all argue that social media hasn't been amazing. For us. So how are you going to prevent that?
Well, I mean, I don't. I was trying to make an analogy about the the betting early on something in the early days. I wouldn't comment on the impact of social media on society. It's not my area of expertise. But one things that we know about AVS so far is the more of them are out there, the more data they collect, the more examples of tough, tough situations they encounter. And so, you know, as a human as humans, we put kids on the road and give them access to cars, maybe when they turn 16. And they maybe have you know, a year worth of driving experience maybe 10s of 1000s of miles. That's not how AVS work. You know, when we hit 5 million miles like we did last night, that mean now means we have the equivalent of several lifetimes of human driving data. And we're using all that data to optimize this AV system. And so you know, over time, it's going to have the collective knowledge of you know, 1000s of individuals and be able to beat human performance on every aspect of driving. And so in that world, you know, more scale, more AVS actually creates this network effect where the benefit to society also increases with the scale.
How does that compare to then the millions of miles that you've done in simulation is one more valuable than the other?
Yeah, I don't really compare them exactly like that are simulations. You know, the goal there is to make sure that when we're changing things in our software, it's getting better on a bunch of different dimensions. Driving is really hard to score. There's certain maneuvers and caution around things and so we want to make sure it's going up into the right and then also to tease out new and challenging situations we haven't even seen on the road yet. So it's not so much the raw miles as opposed to the diversity and randomness that we encounter in simulation. And so I would say it's a critical tool for us when you look at vehicles like the origin that we're hoping to bring to production really soon. That's the first vehicle ever manufactured without a steering wheel in it. And because it has no steering wheel, we can't just drive it around on roads with someone behind the wheel taking over you know if it if it ever does anything strange. That vehicle has been validated almost entirely in simulation where we can do things like compress the millions of miles of on road driving that, you know, took the first time we went to driverless cars into something that can be done in a matter of months. It's really impressive.
Now a few weeks ago, I think just last week, you said publicly that you're close to getting this exemption from NITSA which is the same federal regulators that Pelosi just was, you know, asking for more information. So is there any concern that this new focus on Cruz could potentially derail that exemption?
I mean, hard to say I can't really comment on you know, what, what regulators may or may not be thinking but I will say we work really closely with NITSA. I think they're, I appreciate the difficulty of their job because if you pull up the vehicle code, which they're supposed to hold manufacturers accountable to it was written in the 70s. And there's sections in there that say the driver must you know, look at the rear view mirror or the side mirrors. What does that mean when there is no driver or the drivers computer? And so they've got to, you know, balance on one hand, the promise of this technology and the safety data showing it's, you know, has has a huge benefit over humans and is growing from here. And on the other hand, the statutes that you know, they need to be strict, strict compliance with that never contemplated the type of technology we have today. So it's a tough job. We're trying to figure you know, work with them and help them figure that out.
It's interesting, because there's at least one other company that took a very different path Zooks they chose not to go through the exemption path. They are now getting some questions. Why didn't crews just do that and put the origin on the road?
Well, I mean, work closely with regulators and you know, in this case, and it's and we think this is the ideal path to go but you know, these conversations are evolving, so it's hard to say how it will turn out at the end or you know, in what form these new vehicles will reach the road. You know, in terms of regulatory pathways, there's, there's several ways to get there.
We were talking a bit ago about pressure, sort of specifically in San Francisco. We're not seeing the same amount of pressure in some of these other cities that you're starting to go to, and there is a push to have both Waymo and Cruz's last permit revoked. And if that happens, is Cruz committed to staying in San Francisco?
Well, you know, look, if it gets to that point, we can have that discussion. But you know, there was a democratic process here. Like the Public Utilities Commission who voted on that permit, had hundreds of people come up to speak, you know, dozens of people got to submit letters of support for AVS that were far more letters of support than and against, and that commission ruled defeat will have spoken when it comes to whether or not they want abs.
In terms of the exemption, you have now two origins right where you have the wave which is the wheelchair accessible one, which was just unveiled on Thursday and then the regular origin one when that exemption goes through, presuming it does I know that the origin right now is in Austin. But what are the next cities? What's on your to do list or your priority list of where you want to bring the origins next and specifically, will you do the wave wheelchair accessible one in different markets or both at the same time,
the wave is probably a year or a little more behind development than the primary origin. So just just for context, the origin is a six seater vehicle. first of its kind with no steering wheel. But the reason we built it this way and with seats that face each other and have a ton of legroom was to encourage pooling. So you know the notion of a share a real shared ride, where, you know, one of these vehicles pulls up, there may be someone else in it, and instead of, you know, trying that on a ride hailing service today where you've squeezed into the back of you know, a Ford or compact car next to someone you don't know and you're shoulder to shoulder with them, you actually sit across from them and you can see them and feel comfortable, you know, in your space. That's something we've heard a lot from riders. And if you can get a higher rate of pooled ride adoption in a city, you need less vehicles on the road, you'd actually do a better job in terms of utilizing the road surface in a city that you allocate to vehicles for moving people much better than personal car ownership. And that's the direction we should be going but when it comes to the wheelchair accessible vehicle, which I'm super excited about, because I don't know how many people in the audience have, you know, friends or relatives with various disabilities, but the status quo is pretty broken like paratransit and other things exist, but they're not convenient. They're usually not very expensive. And certainly if you're you know, a person using these you don't feel like you have the same freedom of mobility as someone who can pull out their phone and get into you know, now a driverless car. The challenge with that vehicle is it's never been done before. No one's trying to build a wheelchair ramp that slides out of a car with no driver. And so there's a lot of interesting engineering problems, regulatory problems, safety problems. That we're sorting through. But our goal is absolutely to introduce that alongside origins and cities and, you know, have it be as equitable and accessible as we can.
So what cities are on your priority list then? Once the exemption
Yeah, so well, so Austin is you know, sort of tip of the spear for us we have developments that up there. The regulatory environment is more streamlined compared to San Francisco or California. So we did some of early testing there and it's likely that you see origins you know, in those cities are basically where the, you know, cities that are ready and willing to have this technology will get it first. And I think California has, you know, just a longer process not has nothing to do with whether California and San Franciscans want it's more or less it just takes longer to bring them to market here currently. So
what's that really feedback like where are people really, you know, either city officials or you know, everyday citizens want origin asking for that.
I mean, everywhere. The funny thing is about like, you know, we talk to city officials or regulators, and there's a huge difference between people who initially oppose it, and then people who have actually tried it, you know, the level of opposition, if any, drops through the floor. And I think that's because you ride in these vehicles for a couple of minutes. And you get it. And your mind starts to wonder about all the opportunity for how it can improve mobility in cities, how you can augment last mile transit or transportation for public transit systems. People get that within minutes and then they immediately start saying like, how can we get more of this? How do we scale it up? And then of course, we see that from the riders you know, it's about half of half of the people who first try and AV are naturally very skeptical, and they they're not even sure if they're gonna like it. They're there they start asking themselves like well, what if something goes wrong like what if this happens and I'm in the in the robo taxi, maybe academically, they've heard that there's, you know, safer or better than humans, but they're apprehensive and then after that, first ride two or three minutes in, we survey people again and 90% are ready to do their next ride. It's like an immediate conversion. And that applies universally to you know, the residents of cities. You know, regulators city officials, we see the same thing across the board.
The origin you specifically talk aspirationally about getting people to share. I would like to ask the audience right now, how many people used Uber share or Lyft pool and like sharing with people I see 123 more, okay.
That's actually a fair number.
I'm actually temporize. Yeah. Um, how are you gonna get people to share rides in a robo taxi people don't like sharing.
Well, one of the one of the reasons probably a lot of people raise their hands here is there's economic benefits like when two people share the same vehicle going, you know, along an overlapping route you can lower the cost to both people and when you lower the cost of something like that, it reduces the friction to using it. And, you know, with sufficient densities, vehicles, you see that the impact on trip time is fairly minimal. And when the experience feels actually better even than riding an you know, an existing rental vehicle by yourself, arguably writing in an origin with a stranger stranger is still going to be better than riding in the back of someone else's car where, you know, it may be their space that smelly, smelly or whatever it is, like all the things that can go wrong, or with someone if you don't know if they just got off another shift and they're tired or that you know, they had a few drinks or whatever it is that variable you just take off the table and sitting in a large vehicle is the shared space very comfortable facing each other. I don't think that's going to significantly affect usage compared to what it did for you know, attempt 1.0 at this problem.
You talked about being price competitive from sharing. What do you see it competing with Uber, like how close to parity will you compete with them? initially and then eventually once you're at a larger scale?
Well, I mean, start by looking at ride Hill as it exists today. It is a luxury good you are paying for a timeshare of a human driven chauffeur and overtime we know that we probably expect the you know the wages are going to increase you know the the pool of drivers and everything may shrink as is a popular you know, percentage of the population all these things are driving human driven rideshare prices up. And then when you compare it to an AV, you have a technology that today is probably relatively expensive compared to the cost of a regular vehicle. In a couple years time. We expect that to drop substantially. And so the cost structure is much lower. The experience is good today. And getting better at a rapid rate. And so I think that that delta there like sort of higher price, worse experience, lower price, better experience is going to be really good for consumers and it'll give us a lot of flexibility as a company to you know, moving price down to, you know, go back and forth between the usual trades of like, how important is it to get to cashflow positive in an industry that is really fixated on that versus, you know, expanding this technology getting the societal benefits and larger market share.
You're removing the driver, but there are a lot of people that support a single Robo taxi. How many people win win today in San Francisco and a Chevy Bolt goes out? How many people are supporting that? At that time?
I don't know the number it's probably it's probably less than one as of today when our fleets fully functioning. You know, our peak right now is around 400 concurrent EVs, but it's not really supposed to work at this you know the economics aren't supposed to be optimal at this scale. If you think about this, when you start to get real benefits from moving to Robo taxis, it's when you have a very large fleet relative to you know, a small amount of infrastructure or fixed cost. And today, you know, for the human driven ride Hill, you've got one person in every vehicle, probably a fraction of a person, you know, allocated inside the parent company, and then all the people it takes to support that indirectly, you know, through charging, maintenance, cleaning, all those other things. So it's a lot and even at relatively modest scale, a few 100, AV few 1000 that that flips and moves very in the positive so I think one thing people don't really comprehend is this is a business that at small scale, tiny scale is not interesting as a business at moderate scale, even 1000s of vehicles is very, very attractive. And when you go larger than that, it's an excellent business, but that that advantage and all those economies of scale again pass through to lower prices for the consumer.
So how many Robo cat taxis do you need to have on the road to be profitable? What is that number?
Well, so I can answer that in many different ways. The reality is depends on whether the priority for us is fastest path to profitability, or the whether we want to, you know, go after market share and get more people on the platform more people using it. But it doesn't, you know, if you do the math, like, you know, a single car operating, you know, 1520 20 plus hours a day, looking at, you know, ride hill like revenues, it doesn't take that money.
Can you give me a number?
I'll leave it as an exercise to the to the audience. Okay.
Thank you all now, please. But is it in terms of scale? It's also a number of cities. So you couldn't possibly scale or have that enough cars in a single city? Is that the thinking?
Yeah, for context, you know, San Francisco is a billion dollar right Hill market. That's basically one city and sort of, you know, larger metro area. We're testing in 15 right now. And that will be more next year. And so you know, as we ramp up production factory, you've got a continuous flow of vehicles coming out of a factory. And then we have, you know, some people running an algorithm that decides which cities they go into first, but the goal there is to get to scale as quickly as we can in terms of the total number of AVS to make this business ultimately profitable. And sustainable. But also balance that meeting with community has given them time to acclimate to become familiar with the technology, making sure all the interactions with first responders and local community groups are working. And so we have these things in balance so that we don't think we're ever going to grow in any one. City too fast, which is just not a good thing for anyone.
Is there a certain number of vehicles that a city the size of San Francisco, not specifically San Francisco that one this population base can handle at any given time you said about you have about 400 now and what can they actually sustain?
I don't know the number there. I mean, it'd be depend on the adoption rate, and how much people still prefer to use their own vehicle and park it on roads versus using Robo taxis. I mean, ultimately, if you extrapolate forward and you see that these cars decreasing cost, they continue to prove their safety performance. They get much better on you know, adapting in ways that that cities find agreeable and preferable and the cost goes down. I think that as you see more pool rides, the question will be, do we want as many human driven cars on our road like at what point does it still make sense? You know, if I told you, you know, five years from now AVS are gonna be 100 times safer. And you have regions of your city that have high pedestrian and cyclist traffic. It would almost seem reckless as a city planner to allow, you know, one version of of transportation that's 100 times less safe than the other to coexist in that space. So we're looking forward to these conversations, they're gonna be awkward. They're gonna be really uncomfortable when we come to terms with the fact that humans are not the ultimate and best form of vehicle drivers. And even if you believe that you should have more forms of other transportation, which I'm totally on board with. You know, as long as there is a need for vehicle driven transportation. I think we're going to want them to be AVS
Are you going to would crews put itself in a position where they work with city officials to lobby to actually eliminate human driven vehicles and urban centers? Or will you take a sort of let them figure it out by themselves? Five years from now, the in this scenario,
look, I don't think we're gonna have to do a thing. I think the residents in the community members in a city are going to stand up and say we want safer safer transportation in our cities, AVS of the path, you know, what our city is going to do about it? And if we're called to join that and help in that in some way we'd be honored to but this is really, I think, a question that society is going to have to grapple with. And you may not see it today when I say that, you know, there's only only a 75% reduction in collisions that call it could cause injuries. That's huge. Like if that if that if we add a you know, a couple orders of magnitude to that improvement. I just don't I don't see how it would be possible for cities to act. in good faith. You know, if safety is the number one priority for a city, it will be hard not to have that conversation and to say, look, we you see this in European and other cities where they don't allow cars in certain areas and things like that. We're gonna have to do something like that if we want to keep up and if we want to hold the bar high for safety, and so it's really a question for how much are we going to tolerate in our cities before we demand something better?
You mentioned a moment ago that you're in 15 cities now either small pilots, of course in San Francisco has fared rides and that you have plans to go to more cities next year. So what can we expect? How many more cities do you plan on entering into next year?
Well, let me go back to kind of where we started on the on the technology. So I personally love the technology we work on. It's really fun. But one of the challenges for us is Do you Do you boil the ocean and try to make a technology that works everywhere on day one? Or do you start where the technology is best suited and grow over time? We obviously chose the latter. We started in San Francisco where you know we have we have fog and occasionally smoke in the air like we do right now. But there's no there's very rarely ice or snow or you know, that kind of weather. So, you know, if you look at the cities we've announced and kind of plot them on a map, you'll see they're kind of in the Sunbelt. And so a couple years from now, we have a new version of our vehicles coming out that is adapted for cold weather. And so there's a whole bunch of fun engineering details behind that but a simple example is the sensor pods where all the cameras and radars and lidars are they have a heating elements built in so they can melt ice and snow that would accumulate on them. You can't really operate in a city with tough wet weather unless you do that. So you'll see a start in the lower half of the US and then move up vertically as the vehicles come online to handle more more challenging weather.
Is you said in a couple of years, there's more winter winterized versions, is this going to be the Chevy Bolt or the new Chevy Bolt that we're going to see in a little bit from GM or is this going to be the cruise origin?
The origin is that is the first version we'll see. So you know, something that may not be familiar to a lot of people who work on you know, software based companies or mobile companies is the development cycle in the automotive world is very long. It can take you know, four years to go from when the technology is ready to when you see it in a mass produced vehicle. And part of that is because when you have the technology working, that's not the level of reliability, you need to to have that operate you know, for years potentially without any replacement. You can't have every part on your vehicle randomly failing once a month and needed to swap it out. So the bar is very high. When you put a vehicle out on the road is subject to corrosion. You know, it's like salt in the air, water, spray, mud, all these other things. And so the state, thermal extremes like basically, you know, 120 degrees in Arizona plus it's been sitting out in the sun getting hot, or really cold weather where basically it has to your computer has to boot up when it's freezing outside quite literally. And then on top of that, not only the reliability, but then the supply chain has to be spun up which is often global in nature to have all these components feeding into a factory where you cannot skip a beat and have that factory line, you know, shut down so for years from start to finish. So you can imagine if a couple of years from now, you know, we want that cold weather stuff. This is something that we've been actively developing for years already.
That's interesting. You talk a lot about factories and scaling up and of course right now, there's a strike. What are your thoughts on the United Auto Workers strike happening right now?
I'm not gonna comment on that. What I what I will say is, you know, I think General Motors is working in good faith to try to find the best outcome they can for everyone. But that's all I can really say about that.
Can you talk at all about whether it will affect crews because it could, as you talk just mentioned, a global supply chain needing workers on a line to bring you know, something you've been working on for years to public roads is winterized version which sounds pretty cool. Could that could this strike threaten cruisers plants?
Yeah, I appreciate the relevance of that. But, you know, it's still not my place to comment on that. And, you know, I'm going to resist the temptation to want to speculate Hey, sure.
Okay. We're talking about winterizing. So can we expect you said in two years, will you sort of ahead of that, go into some Midwestern cities potentially next year start, you know, looking around, maybe testing a little bit in anticipation of this winterized version of the origin
is that's likely I mean, one of our strategies, as I mentioned is, well, we started in San Francisco, pretty complex city. We see a lot of things here that we see in other places. But each time we go to a new city, we usually see at least one thing that's new. So you know, in Austin, I think it was the traffic lights, some of them are horizontal instead of vertical. And so we have a scouting process now where we do either send, you know, send ABS or other data collection vehicles, to kind of seek out these whatever might be unique about a new city well before we intend to operate there just to decrease the probability that we have a delay because something something blindsided is something new. But what we're finding is as we go, you know, we learned a lot going from San Francisco to Austin and Phoenix. Those are the two cities that we did just over a year ago. By the time we got to Houston, it worked almost completely out of the box. Like each time, I think there's going to be a smaller and smaller number of new things or changes we need to make to adapt to cities. So we'll still scout and you may see a views on the ground and some of the cities but you know, from an engineering standpoint, we're at the point where, you know, we build a system or a capability or improvement once and it gets picked up in all the cities. We operate in.
I'm going to try and ask this question a different way. You mentioned that the cities that we if we plotted on the map, they're all in the Sunbelt. So next year, will we see cities that are beyond the scouting we're actually piloting and maybe looking to scale outside of that Sunbelt?
Less Likely because you know, if we if we did that we'd have a service that we'd have to flip on when it's sunny and turn off if it if it starts to snow. And while that is possible, it's something we'd explore it's not ideal from a customer standpoint
that we saw it didn't really work out well for scooter companies, right? So sure.
And we actually had that in the early days of cruise in San Francisco. We started very conservatively on fog, for example. And so if an AV would go, you know, towards the sunset region of the city, and then there's a sudden onset of fog. That vehicle would have to return back to homebase. But after we've collected enough data and improve some of our perception models, at this point, we can handle 99% Of all the foggy situations in San Francisco. And so this is another example of how you see these weather capabilities. It's not necessarily like a light switch, you turn on and off. We have a great deal of control on you know, setting safe thresholds for operation and having all sorts of mitigations like the ABS just deciding to go back to home base and park until the storm clears. You know, as we take on more and more challenging weather conditions.
Five years ago, you said our christmas master plan and use it actually Chris's master plan 2.0 Because originally it was a slightly different company. You said the plan was to go into dense urban areas and then you said something that was surprising to me would you said and then roll and I'm wondering if that's still in the crews master plan.
Taking a step back and I mean, our our mission is not just about improving transportation in the subset of the population that uses rideshare. I look forward to the day when any of us can get this on the vehicle. We own a vehicle maybe there's a new ownership model where you lease the vehicle or you share it with a community or group of friends. There's all sorts of opportunities that exist beyond ride hill and you know at the moment we we saw evidence that this technology is reducing collisions, that reducing collisions overall and the ones that results in injury. We felt even more of a poll and this this need to get this into every form of transportation that we have. And so you know, there are some technical hurdles we have to drop the cost of the technology so that it works in these new models are different different business models and ryedale. And we have to adapt the technology to work you know, rely less on things like HD maps, which we use some version of today. But just like you know, we started a few years ago on this on this cold weather capability. reducing our reliance on maps are getting to the point where you don't really need an HD map as it exists today is something that's an active development, I think going to unlock not just you know, better operation in suburban and urban environments, but will lead to rural environments. So this stuff is going to work everywhere. Okay.
So it's fair to say that you're working on it. Are you working on it closely with GM as a parent company because Mary Barra, CEO of the company has talked about the the likelihood of personal Robo taxis So is this something where the two companies are working together?
Yeah, it's inevitable. The question is exactly when and in what forms but it's it's a it's it's not only you know, an unprecedent business opportunity, as we've seen, with companies dabbling in cars with various forms of autonomy today, but one of the most fundamental shifts we'll see I think in our society in a long time, not not just safety, but you know, if you think about the value proposition of, you go to a car dealer and there's one car on the lot that you have to get in and drive yourself 100% of the time. Maybe there are moments when you can take your hands off the wheel, but you have to be you know, and you can't be doing anything other than driving. And then maybe there's another car in the lot that you sit in the backseat and you can sleep or do work or talk to your family or even send it completely empty to go pick up your groceries or you know, take your kids to school like that is going to be very profound. And I think that's, that's within reach for this kind of technology.
Where's delivery on the roadmap, because you do have a pilot, I believe with Walmart right now.
Yeah, so we're working closely with Walmart which you know, people know that brands but they often don't realize it's the largest retail company in the world. And we've done driverless deliveries with them. So experimenting with different uses there. And one of the things we've done with the origin is make it so that there's a way that it can very quickly convert from moving people to moving things and the benefit there when you do both things using the same fleet of vehicles. is you get a higher utilization of that fleet, and then a lower cost per trip. So delivery and movement of people are not just two very good businesses, but when you put them together on the same fleet, it makes both of them better.
You have 3900 and employees right now. So it's a lot but you've been talking a lot about scale. How many more people do you need to hire based on all these cities you've been in? And are they mostly engineers or an operations?
Well, so yeah, that's, you know, one of the things we look at as we went from an r&d company to one that thinks more about operating large fleets, is reducing the ratio of, you know, engineers to vehicles. And, you know, we expect that to grow. It's already growing, I guess, sub linearly meaning we're adding cars faster than we're adding engineers, but we expect that gap to increase dramatically. I think we've got an incredibly talented engineering team. And we will not need, you know, to significantly grow it to handle fleets 10 times 100 times, you know, even 1000 times larger than what we have today.
A big part of scaling a company and hiring people is culture. And we've talked about this before, this is your third company, Justin TV. Twitch. Did you get it right this time?
I'm a believer that every company founder or executive is on a continuum of personal growth. And so I would say that the culture of cruise and the sincerity with which we treat our values and behaviors is much higher than I've ever had in my career. You know, we try to live them every day little details like people you could probably all use Slack for communication and things like that. But random employees across the company will will insert little hashtags of our behaviors like in mid sentence in like half of the community or the company communications that we have. So little details like that, where there's constant reminders of why we do things and people acting in accordance with those behaviors has been super compelling. And to have a mission that is so I mean, we're lucky we have a mission like almost everyone has someone in their family who has been hurt in a car accident or know someone who died in a car accident. I've high school friends who did. And so just to know that you're doing something with the talents you have the talents you've developed that actually matters, and it's not building toys or fun things to play with on the internet. These are, these are real visceral problems that we can solve. It makes it maybe it's unfair, because it makes it a little bit easier to build a really strong culture around that and strong dedication to what we're doing.
So you've mostly answered my question, but have you learned anything recently? Do you have it perfectly dialed or is there still more work to go?
Well, I guess I I don't think that's possible. Perfection. You know. I have a long list of of improvements. I like to make both for myself, other leaders in how we run the company and the culture. Where do we start? I mean, I think one of the things that we're particularly bad at, as you mentioned, we had, you know, dozen announcements, big announcements in the last few months, where we're so accustomed that and we're so driven and heads down and focused on the next thing that we never take time to sit back and actually appreciate what we have accomplished. I think one of the most important things for anyone working on a team is to feel that after you you put your you know, blood sweat and tears into a project or a deliverable and you hit it, you gotta you gotta feel rewarded for that. And, you know, I wish we did more of that we're working to do more of it, but you know, it's hard when you're moving fast to also take a step back and be deeply appreciative for you know, the effort that everyone did put into getting to where we are. Right.
Thank you so much, Kyle, for joining us. Thank you for everyone for sticking around. Thanks for Kyle. Appreciate it.
