Full text: Tesla Q 2 results meeting transcript: Robotaxi is extremely safe
Tesla’s total revenue in the second quarter was US$28.236 billion, a year-on-year increase of 26%, driven by growth in vehicle deliveries, growth in services and other businesses, and positive foreign exchange effects. Tesla’s Q2 revenue increased by 26% year-on-year, profits were cut in half, and capital expenditures increased significantly After the financial report was released, Tesla CEO Elon Musk, CFO Vaibhav Taneja, Vice President of Supply Chain Karn Budhiraj, Vice President of Automotive Engineering Lars Moravy, Director of Autonomous Driving Software Ashok Elluswamy, and General Counsel Brandon Earhart Ehrhart, as well as head of investor relations Travis Axelrod and other executives interpreted the financial report and answered questions from analysts. Investor Q&A session: Mainly about Robotaxi, Cybertruck and Optimus. Ashok Eluswamy: First of all, I want to say that the Robotaxi program works very well. Especially when it comes to safety, the project maintains an impeccable safety record. Currently, our self-driving taxis have driven more than 380,000 miles without supervision in six cities in two states. We did not have any major incidents during this period. All reports were of us being impacted by other vehicles while stationary. Let me stress again that to date, our operations have been very safe, with no major incidents in more than 380,000 miles. Historically, so-called experts have always claimed that to drive safely you need lidar, radar, high-definition maps, and every other technology imaginable. Here we show that this is not the case. With cameras alone, you can achieve safe, comfortable and affordable autonomous driving. This record is a huge validation of Tesla's entire approach to artificial intelligence (AI). Secondly, I want to discuss scaling. About a year ago, we launched a self-driving taxi program in Austin. At that time, we installed safety monitors in the passenger seats. Around the end of last year, we launched one of the first fully driverless self-driving taxis in Austin. The number of unsupervised miles driven by fleets per week has increased by double digits since the start of the year. I'll say it again, we've been growing at such a high compound growth rate week over week over the last few months. Not only that, but we expect this growth rate to continue at this high level for the rest of the year. Another thing worth noting is that in In terms of expansion, we put in relatively little effort. We anticipate that the time to expand into new cities will continue to approach zero until we are able to operate across the state rather than advancing city by city. We are able to achieve this kind of growth because we have a world-class AI cutting-edge development engineering team at Tesla. The self-driving taxi fleet currently operating is already running an early version of the v15 FSD software we mentioned earlier. For the v15 version, we have roughly planned seven major improvement directions, and they are all being carried out in parallel. Currently, 40% of the improvement directions in the early v15 version run by the self-driving taxi fleet have been integrated, and this is the system the fleet is currently running. As we continue to complete work on the v15 version, we will see that this car will have amazing safety and powerful performance. We're already seeing internal signs that all of our efforts to improve our technology are continuing to pay off, and we expect the returns will be significant, helping us maintain the same growth rate throughout the remainder of the year and into the future. Finally, a word about Cybercab. First of all, it is an extraordinary product. Anyone who tries it will instantly fall in love with the experience. We have adjusted our manufacturing targets to roughly match the projected growth rate of unsupervised driving miles. The v15 models that power the Model Y and other platforms will also be available for Cybercab.
All in all, robotaxi growth has been exponential so far while maintaining an impeccable safety record. Our technology investments are paying off, helping us sustain this exponential growth for the foreseeable future, and Cybercab is going to be great. About Optimus Figure , especially the Optimus 3rd generation model, which was designed to match not only human appearance, but also human functionality and dexterity. One of the main reasons for this approach is that having human form and function allows us to learn from humans how to perform a wide variety of tasks. This provides the entire world of data for training Optimus. Just like the fire department, we have access to a large group of people who provide us with data on all the workers in the factory. Optimus could learn a lot from watching them, performing their tasks in the factory. We also have a dedicated data collection team who can provide relatively small but very high-quality demonstrations for post-training of these models. This is all on top of any training we can do from data across the internet, which shows demonstrations of people performing various tasks in both industrial and domestic settings. The second phase is that we have a large number of Optimus robots practice tasks in what we call the "Optimus Academy." Data from the robots themselves performing tasks will be invaluable and help us close any small form factor gaps that may exist between robots and humans. This is also when the reinforcement learning cycle begins, with the robot initially trying some tasks, sometimes failing, learning from its successes and failures on these tasks, and eventually learning to master these tasks at a superhuman level. Our AI strategy for Optimus is consistent with this. This is the same end-to-end strategy that drives FSD, pixel input, and control output. Like the Fire Service, we want it to be widely available. In FSD, you get in your car, enter your address, hit start, and then it's just like you do with all your driving from park to park. The same goes for Optimus. You just let it do whatever it takes, and then you can complete the entire task yourself without having to do anything along the way. It is this team that developed the FSD, v12, 13, 14 and now v15 software that now also runs on Optimus. I'm very confident that Optimus will be a very capable robot. Analyst Q&A session: Morgan Stanley analyst Andrew Percoco: Thanks for the opportunity to ask questions, and the first question is around the Optimus humanoid robot. Elon, you mentioned before that the complete supply chain for Optimus is almost non-existent at present, and most of the components need to be developed and produced by Tesla. I would like to ask, in the process of promoting self-research and production, are there any external suppliers willing to invest simultaneously and build factories in the United States to help Tesla expand production faster? We understand that Tesla will control core technologies, but from the perspective of investment efficiency, supply chain partners share production capacity construction and can optimize capital expenditures. I would like to know your overall thoughts on cooperation with external suppliers, as well as the progress of communication with suppliers at this stage. Musk: Overall, supplier cooperation is very smooth, and major manufacturers continue to invest huge amounts of money to support the implementation of Optimus, Robotaxi and other projects. Among them, Samsung and TSMC are particularly prominent. TSMC's Arizona factory and Samsung's Texas factory have invested tens of billions of dollars to specialize in the production of digital AI computing chips suitable for humanoid robots and driverless taxis. Amazon has also invested billions of dollars in expanding its battery production lines. Although I feel a little unwell today and my voice is not in good condition, I still want to thank all partners for their financial investment and implementation support.
Kahn Budiraj: Just to add to what Elon said. The production capacity of Samsung's new wafer fab will be heavily tilted towards Tesla's future projects, with an investment scale of tens of billions of dollars. New robot-specific industry chains such as memory chip manufacturers, new metal injection molded parts, and flexible printed circuit boards are all simultaneously increasing their investment, which is completely different from the traditional vehicle supply chain. If it cannot find a high-quality supplier that matches its needs, Tesla will not hesitate to develop and produce it itself. We have top-notch manufacturing engineering and product design teams, and are fully capable of independently expanding our parts production capacity. Musk: Also thank you to Micron for reserving memory chip production capacity for Tesla. The supply of the current memory chip market is extremely tight. Micron has made a large amount of internal production capacity deployment and provided Tesla with a large long-term production capacity quota at a reasonable price. We are very grateful. Morgan Stanley analyst Andrew Perko: Thanks for the addition, my second question is related to driverless taxis. At present, regulatory policies regarding the mandatory requirements for autonomous driving sensors in various states in the United States continue to change. I would like to ask Tesla what regulatory policy demands it has when launching Robotaxi? Moderate regulation is conducive to the development of the industry, but excessively rigid regulations will restrict innovation. How does Tesla want the regulatory framework to be reasonably formulated? Kahn Budiraj: First let’s talk about regulation at the federal level in the United States. Our progress has been very smooth. The U.S. National Highway Traffic Safety Administration (NHTSA) has issued a new version of the motor vehicle safety standard FMVSS and recognized the purely visual autonomous driving solution. We are very grateful to NHTSA for its support. State-level regulations are clearly divided, and the stringent policies introduced in New Jersey are disappointing. As Ahsok said before, our core reliance is on the actual safety performance of vehicles to convince regulators and the public. The optimal supervision model is to set safety goals and leave it to enterprises to independently develop and implement paths. If supervision directly and rigidly specifies hardware solutions without focusing on safety results, we do not agree with it. Tesla has always used actual operational data to prove product safety, which is also the core support for our subsequent cross-regional expansion. Bank of America analyst Alexander Perry: Thanks for the opportunity to ask, I have two questions about driverless taxis. First, what key milestones do investors need to track that will drive Tesla to accelerate its fleet rollout and expand into new cities? Is it safety data (accident rate per million miles), or some other metric? Secondly, will Tesla reach third-party cooperation with online ride-hailing platforms to improve vehicle utilization? Or should we insist on vertical self-research and self-operated Robotaxi business in the entire industry chain for a long time? Musk: Yes. We want to be vertically integrated in our self-driving taxi business, just like we have done in other areas of our business. So I don't think we will have any demand challenges in the autonomous taxi business. So from an economic perspective, I think the appeal of a self-driving taxi service will be so great that we'll really be more willing to use it, to the point where demand will exceed our ability to serve it. So obviously, in this case, your only option is vertical integration. So, this is actually going through what we call the "five nines reliability journey." In this process, what you need to consider is, what level of reliability do you need to achieve in order to achieve scale? Ideally, you want to achieve 99.999% reliability. So, this is the “five nines” journey of reliability. I think this is the only real constraint on our development of self-driving taxis. Colin Langan, Wells Fargo Analyst: Thanks for the opportunity to ask. There are a lot of rumors in the market that Tesla and SpaceX will merge. At present, the two parties have in-depth collaboration in many fields. I would like to ask Elon whether there is synergistic value in the merger between the two companies in the long run, and is the merger feasible? Musk: I mean, you can see from our many collaborations with SpaceX in multiple areas, the scope of our cooperation is getting wider and wider, especially with Terafab, which is really going to be a huge project. But obviously, we can't discuss things like corporate mergers on earnings calls. This must be done through appropriate procedures. On that note, I'm going to turn the microphone over to Brandon, our general counsel.
Brandon Earhart: Thanks Elon. We continue to benefit from working with SpaceX, and they have been a great partner and we have had many beneficial transactions with them. Earlier this year, we deepened our partnership with an investment and a framework agreement. This will allow us to continue working with them on the projects Elon mentioned, such as Terafab and Digital Optimus. Musk: Yes. And there are many other things. Apparently, you have a Grok in your car. So - Grok is helping pilot Digital Optimus. You also have Starlink integrated into Cybercab, and Starlink will be integrated into all of our vehicles, at least in markets where Starlink is active. Because in the case of self-driving taxis, you need ubiquitous coverage. Even in Silicon Valley, surprisingly for Silicon Valley, there are a lot of places where cellular coverage is terrible or sometimes nonexistent. But I know that when I'm driving to work for the first 10 to 15 minutes, I can't actually make a phone call because the cellular connection is so bad. So we can’t have self-driving taxis stuck in areas like the Bermuda Triangle that lack cellular connectivity. So, Starlink with the ability to provide connectivity anywhere is actually very important. This way, our self-driving taxis won’t go missing while on a mission. Obviously, if people are sitting in a car, they want to do something productive or recreational. With Starlink, you can watch live sports events in 4K in your car at a very low cost per gigabyte of data that is simply not feasible with cellular systems. In addition, there are many other situations. Vaibhav Tania: Add to the Cybercab product logic. Like Ashok mentioned about Cybercab, the experience is really great. In a world where you don’t have to focus on driving, you can have a conference call, watch a movie, or anything else in your car at any time. That's why when you look at the design of Cybercab, you'll notice it has a big screen. We've started producing Cybercab at our facility in Austin, and customers will be able to experience it in the near future, and you'll understand why this connection is so important. Wells Fargo analyst Colin Langan: The second question is about the pace of driverless taxi deployment. At present, Tesla has chosen to expand to multiple cities simultaneously, but only dozens of vehicles will be launched in each city instead of hundreds. Why not first develop a single city in Austin, launch a large fleet, and then expand outward? What are the core obstacles to launching large-scale bicycle cities? Ashok Eluswamy: The reason why we are expanding across different cities, rather than doubling down on just one city, is because we want to ensure that our portfolio is broadly versatile. It does have broad versatility. We just wanted to prove to ourselves and others that it could work in a lot of different cities without putting too much effort into each city. This is what we see internally. And then, as Elon mentioned before, the growth rate is actually exponential. It’s just that it’s still in the early stages of exponential growth. That's why it's so hard for others to understand. In terms of mileage and number of vehicles, since the vehicles in the self-driving taxi fleet are basically continuously operating, and human drivers may only use the vehicles for a few hours or so every day, most of these vehicles are in continuous operation, which means that even with only a few vehicles, a lot of miles can be driven. That's why we track unsupervised miles driven, not just vehicle counts or other similar metrics. We want our fleet to operate efficiently. As a result, all efficiency efforts result in more miles per vehicle in the fleet.
Vaibhav Tania: Yes. I would add further, like - think about it, we're trying to scale. We're trying to make sure there are different bottlenecks. This is not only reflected in the software, but also in terms of operations, we are also working hard to solve some problems. So we're trying to expand our footprint to make sure we can address these issues in a controlled way on smaller fleets before we really deploy them at scale. Musk: Yes. One thing to point out about Cybercab is that since it is a new vehicle chassis, we need to accumulate Cybercab-specific driving data before it can be put into mass use. So unlike Model 3, Model Y, and our other vehicles, those are already on the road in large numbers, millions of them. Cybercab has not reached this scale yet. Therefore, we actually have to accumulate Cybercab mileage by modifying the steering wheel, brakes, acceleration brake and other components to calibrate the Cybercab chassis. So you're going to see -- we're very confident about it, there's going to be a significant increase in the number of Cybercabs in cities. Lars Morave: So, I would like to add a little thought here. Another reason I want to point out is that traffic control (TCO) situations vary from city to city and from state to state, and there is no federal framework to govern them. So the reason why we're expanding city by city is to make sure that we can meet all of those requirements one by one and meet the various needs of each city to the best of our ability. We must expand to every city to meet these requirements, and we will continue to do so. LightShed Partners analyst Walt Piecyk: I would like to ask about the regulatory topic Lars mentioned. The head of NHTSA recently appeared on CNBC and stated that he planned to eliminate the mandatory steering wheel regulations in vehicles. Will the elimination of the steering wheel become a core sticking point for Cybercab’s large-scale launch? In addition to steering wheel regulations, what policy gaps are there at the federal level that will limit the expansion of production of private cars without steering wheels? Lars Morave: The short answer is "no". I mean, I think we have a great relationship with NHTSA and especially Commissioner Morrison. I think they're just going with the public's expectations because the world knows what's coming and they're trying to get ahead of it and make sure they're taking the right steps. So we've been honest with them over the past few years about what our plans are and what we're doing. I don’t want to say we’re in lockstep, but I feel like we have a partner and we’re working together. Walter Pisik: Got it. And then, Elon, I want to ask a question about the road map. Obviously, Starlink has been integrated into Cybercab, and you just talked about the connectivity for watching movies, but what about the other side? Can these online taxis become remote hotspots for Starlink Mobile? Likewise, regarding Semi, I want to ask a question about the roadmap, when do you expect to consider Semi's self-driving capabilities? Because the market seems to be quite large as far as self-driving trucks are concerned, and the demand for self-driving trucks is huge considering the current situation in our country. Musk: Yes, indeed. As you mentioned, there is a severe shortage of truck drivers. As a result, not enough people are willing to drive trucks, which are.
All in all, robotaxi growth has been exponential so far while maintaining an impeccable safety record. Our technology investments are paying off, helping us sustain this exponential growth for the foreseeable future, and Cybercab is going to be great. About Optimus Figure , especially the Optimus 3rd generation model, which was designed to match not only human appearance, but also human functionality and dexterity. One of the main reasons for this approach is that having human form and function allows us to learn from humans how to perform a wide variety of tasks. This provides the entire world of data for training Optimus. Just like the fire department, we have access to a large group of people who provide us with data on all the workers in the factory. Optimus could learn a lot from watching them, performing their tasks in the factory. We also have a dedicated data collection team who can provide relatively small but very high-quality demonstrations for post-training of these models. This is all on top of any training we can do from data across the internet, which shows demonstrations of people performing various tasks in both industrial and domestic settings. The second phase is that we have a large number of Optimus robots practice tasks in what we call the "Optimus Academy." Data from the robots themselves performing tasks will be invaluable and help us close any small form factor gaps that may exist between robots and humans. This is also when the reinforcement learning cycle begins, with the robot initially trying some tasks, sometimes failing, learning from its successes and failures on these tasks, and eventually learning to master these tasks at a superhuman level. Our AI strategy for Optimus is consistent with this. This is the same end-to-end strategy that drives FSD, pixel input, and control output. Like the Fire Service, we want it to be widely available. In FSD, you get in your car, enter your address, hit start, and then it's just like you do with all your driving from park to park. The same goes for Optimus. You just let it do whatever it takes, and then you can complete the entire task yourself without having to do anything along the way. It is this team that developed the FSD, v12, 13, 14 and now v15 software that now also runs on Optimus. I'm very confident that Optimus will be a very capable robot. Analyst Q&A session: Morgan Stanley analyst Andrew Percoco: Thanks for the opportunity to ask questions, and the first question is around the Optimus humanoid robot. Elon, you mentioned before that the complete supply chain for Optimus is almost non-existent at present, and most of the components need to be developed and produced by Tesla. I would like to ask, in the process of promoting self-research and production, are there any external suppliers willing to invest simultaneously and build factories in the United States to help Tesla expand production faster? We understand that Tesla will control core technologies, but from the perspective of investment efficiency, supply chain partners share production capacity construction and can optimize capital expenditures. I would like to know your overall thoughts on cooperation with external suppliers, as well as the progress of communication with suppliers at this stage. Musk: Overall, supplier cooperation is very smooth, and major manufacturers continue to invest huge amounts of money to support the implementation of Optimus, Robotaxi and other projects. Among them, Samsung and TSMC are particularly prominent. TSMC's Arizona factory and Samsung's Texas factory have invested tens of billions of dollars to specialize in the production of digital AI computing chips suitable for humanoid robots and driverless taxis. Amazon has also invested billions of dollars in expanding its battery production lines. Although I feel a little unwell today and my voice is not in good condition, I still want to thank all partners for their financial investment and implementation support.
Kahn Budiraj: Just to add to what Elon said. The production capacity of Samsung's new wafer fab will be heavily tilted towards Tesla's future projects, with an investment scale of tens of billions of dollars. New robot-specific industry chains such as memory chip manufacturers, new metal injection molded parts, and flexible printed circuit boards are all simultaneously increasing their investment, which is completely different from the traditional vehicle supply chain. If it cannot find a high-quality supplier that matches its needs, Tesla will not hesitate to develop and produce it itself. We have top-notch manufacturing engineering and product design teams, and are fully capable of independently expanding our parts production capacity. Musk: Also thank you to Micron for reserving memory chip production capacity for Tesla. The supply of the current memory chip market is extremely tight. Micron has made a large amount of internal production capacity deployment and provided Tesla with a large long-term production capacity quota at a reasonable price. We are very grateful. Morgan Stanley analyst Andrew Perko: Thanks for the addition, my second question is related to driverless taxis. At present, regulatory policies regarding the mandatory requirements for autonomous driving sensors in various states in the United States continue to change. I would like to ask Tesla what regulatory policy demands it has when launching Robotaxi? Moderate regulation is conducive to the development of the industry, but excessively rigid regulations will restrict innovation. How does Tesla want the regulatory framework to be reasonably formulated? Kahn Budiraj: First let’s talk about regulation at the federal level in the United States. Our progress has been very smooth. The U.S. National Highway Traffic Safety Administration (NHTSA) has issued a new version of the motor vehicle safety standard FMVSS and recognized the purely visual autonomous driving solution. We are very grateful to NHTSA for its support. State-level regulations are clearly divided, and the stringent policies introduced in New Jersey are disappointing. As Ahsok said before, our core reliance is on the actual safety performance of vehicles to convince regulators and the public. The optimal supervision model is to set safety goals and leave it to enterprises to independently develop and implement paths. If supervision directly and rigidly specifies hardware solutions without focusing on safety results, we do not agree with it. Tesla has always used actual operational data to prove product safety, which is also the core support for our subsequent cross-regional expansion. Bank of America analyst Alexander Perry: Thanks for the opportunity to ask, I have two questions about driverless taxis. First, what key milestones do investors need to track that will drive Tesla to accelerate its fleet rollout and expand into new cities? Is it safety data (accident rate per million miles), or some other metric? Secondly, will Tesla reach third-party cooperation with online ride-hailing platforms to improve vehicle utilization? Or should we insist on vertical self-research and self-operated Robotaxi business in the entire industry chain for a long time? Musk: Yes. We want to be vertically integrated in our self-driving taxi business, just like we have done in other areas of our business. So I don't think we will have any demand challenges in the autonomous taxi business. So from an economic perspective, I think the appeal of a self-driving taxi service will be so great that we'll really be more willing to use it, to the point where demand will exceed our ability to serve it. So obviously, in this case, your only option is vertical integration. So, this is actually going through what we call the "five nines reliability journey." In this process, what you need to consider is, what level of reliability do you need to achieve in order to achieve scale? Ideally, you want to achieve 99.999% reliability. So, this is the “five nines” journey of reliability. I think this is the only real constraint on our development of self-driving taxis. Colin Langan, Wells Fargo Analyst: Thanks for the opportunity to ask. There are a lot of rumors in the market that Tesla and SpaceX will merge. At present, the two parties have in-depth collaboration in many fields. I would like to ask Elon whether there is synergistic value in the merger between the two companies in the long run, and is the merger feasible? Musk: I mean, you can see from our many collaborations with SpaceX in multiple areas, the scope of our cooperation is getting wider and wider, especially with Terafab, which is really going to be a huge project. But obviously, we can't discuss things like corporate mergers on earnings calls. This must be done through appropriate procedures. On that note, I'm going to turn the microphone over to Brandon, our general counsel.
Brandon Earhart: Thanks Elon. We continue to benefit from working with SpaceX, and they have been a great partner and we have had many beneficial transactions with them. Earlier this year, we deepened our partnership with an investment and a framework agreement. This will allow us to continue working with them on the projects Elon mentioned, such as Terafab and Digital Optimus. Musk: Yes. And there are many other things. Apparently, you have a Grok in your car. So - Grok is helping pilot Digital Optimus. You also have Starlink integrated into Cybercab, and Starlink will be integrated into all of our vehicles, at least in markets where Starlink is active. Because in the case of self-driving taxis, you need ubiquitous coverage. Even in Silicon Valley, surprisingly for Silicon Valley, there are a lot of places where cellular coverage is terrible or sometimes nonexistent. But I know that when I'm driving to work for the first 10 to 15 minutes, I can't actually make a phone call because the cellular connection is so bad. So we can’t have self-driving taxis stuck in areas like the Bermuda Triangle that lack cellular connectivity. So, Starlink with the ability to provide connectivity anywhere is actually very important. This way, our self-driving taxis won’t go missing while on a mission. Obviously, if people are sitting in a car, they want to do something productive or recreational. With Starlink, you can watch live sports events in 4K in your car at a very low cost per gigabyte of data that is simply not feasible with cellular systems. In addition, there are many other situations. Vaibhav Tania: Add to the Cybercab product logic. Like Ashok mentioned about Cybercab, the experience is really great. In a world where you don’t have to focus on driving, you can have a conference call, watch a movie, or anything else in your car at any time. That's why when you look at the design of Cybercab, you'll notice it has a big screen. We've started producing Cybercab at our facility in Austin, and customers will be able to experience it in the near future, and you'll understand why this connection is so important. Wells Fargo analyst Colin Langan: The second question is about the pace of driverless taxi deployment. At present, Tesla has chosen to expand to multiple cities simultaneously, but only dozens of vehicles will be launched in each city instead of hundreds. Why not first develop a single city in Austin, launch a large fleet, and then expand outward? What are the core obstacles to launching large-scale bicycle cities? Ashok Eluswamy: The reason why we are expanding across different cities, rather than doubling down on just one city, is because we want to ensure that our portfolio is broadly versatile. It does have broad versatility. We just wanted to prove to ourselves and others that it could work in a lot of different cities without putting too much effort into each city. This is what we see internally. And then, as Elon mentioned before, the growth rate is actually exponential. It’s just that it’s still in the early stages of exponential growth. That's why it's so hard for others to understand. In terms of mileage and number of vehicles, since the vehicles in the self-driving taxi fleet are basically continuously operating, and human drivers may only use the vehicles for a few hours or so every day, most of these vehicles are in continuous operation, which means that even with only a few vehicles, a lot of miles can be driven. That's why we track unsupervised miles driven, not just vehicle counts or other similar metrics. We want our fleet to operate efficiently. As a result, all efficiency efforts result in more miles per vehicle in the fleet.
Vaibhav Tania: Yes. I would add further, like - think about it, we're trying to scale. We're trying to make sure there are different bottlenecks. This is not only reflected in the software, but also in terms of operations, we are also working hard to solve some problems. So we're trying to expand our footprint to make sure we can address these issues in a controlled way on smaller fleets before we really deploy them at scale. Musk: Yes. One thing to point out about Cybercab is that since it is a new vehicle chassis, we need to accumulate Cybercab-specific driving data before it can be put into mass use. So unlike Model 3, Model Y, and our other vehicles, those are already on the road in large numbers, millions of them. Cybercab has not reached this scale yet. Therefore, we actually have to accumulate Cybercab mileage by modifying the steering wheel, brakes, acceleration brake and other components to calibrate the Cybercab chassis. So you're going to see -- we're very confident about it, there's going to be a significant increase in the number of Cybercabs in cities. Lars Morave: So, I would like to add a little thought here. Another reason I want to point out is that traffic control (TCO) situations vary from city to city and from state to state, and there is no federal framework to govern them. So the reason why we're expanding city by city is to make sure that we can meet all of those requirements one by one and meet the various needs of each city to the best of our ability. We must expand to every city to meet these requirements, and we will continue to do so. LightShed Partners analyst Walt Piecyk: I would like to ask about the regulatory topic Lars mentioned. The head of NHTSA recently appeared on CNBC and stated that he planned to eliminate the mandatory steering wheel regulations in vehicles. Will the elimination of the steering wheel become a core sticking point for Cybercab’s large-scale launch? In addition to steering wheel regulations, what policy gaps are there at the federal level that will limit the expansion of production of private cars without steering wheels? Lars Morave: The short answer is "no". I mean, I think we have a great relationship with NHTSA and especially Commissioner Morrison. I think they're just going with the public's expectations because the world knows what's coming and they're trying to get ahead of it and make sure they're taking the right steps. So we've been honest with them over the past few years about what our plans are and what we're doing. I don’t want to say we’re in lockstep, but I feel like we have a partner and we’re working together. Walter Pisik: Got it. And then, Elon, I want to ask a question about the road map. Obviously, Starlink has been integrated into Cybercab, and you just talked about the connectivity for watching movies, but what about the other side? Can these online taxis become remote hotspots for Starlink Mobile? Likewise, regarding Semi, I want to ask a question about the roadmap, when do you expect to consider Semi's self-driving capabilities? Because the market seems to be quite large as far as self-driving trucks are concerned, and the demand for self-driving trucks is huge considering the current situation in our country. Musk: Yes, indeed. As you mentioned, there is a severe shortage of truck drivers. As a result, not enough people are willing to drive trucks, which are.