Tesla's Q1 2026 Earnings Call

[00:00:00] Unfortunately, Hardware 3, I wish it were otherwise, but Hardware 3 simply does not have the capability to achieve unsupervised FSD.

[00:00:12] Good afternoon everyone and welcome to Tesla's first quarter, that 2026 Q&A webcast.

[00:00:40] My name is Travis Axelrod, Head of Investor Relations, and I'm joined today by Elon Musk, Vebov Taneja, and a number of other executives. Our Q1 results were announced at about 3 p.m. Central Time in the update deck we published at the same link as this webcast. During this call, we will discuss our business outlook and make forward-looking statements. These comments are based on our predictions and expectations as of today.

[00:01:05] Actual events or results could differ materially due to a number of risks and uncertainties, including those mentioned in our most recent filings with the SEC.

[00:01:34] Hello everyone and welcome to Kilowatt, a podcast about electric vehicles, renewable energy, autonomous driving, and much, much more. My name is Bodhi and I am your host and today we've got the much, much more. Today we're going to talk Tesla's Q1 2026 earnings call. We have interesting details as you heard. Elon admitted that Hardware 3 is not going to work. I think, you know, he was kind of leaning that way for a while.

[00:02:02] I don't know that this is a big revelation like people are making it out to be. But we'll get into that when we get to that part of the earnings call. I have listened to this earnings call all the way through. So I think I have a pretty good idea of how it went. And in my mind, it was a very kind of subdued earnings call.

[00:02:27] There wasn't a lot in this that was like light your hair on fire news. There was a lot of like, there's a lot of like, this is the best product ever type stuff. But I don't know. I'm going to try to, I'll include some of that, but I'm going to try to, you know, just include the nutritious part of that.

[00:02:52] Another thing is Vivhev, who is the CFO, his CFO opening remarks tend to sound like commercials. So while the first part of his opening remarks were commercial like, I'm not sure I'm going to use any of the second part of his opening remarks. But we'll see. We'll see. When I listen to it again, maybe I'll change my mind on that.

[00:03:20] But let's go ahead and start off with Elon's opening remarks. I think we've got a very exciting year ahead of us with 2026. We're going to be substantially increasing our investments in the future. So should expect to see significant, a very significant increase in capital expenditures. But I think well justified for a substantially increased future revenue stream. And obviously Tesla is not alone in this.

[00:03:49] I think you've seen in most, if not all, certainly the major technology companies substantially increasing their capital investments. And we're going to be doing the same. I think it's going to pay off in a very big way.

[00:04:08] So we're investing in and improving our core technologies, battery powertrain, AI software, AI training, chip design, manufacturing, you know, laying the groundwork for significantly increased manufacturing production. We're also strengthening our supply chain across the board, batteries, AI, AI, silicon, everything.

[00:04:38] And laying the groundwork, like I said, for what we expect to be a significant increase in vehicle production in the future. And of course, a very significant increase. Actually releasing Optimus, but increasing our internal production for testing. And then probably being able to have Optimus be useful outside of Tesla sometime next year.

[00:05:06] As you've heard me say a few times, I think Optimus will be our biggest product, not just Tesla's biggest product ever, but probably the biggest product ever. And I remain convinced of that conclusion.

[00:05:26] So on the RR vehicle side, it's always, I think, worth noting that a Tesla car is incredibly, incredible value for money and they're all autonomy ready, depending on what part of the world you're in. And the supervised full self-driving is getting extremely good. We have just started production of CyberCab and will begin production of our semi truck soon.

[00:05:57] Now I should say, whenever you have a new product with a completely new supply chain, new everything, it's always a stretched out S-curve. So you should expect that initial production of CyberCab and semi will be very slow, but then ramping up and going kind of exponential towards the end of the year and certainly next year.

[00:06:23] And in fact, we'll be ramping up production of all vehicles and all factories to the best of our ability through the balance of this year. On the energy front, the United States and the whole world will need a lot of energy storage to meet growing electricity demand. Demand for our Megapack is very strong. And we're excited to begin production of Megapack 3 later this year in our new world class factory outside Houston.

[00:06:53] For full self-driving and robotaxi, version 14.3 was a major architectural update. And we have a whole pipeline of major improvements to full self-driving that we believe will lead to unsupervised full self-driving being available anywhere in the world that it is legal to do so.

[00:07:24] And then there's a version 15, hopefully by the end of this year, but certainly by early next year. We'll be able to run on AI4. And that's a complete overhaul of the software architecture and we'll run on AI4.

[00:07:44] And at that point, we're really just increasing the safety level of FSD above human safety level even more. Meaning, I think even within version 14, we're significantly safer than human, but B15 will take that to another level. Okay.

[00:08:12] Lots of investments in 2026. And Vivek talks about this quite a bit. So like I said, maybe I'll include some of that, but significant capital expenditures. You know what? I mean, Tesla's working on a lot of stuff that makes sense. You know, powertrain, AI, chip design. I'm guessing that's where most of the money is going.

[00:08:40] You know, into their big fab that they're building. So yeah, I mean the silicon and the whole bit, you know, there's a lot going into that, what they're doing. Optimus out this year, be useful outside of Tesla. Tesla sometime next year. I mean, maybe.

[00:09:01] I, you know, he said that Optimus was going to be Tesla's biggest product ever, which, you know, he's also said that about Tesla AI. He said that about Tesla energy. He has said that about a lot of different products that Tesla has produced. A Cybertruck was going to be a huge hit. And it's not to say that he's lying or anything like that.

[00:09:28] It's just, you know, will companies want to purchase a robot from Tesla knowing that the information that's fed into that robot on the manufacturing line is going back to Tesla, a company that manufactures a wide variety of products. I don't know. I mean, that's a decision that those companies have to make.

[00:09:52] Like, is it worth saving the few dollars that you spend on people to potentially unwittingly send data back to the company that you're buying the robots from who might at some point in time compete with you? Now, if they're building, you know, toy trucks, I don't think Tesla is going to be competing with you on that level. But I'm just saying, like, there are some things out there that are kind of sensitive.

[00:10:21] So we'll see where this goes. And based on how Tesla has handled their data, I mean, when Optimus first came out, I was all for this. There's recent revelations about how they've handled their data and, you know, spreading videos around from the FSD cameras internally between teams. I wouldn't not have one of these. I don't have cameras in my house.

[00:10:47] I wouldn't have an Optimus in my house based on that. Now, is that unique to Tesla? No, that could happen anywhere. So I would have to really rethink one, what I'd have to, I would like somebody to do my vacuuming. Maybe when I'm like 80 years old and I need somebody to help me with different things, that will be an easier sell for me.

[00:11:14] But right now, like I have young kids and I have a wife and stuff like that. I don't want videos of them being passed around Tesla headquarters or any headquarters for that matter. I do agree that Tesla is a huge value for the money. He talked about every car being autonomy ready. Maybe. Maybe AI4. We'll talk about that with Hardware 3.

[00:11:43] But yeah, maybe it's autonomy ready. I don't know. It doesn't, I'm not putting a lot of, I don't put a lot of stock in that statement because I've heard it twice before with Hardware 1 or Hardware 2 and Hardware 3. Megapack version 3 is coming out. That's great. Version 15. I've kind of already talked about Tesla's versions for the FSD. You know, 14.3 comes out. That's supposed to make this huge improvement.

[00:12:11] And 15 is now the new target of what's going to be amazing and fix all the problems. I just think we're, let's appreciate what we have right now and not worry about what's coming down the line. Again, and I've said this plenty of times. Let's go ahead and jump into our next Elon clip. We've expanded RoboTaxi to Dallas and Houston using the same software as Austin and the Bay Area.

[00:12:41] And the limiting factor for expansion is really rigorous validation, making sure things are completely safe. We don't want to have a single accident or injury with the expansion of RovoTaxi. And we have, to the credit of the team, not had a single one to date. Now, we know that they've had at least 18 accidents and a few of those have been injuries. So, I don't know quite what he means here.

[00:13:11] He is, he's on an earnings call. So, he cannot lie without having some sort of legal repercussion. But they have had accidents with injuries. And this is something that somebody else on the team repeats as well. So, I'm a little confused. I'm not saying they're lying. I'm saying I'll have to do a little bit more research and see what exactly they're talking about. Because if they're saying they haven't had any accidents or any injuries, that's simply untrue.

[00:13:40] So, we'll keep an eye on this and see where it goes. Let's continue on with the call. Optimus, we're preparing Fremont for starter production later this year with Optimus. It's, again, totally new supply chain, totally new technology. So, therefore, the production S-curve is always very slow in the beginning. But we'll ramp up to significant numbers next year.

[00:14:05] And we're constructing a second Optimus factory at our Giga Texas location. And that will probably start production around summer next year. The V3 Optimus design is almost ready to demonstrate. I think we want to just make sure it's, like, polished. Like, it works functionally, but there's some aesthetic elements that need to be finalized.

[00:14:37] And I think probably middle of this year we should be able to show it off. We're also a little hesitant to show V3 off because we find our competitors do a frame-by-frame analysis whenever we release something and copy everything they possibly can. So, I think there's some value to, you know, not showing new technology until it's close to production.

[00:15:04] I think there's value in not talking about technology until it's nearly fully baked. Like, one of the cool things is that Tesla has this very, or used to have this very transparent way of doing things because Elon was just unable to not talk about stuff. And he's gotten more disciplined over the years. But early on, it was just like, he would just talk about whatever.

[00:15:26] And while that was fun, you know, now we're kind of at a point where talking about, you know, future products or talking about current products and what they're going to be in what he calls, you know, next year, next month, next whatever. But they don't end up that way. And I feel like I'm just beating this to death. But yeah, don't talk about it.

[00:15:55] On the other side, when he says that people, competitors are taking frame-by-frame analysis of optimists, you're only seeing the outside of the optimists. You're not seeing the insides. You're not seeing the programming. You're not seeing the hard parts. You're seeing the outside. Like, I would imagine that robotics engineers probably already have a good idea of what Tesla's doing.

[00:16:22] One, because they're robotics engineers and they're not dumb. But two, they're probably working on similar things. And they're probably in similar places. And yes, we don't want to give a bunch of information to our competitors. But on the other side, showing off Optimus 3 walk from one place to another,

[00:16:46] you don't think that's enough to tilt the scale one way or the other when it comes to a company that's working on similar products. Let's say there's a company that's 80% along on this one problem. And they frame-by-frame one of Tesla's videos, one of their Optimus robots hanging out serving drinks to people. They frame-by-frame that video. And they were like, oh, we're doing it all wrong.

[00:17:15] We're going to scrap 80% of the work we've done. No, they might rethink something if they kind of get an idea. But really, I don't think any of this stuff is something they haven't already thought of. It's like Tesla has their way of doing things and these other companies have their way of doing things. I just don't know. Like, unless they're starting from scratch, I don't know what the value is in completely ripping up what you've done because you saw three frames of the Optimus robot.

[00:17:45] Now, if you've got a hold of an Optimus robot and you are able to completely rip it apart, yeah, I totally get that. But I don't get the, if we show it off, they're going to steal from us. I don't think there's enough information there to steal that they don't already know. All right, let's continue on with the earnings call. Congratulations again to the Tesla AI chip team for taping out AI-5. That's going to be a great chip.

[00:18:15] I think probably the best AI inference chip for edge compute that exists. And certainly, I think unequivocally the best value for money. Team did a great job. And we already have a lot of momentum for designing AI-6. And we've begun to discuss ideas for Dojo 3. So this is all very exciting.

[00:18:46] We've also finalized plans for the chip fab, the research chip fab on the Giga Texas campus. And we'll start construction of that this year. In conclusion, Tesla is working on a lot of large, ambitious projects. They're all very challenging, but I think they're going to be revolutionary.

[00:19:15] And this is what the team does best. Solve the hardest problems and build amazing products. And I'd like to thank the Tesla team for all the hard work. And thank you to all of our supporters. All right. I don't have any more to add. I mean, I'm sure you could imagine what I'm thinking with that. But that's a nice update. And I'll just leave that as it is. Next up, we're either going to get a few clips from Vibhev's opening remarks. I'll re-listen to it again.

[00:19:40] And if I still think there's nothing worth talking about or listening to, then I'll go ahead and move on to our retail investor questions. So let's go ahead and listen to our next clip. On the FSD adoption front, we continue to see improvement, reaching nearly 1.3 million paid customers globally.

[00:20:05] The bulk of the growth came from subscriptions, while upfront purchases only increased 7% as we removed the purchase option in some markets in Q1. We recently received approvals for FSD in Netherlands. This sets up us well for an EU-wide approval later in Q2. And, you know, we're just gated by how the regulators go about it. Additionally, we've also received approvals in China.

[00:20:36] The broader approval is still not there, but we're working with the regulators in the country. And we're hoping that we can get approval by Q3. With these approvals coming through, we expect the broader adoption of the software in the existing fleet and incremental demand for our vehicles. With all this in mind, we have evolved our vehicle sales strategy, where we now emphasize FSD as a product and vehicle as only the delivery mechanism.

[00:21:06] As we have noted previously, the energy storage business is inherently lumpy, tied to customer deployment timelines. In Q1, we deployed 6.8 gigawatt hour of energy storage, a 38% sequential decline. However, we still expect 2026 deployments to be higher than 2025.

[00:21:29] We set yet another record with gross margins in this business over 39.5% due to some one-time benefits from certain tariff recognitions of more than $250 million from certain tariffs which you had paid in prior quarters. On a normalized basis, we continue to expect energy compression from here with increasing competition and tariff impacts.

[00:21:53] As previously discussed, tariffs in this business can have outsized impacts as most of the battery cells are procured from China. All right. So he goes on a little bit more to talk about tariffs and all that stuff. They haven't, as far as I know, they haven't applied. Sounds like they haven't applied to get their refunds on tariffs yet.

[00:22:17] When it comes to 1.3 million paid FSD customers worldwide, that obviously includes China and Europe and all that stuff. I would like to know how many subscribers they gained in Q1 of 2026. And the paid customers, you know, that wasn't very many. 7% is what he said.

[00:22:46] So that's, you know, I would like to know how many they picked up, how many subscribers they picked up in Q1 2023 and what their churn is, like when people decide to stop using it. But I mean, I didn't think we'd be getting these numbers, so that's a positive, but I doubt we'll get the churn anytime soon. Tesla, we talked about when Elon said that they're going to do investments, big CapEx investments.

[00:23:16] Vivev said that that's going to be $25 billion in 2026. That is not so, so big investments in 2026. All right, let's move on to our retail investor questions. Now we're going to go to investor questions, starting with a question from say.com. The first question is, when will we have the Optimist 3 reveal, which we already touched on?

[00:23:45] But the rest of the question is, when will Optimist production start since we ended the Model X and S production earlier this mid-year? And then what's the expected Optimist production rate exiting this year? And what are the initial targeted skills? Well, as I was saying, what we found is that when we've unveiled various Optimist versions, we found out how competitors literally do a frame-by-frame analysis and copy everything we're doing.

[00:24:13] So I think we want to push the Optimist 3 unveil maybe closer to production. Startup production is, we're assuming, is somewhere around the late July, August timeframe.

[00:24:36] And, I mean, just to inject some reality into these questions, since these questions are not, you know, whoever does this question does not fully understand what happens with the production line. The last SX production will be in early May. But you have to look at the entire upstream portion of the production line.

[00:25:01] So you start with, you know, cells, battery packs, you know, motor production, all the parts production. And so we've been dismantling the SX production line, you know, from the more base-level parts, the more basic-level parts to, as you get to more larger sub-assemblies, you start dismantling the line from the small parts first, not from the final assembly first.

[00:25:27] So the final assembly line will, that'll be dismantled next month and after the last of the SX vehicle's done. Now, you can't dismantle some gigantic production line, like, overnight. It takes at least a few months to do so.

[00:25:49] And then you've got to install a new production line, and you've got to provide all of the wiring and communication, you know, test out the machines of the new production line for Optimus. So that also takes several months. So, frankly, if we're able to go from stopping production on one line, dismantling that entire line, reinstalling a whole new line,

[00:26:19] and turning that on in a matter of four months, that is an insanely fast speed. I don't think any other company on Earth has ever done that before, just to put things into perspective and inject some reality into the situation here. I don't know what the production rate of Optimus will be this year. It is impossible to predict these things.

[00:26:52] When you have a brand new product and an entirely new production line, and you have 10,000 unique items, all of which have to go right to ramp production, it'll move as fast as the least lucky, slowest, dumbest part in the entire 10,000. And Optimus is a completely new product with a completely new production line.

[00:27:23] So, it's just literally impossible to predict. Except that I think it will be quite slow at first, as we iron out the 10,000 plus unique items that have to be sold for Optimus to reach volume production. Initial skills will be, obviously we're going to start with simple skills in the factory,

[00:27:52] and then build up from there. I don't think this was a very kind response to the person that asked that question. Elon said in the beginning, like, something about this person knows nothing about assembly lines, or something to that effect. He also, you know, made mention that these things are, I want to inject a little bit of reality into this. And, what was the other thing that he said?

[00:28:21] I can't predict how many, it's impossible to predict how many products we'll build this year. It's not, but also, you know, it's probably not going to be very many, and I think we can, you know, Tesla could be forgiven for it not being very many, because they are building a brand new product. But I just really feel like the way that he explained this was condescending and kind of jerky.

[00:28:52] And maybe I'm, I miss, maybe he's just trying to be really thorough. I don't know. But I didn't like the way he did it. You know, let's just inject some reality into this. Just very irritating to me. Also, and again, the prediction. He predicts all the time. All the time. He has no problem throwing numbers out that don't end up becoming true. He's going to do it this earnings call.

[00:29:20] So, for that, you know, I just don't, like, he's the one that sets the expectations. He's the one, like, he's done this before, which is why people are asking. There's where I have the problem. Don't get irritated because you created the monster. So, don't get mad when it goes into the village and wrecks all the villagers' homes, because you're the one that created the monster. All right, let's go on to our next question.

[00:29:50] The next question is, What milestones are you targeting for unsupervised FSD and RoboTaxi expansion beyond Austin this year, and how will that drive recurring revenue?

[00:30:13] Well, we certainly hope to have unsupervised FSD or slash RoboTaxi operating, you know, in, I don't know, a dozen or so states by the end of this year.

[00:30:36] Initially, you know, we're taking a very cautious approach to the rollout here. Like, we haven't had any injuries and certainly no fatalities to date with the unsupervised FSD and RoboTaxi expansion. We want to keep it that way.

[00:30:54] And so, I think probably unsupervised FSD or RoboTaxi revenue will not be super material this year, but I do think it will be material probably in a significant way next year. Okay, so we said again, no injuries.

[00:31:22] I don't know what they're qualifying as an injury because it clearly states there were minor injuries. Now, I don't think if they're minor injuries, we should be making that big of a deal about it. But if you're saying no injuries, an injury could be something like a stiff neck when you get into an accident, or it could be, you know, a little tissue damage when the seatbelt yanks against your chest or around your abdomen.

[00:31:52] I would really like to know what they qualify as an injury because I think that might clear it up for me a little better. And what they qualify as an accident. I would say anything that you had to report to the National Highway Traffic Safety Administration, I would qualify that as an accident or an injury. But, you know, who am I? All right, let's move on to the next question, which is about unsupervised full self-driving. And when will normal people get it in their cars?

[00:32:22] The next question is, when do you expect FSD unsupervised to reach customer cars? I'm just guessing here, but probably in the fourth quarter.

[00:32:45] It's difficult to release this to everyone everywhere all at once because we do want to make sure that they're not unique situations in a city that, you know, particularly complex intersection or, you know, actually they tend to be places where people get into accidents a lot because they're just, you know, perhaps there's, like I said, an unsafe intersection or bad road markings or, you know, a lot of weather challenges.

[00:33:13] So, so I think we would release unsupervised gradually to the customer fleet, but, you know, yeah, as we feel like a particular geography is confirmed to be safe. Okay. So, yes, where people tend to get into accidents is where the robo taxis get into accidents. There he says that the cars get into accidents.

[00:33:42] This, this is where, I don't know, this is where my problem is. I really need to know what exactly he means by no injuries, no accidents, but yet there's accidents. Anyway, that does make sense because that's where Waymos tend to get into accidents the most. Unsupervised FSD for regular people in the fourth quarter, I believe is highly unlikely.

[00:34:11] And where unsupervised for regular people, full self-driving, when it will come out, I would be very, I'd be willing to bet a coffee, and I can't go one to many, but I'll bet one person a coffee that it's going to get rolled out where robo taxis approved first, because that's a less of a lift

[00:34:34] for them to get that approved versus a state or area where they won't even allow robo taxi, where are they going to allow you to have full self-driving cars that are unsupervised full self-driving cars. So, yeah, that's my guess. All right, here comes the big question. Let's go ahead and listen to that. The next question is, how will hardware three cars reach unsupervised FSD?

[00:35:02] Unfortunately, hardware three, I wish it were otherwise, but hardware three simply does not have the capability to achieve unsupervised FSD. You know, we did think at one point it would have that, but relative to hardware four, it has only one-eighth of the memory bandwidth of hardware four.

[00:35:26] And memory bandwidth is one of the key elements needed for unsupervised FSD. It's just generally a thing that's needed for AI. If you're doing an auto-aggressive transformer memory bandwidth is the true point. So, for customers that have bought FSD, what we're offering is essentially a trade-in,

[00:35:55] like a discounted trade-in for cars that have AI4 hardware. And we'll also be offering the ability to upgrade the car to replace the computer. And you also need to replace the cameras, unfortunately, to go to hardware four. So, to do this efficiently, we're going to have to set up,

[00:36:22] like, kind of micro-factories or small factories in major metropolitan areas in order to do it efficiently. Because if it's done just at the service center, it is extremely slow to do so and inefficient. So, we basically need, like, many production lines to make the change.

[00:36:47] And I do think over time, it's going to make sense for us to convert all hardware three cars to hardware four because that's what enables them to enter the robot taxi fleet and have unsupervised FSD. And for what it's worth, in the meantime, we're going to also release a V14 version for hardware three. This will be a distilled version of the same V14 software that we released for hardware four.

[00:37:15] And people should be able to start the drives from Park State and basically have all the features that V14 for hardware four has. And that's expected to come end of June. Maybe it'll have all the features that V14 hardware four has. I mean, maybe V14 without the .3 or the .2. I don't know. That seems... We'll see. We'll see if it comes out in June.

[00:37:45] For folks who have hardware three vehicles, how does that make you feel? Are you... I mean, obviously, it would be great if you could take your car, especially if you purchased it in, like, 2019. You could take your car and have it... The hardware updated for free. I guess I should say people with FSD hardware three vehicles. Like, if you paid for FSD, how does it make you feel? I mean, that sounds cool.

[00:38:15] I really think Tesla's going to look at that. And they're going to have to see, because this is not going to be a cheap fix for them. Is it make more sense to just give people their money back and say, see you later, or try to get them into a newer car, than it does to replace the hardware three vehicles with hardware four, you know, hardware with the actual components. Because it...

[00:38:44] I mean, if you have to set up your own factory, if you can't do this in a... If you can't do this at a service center, because it'll be too slow, that seems like it's going to cost a little bit of money. To get it all set up and stuff. So, yeah, that seems like it's going to be harder than... What some people might expect. Harder than what I thought, because I thought they were just going to replace the computer. I didn't think they were going to replace the cameras. So, it remains to be seen,

[00:39:13] once they crunch all the numbers out and actually get ready to roll out the change into hardware four, whether or not Tesla thinks the juice is worth the squeeze and they just don't give you your money back. All right, let's go ahead and move to our next question. Great, thank you very much. The next question is, what enabled you to finish the AI-5 tape out early? And were there any changes to the original vision?

[00:39:42] Last week, Elon said AI-5 will go into Optimus and the supercomputer, but one month ago said it would go into the Robotaxi. Has AI-5 been dropped from the vehicle roadmap? Well, the reason AI-5 tape out finished early was because the team worked incredibly hard to make it happen. And just over time, we gathered a lot of momentum. But we did have to work every weekend for six months straight,

[00:40:11] including every holiday. So, it was a lot of sacrifice by the team, and I was there, of course, myself every weekend. And, you know, fortunately, we didn't encounter any major... We didn't make any major mistakes, at least that we're aware of, that required pushing out the tape out. So, the team just did a great job and worked incredibly hard is the reason.

[00:40:44] Yeah, I do expect that AI-5 will go into Optimus and into the data center because it's looking like we'll be able to achieve unsupervised self-driving with AI-4 that is far greater than human safety levels.

[00:41:10] Which means it's certainly not immediately needed in the car. At some point, I think it will make sense for us to switch to AI-5 in the car, but there's not a pressing issue to do so. So, but at some point, the AI-4 hardware is going to get so old that it's like, okay, you know, the only reason they're keeping the factory open is for AI-4.

[00:41:40] We are planning an AI-4 upgrade to use a newer generation RAM. So, it'll go from 16 gigabytes to, I think, 32 gigabytes per SOC. So, a total of 64 gigabytes. And probably a 10% increase in compute and sort of trillions of operations per second

[00:42:09] and in memory bandwidth. So, that's AI-4.1 or AI-4 plus probably goes into production in the middle of next year, I think. Depends. It depends on... Samsung's doing the modifications for us, so it sort of depends on when they're able to finish that, finish those modifications and bring it to production. It was my understanding,

[00:42:38] sometime last year, Elon had kind of talked a little bit about the AI platform, like AI-5, AI-6, and what the rollout would be. And the way he made it sound like, you know, they were going to be coming out yearly and it might even be sooner than a year or sooner than 12 months in some cases. And at that time, I was like, that's weird because you don't spend billions of dollars developing something to make it obsolete in eight months.

[00:43:06] Like, you want to get as much money out of that product as you can. So, it sounds like that's what they're going to do. Putting AI-5 and Optimus, okay. I mean, that's great. I think that's cool. I would not be surprised if they don't go, you know, once AI-4 has run its course, I would not be surprised if we don't get AI-6 in the cars and we just skip AI-5 in the cars altogether. Um,

[00:43:35] I might be wrong on that, but if that's kind of how they're doing things, uh, and they think if, if this all is predicated on them actually, um, hardware for AI-4 actually being able to have as long of a life as he's saying it will have here with the, the new update to AI-4, um, the new hardware update to it. And if, if that's the case, you know, when AI-6 is ready, why wouldn't you put it in the car

[00:44:05] and just skip AI-5 altogether if it's going into Optimus? And I'm sure there's, you know, lots of things that Optimus can do and AI-5 sounds like a promising chip, but maybe that's a purpose-built chip for robots and not so much for cars. All right, uh, let's go ahead and go into our next question. The next question is, uh, given the recent NHTSA incident filings, can you update us on the robot taxi safety data?

[00:44:35] If safety validation remains the primary bottleneck, why not deploy thousands of vehicles to accelerate removal of the safety driver? Yeah, Shark, do you want to take that? Yeah, we are increasing the amount of our QA fleet, but we also want to use the customer fleet to, um, give us, uh, the useful metrics back so that we can scale it safely. Like you don't mention, we are absolutely focused on safety, uh, and so far we have zero, uh, incidents, um, and that's what the NHTSA filing also shows. Uh,

[00:45:05] in addition to, um, safety, we are also solving some of these, uh, so-called scaling issues. For example, you do not want the robot taxi to be stuck blocking intersections, uh, or don't want to be dropping people off at, uh, slightly incorrect locations and so on. So we are simultaneously solving the long tail of safety, uh, by monitoring the metrics across the entire Tesla customer vehicle fleet, which are, you know, uh, is close to driving 10 billion miles on FSD, uh, in the next, uh, few weeks, uh,

[00:45:34] and also scaling up the amount of QA fleet that we have across the entire U.S. to accelerate our safety validation, uh, while also scaling, uh, the rest of the factors that can, um, you know, throttle the, um, increase of unsupervised vehicles. Huh. I mean, again, he says there's no incidents, but there's been at least 18 incidents. I just don't, I,

[00:46:03] I don't know why this is not, I don't know why this is bothering me, but not a Tesla app, which is a fairly, um, uh, fairly, I don't want to say apologetic, but they put Tesla in a positive light in most of the articles. And by the way, I love not a Tesla app. The, not a Tesla app.com. They do great job, but you know,

[00:46:30] they report 18 accidents based on NHTSA and to, um, transports to the hospital. Like we talked about this the other day. I, I genuinely don't know what they mean when they say there's not been any incidents. And then they refer back to net the national highway chance traffic is safety administration. there are accidents. I, uh, they're minor,

[00:46:59] but man, where's, I don't know. I don't know why we're, I don't know where this information is coming from. Um, on the other side, I do think there's a little bit of cover going on with, it's like, well, we don't need more robo taxis, uh, out in the field simply because, you know, we have, uh, data from you driving your car around and we use that to kind of augment. And maybe they do use it to augment, but also you've had people

[00:47:29] driving around with FSD for a really long time now. And, you know, uh, it's still not perfect as, uh, you know, uh, as we've heard and, and it still needs work and it still needs, like, I, I just don't understand this at all. Uh, by the way, I also don't understand the, the, the person who asked the questions, like, why not just add more to the fleet? I don't know that that's the right solution either. So just FYI,

[00:47:57] I think this is a problematic question, but I also think it's a problematic answer. All right. Uh, moving on to our next question. All right. Uh, the next question is, is V, uh, 14.3 still the last piece of the puzzle to enable, uh, large scale unsupervised FSD and robot taxi, or do we have to wait until V 15? Well, um,

[00:48:27] you know, I, I, I, I think, 14.3 is, is, is last piece of the puzzle, uh, for unsupervised FSD. Now the question is like degrees of safety, um, like how safety and convenience, I suppose. Um, we, we have a lot of known improvements, um, like major architectural improvements that we know would improve the probability of safety significantly.

[00:48:57] So I think it's, it's, it's not going to make sense for us to deploy, you know, unsupervised FSD or rover taxi at large scale when we, we know that there are major architectural improvements to the software that can improve safety. Um, so, so, so I think we're, we're, we're going to want to, uh, finish writing that software, validate it and release it before, uh, going to large scale unsupervised FSD,

[00:49:27] depending on what large scale means. I mean, we, we, we, we are, of course, as I mentioned earlier, uh, doing unsupervised FSD in three cities and we'll expand on it to, like I said, probably a dozen states or more later this year. Um, so kind of depends on where your definition of large scale is. Um, but I do think it, it wouldn't, it wouldn't be right for us to go to, go to,

[00:49:55] go to like very large scale unsupervised FSD when we know that there are software improvements in the pipeline that would improve safety. Yep. And I'd like to note that the version of robot taxi that's running in Austin, uh, Dallas, et cetera, those are essentially 14.3 variants. Um, uh, and it's obviously, uh, safe that that's why we're able to launch in those cities. And we continue to expand based on the V13, for V14.3 base, um,

[00:50:25] for a while until V15 lands. And V15 is going to be a major upgrade. All right. I think I've beat that to death. I'm not going to cover it again. Let's go ahead and go to our next question. Considering the various parties involved in the TerraFab project, uh, I'm hoping you can provide, uh, some details for investors about which party is going to take responsibility for each aspect of that project, funding it, designing it, uh, building it,

[00:50:55] operating, taking production. Uh, and the like, uh, we'd love to hear some, some more details. Yeah. So we're, we're still working out the, um, the details of the TerraFab uh, deployment. Um, in, in, in the near term, uh, Tesla will be, um, building the, uh, the research fab, uh,

[00:51:24] on our gigatexas campus. Um, this is something we expect to be probably, you know, a $3 billion-ish initiative, um, um, and capable of maybe a few thousand wafers per month. Uh, but it's really intended to try out ideas, the research fab, um, both, both in terms of maybe, we have some ideas for improving the fundamental technology

[00:51:53] of how chips are made, um, and some of the, there's some new physics we'd like to test out. Um, but, but we also want to test out the ability to, um, to see if something is working in production. So you need kind of like a few thousand wafers starts a month, uh, to make sure that, a production process is sound. Um, and, uh, and then SpaceX is, uh,

[00:52:23] uh, didn't take care of like the initial phase of the, the, the scaled up terafab. Um, and, um, that, that's, that's what we figured out thus far. Um, you know, any, any kind of intercompany thing has to be approved by both the SpaceX and Tesla board of directors. It's got to go through a conflict resolution. It's, it's kind of a lot of, unfortunately, a lot of complexity because,

[00:52:53] we've got to make sure Tesla shareholders are served and SpaceX shareholders are served. And, strike the right balance there. Uh, so it, it takes a while to work through the, the, the kind of independent director, uh, reviews on, on this. So that, that's basically what we've figured out thus far is Tesla's doing the research fab, SpaceX doing the initial part of the large scale terafab. And, um, and then we,

[00:53:23] we've got to figure out the rest. Yeah. And what about Intel's involvement? Uh, yeah. So, uh, Intel is, uh, excited to partner with us on, um, the, some of the core manufacturing technologies. Um, so, uh, we, we plan to use Intel's 14a, uh, uh, process, uh, which is, uh, state of the art and in fact, not yet totally complete. Um,

[00:53:53] so, but, uh, given that by the time terafab scales up, uh, 14a will be probably fairly mature or ready for prime time. So I, I don't have a lot to add to this daily tech news show had, uh, and I'll have to look. Oh, I was on the daily tech news show today, by the way. Um, I wasn't talking about this. I was talking about C ATL, TL's new battery, but anyway, I'll,

[00:54:22] I'll try to remember to find the shows that where they talk about this. If you want more information on it, they do a better job that I can do, uh, talking about this topic. All right, let's go ahead and move on to our next one. Uh, recognize the importance of FSD and that FSD can help to drive vehicle sales. And nice to see some of the, um, improvements in the FSD technology more recently with version 14. Um, however, I'm also hoping to understand if the company, the view on new vehicle models has evolved. And I asked, given that Elon you posted on X recently,

[00:54:52] that Tesla could develop a family vehicle. There's also been some past discussion about a compact vehicle. Well, I mean, cyber cab is the compact vehicle. It's actually, I mean, it's very roomy, but it's a two, it's a two person vehicle. And we do think probably most of our production longterm will be cyber cab because 90% of miles driven are with two or one or two people. So it would mean that, you know,

[00:55:19] you'd want to vast majority of your production to be cyber cab. Um, then, um, but over time, it's going to make sense for our whole lineup to be autonomous vehicles of different sizes. And I, I did talk a bit about this, um, um, when we did the kind of AI day in LA at, uh, Warner brothers and, you know, showed like, this is our current lineup and this is what, you know, some idea of what our future lineup will be,

[00:55:48] which is that it's going to be almost entirely autonomous. In fact, longterm, the only manually driven car will be the, the new Tesla roadster. Uh, speaking of which we may be able to debut that in a month or so. Um, it requires a lot of testing, uh, and validation before we can, uh, actually have a demo and not, you know, have something go wrong with the demo. Uh,

[00:56:16] but I think it will be one of the most exciting product unveils ever. Um, unless it, I don't think it moves the needle massively from a revenue standpoint. So, but it is very cool. Um, I think, I think it might be one of the most spectacular demos ever. Okay. Let's start with the last thing first. Um, in the, we were promised a demo in December. We've been promised demos before, but we were promised to demo in December,

[00:56:47] coincidentally enough, right before people were set to vote on Elon's pay package. And then that demo got kicked down the road and then it got kicked down the road. And now it's kicked down the road again. Uh, I do not expect to see this in a, in a month. Um, I just don't, we'll see. Then on the other side of things, uh, you know, this is kind of a thing that I was talking about before with Elon, Bill, Elon created this monster.

[00:57:16] Like he says things like, you know, the, the family mover or the, the minivan type of thing. And then he says, yeah, but then, you know, everybody's going to want to be driving full self-driving and you don't even need a car. And then, you know, Tesla was the one, Tesla was the company that came out and had a whole big production about how we had this affordable, compact car that more people can afford. And now that's the cyber cab.

[00:57:45] I'm going to be honest with you. If the cyber cab only has two seats, not very many people are going to buy it. And, and the people who are going to buy it are probably going to be the people who can put it on the robo taxi. If, if that ever becomes an actual thing that they can do. And honestly, the people who buy it will be the people who think they can put it on the robo taxi. And maybe they never will be able to put it on the, on the robo taxi network. Well, that remains to be seen,

[00:58:14] but Elon says these things, people respond to them. And then this is like, well, honestly, you know, you only need to, there's only usually two people in the car. That's probably true, but there are occasions when you want to take a third person. And if you're the person that has a car and what are you going to do? I mean, I guess in that point you're going to take an Uber, but if you want to take a road trip or whatever,

[00:58:41] it doesn't cost that much extra to put an extra two seats in the back or a bench seat in the back. Like, I don't, I, I don't know. I think this is an out of touch. Um, I, I think that kind of thought process is out of touch. I'll just leave it at that. Let's move to our next question. You're, you moved the safety driver in Austin and you're now expanding into Dallas and Houston.

[00:59:04] What are the key safety metrics that you're tracking that gives you confidence that robo taxi is safe in us to expand? Is it sort of miles per intervention, miles per accident, per fatality? And, and where do you stand on that now? Sure. Yeah. We track basically all the metrics that you mentioned. Uh, we have a pretty large QA fleet, uh, spread across all of the United States. Uh, and then we, uh, you know, look at, uh,

[00:59:31] any intervention that could happen and then sort of simulate, uh, uh, both in practice and also in our simulators that are very, very good nowadays using neural networks, uh, as what would have happened. Uh, and then based on all these analysis, we in the end make the call, uh, to expand. Uh, and so far, uh, all of the expansions have gone according to our expectations. Yeah. A lot of, a lot of the limiting, a lot of what limits wider deployment of robo taxi are actually not safety issues, but,

[01:00:01] uh, convenience issues or, or the car basically gets paranoid and gets stuck. Um, like sometimes it gets, cause it's, it's programmed for maximum safety. So the problem is that then it sometimes just, uh, gets scared to do things. So like it sometimes gets scared to cross railroads, for example. Um, or it'll get stuck at, uh, you know, a, a light where there's, uh, the light,

[01:00:29] the light never changes from red or, I mean, there was one kind of amusing situation where a whole bunch of, uh, um, robo taxis got stuck in the left turn lane in Austin because I kid you not, a Waymo had crashed into a bus. Um, and so they could not turn left because the way mode crashed into the bus. And so you had this like long line of like, I don't know, it does no more. It tells the robo taxis that we're waiting for the bus to move,

[01:00:58] but the bus was never going to move because the way more is crashed into the bus. So, so that obviously drives people crazy. If there's a whole bunch of robo taxis blocking the whole road. Um, so it's, it's a ton of, ton of things like that. That's the single biggest thing is, is just the car being scared to move or getting kind of stuck in situations like that. Uh, we've also had literal infinite loops where, you know, the car might want to make a turn into a road,

[01:01:28] but there's construction and then it goes around the block, tries to turn into the road with construction, goes around the block, tries to turn the road. Um, and, and so you have to stop the infinite looping, literal infinite looping. Um, so that those actually, and that those are by far the, um, issues that we have to resolve as opposed to direct safety issues. Okay. Um, I mean,

[01:01:57] this sounds a lot like Waymo. It sounds a lot like the problems that Waymo has. And, you know, some of the things that they're talking about here, I remember, you know, members of the, the people on this call, uh, making fun of Waymo because of those, uh, similar things. And yeah, I mean, it's, um, I, I just, I just don't, I don't know. Like I,

[01:02:27] I, I, this, I don't know if it's because it's late in the day, but this earnings call is starting to irritate me. Uh, fortunately it's done though. Uh, we have had our last question, but yeah, I, I just like, if you can't, if you can't fix this, and then you're saying, well, there's nothing really stopping us except for regulation for, uh, expanding. I don't know.

[01:02:55] There's just a lot of stories being told and it's exhausting. So we're going to end it here before I say something that I shouldn't say. Uh, if you want to email me though, and you can say it, Bodie, B-O-D-I-E at 918digital.com. You can also find me on X at 918digital, but I'm never there. But if you want to go somewhere where I actually am, it is, uh, LinkedIn, which is Bodie Grimm,

[01:03:24] B-O-D-I-E G-R-I-M-M. And when I started this recording, when I started recording this episode, it was light outside and it is currently dark. We are, I started at about 545. It is 820. So that might be part of the reason why my brain is melting. Uh, so we'll just, that's, that's the excuse we'll use for today. So, uh, thank you everybody for listening to the episode. Thank you everybody for supporting the show, all the, the Patreon supporters.

[01:03:51] I hope you enjoyed this episode and I will talk to you next week. Oh, next week, we're going to start off with Kim Lundgren. She owned an EV1, uh, which is super cool. So, uh, Kim is going to talk about, uh, just, you know, practical things that cities can do or a local municipal, not municipalities, local localities can do to kind of improve their carbon footprint.

[01:04:20] And she's got some good practical advice. It's not like, you know, we just built a whole building and it's, it's not efficient. So let's tear everything out and make it more efficient. It's like, Hey, if, if old equipment is, is on its way out, let's pick a product. Let's pick something to replace it. That is going to be more efficient, that kind of thing. And man, uh, Kim's a lot of fun to talk to you. And then I get a chance to talk to Max Patton. Jean put this together, but, um,

[01:04:50] Max is with EVs for all America, which is a bipartisan, uh, uh, organization that is their, their specific goal is to, you know, bridge the divide when it comes to EVs, which I think is just super dumb. Uh, not that what they're doing is dumb. I think that, you know, how a car is, is powered, uh, is a political statement. I think that's kind of dumb,

[01:05:15] but he actually does a really good job of explaining kind of maybe why and where that all started, which is cool. And then we have Jill Simonello and man, Jill and I talked yesterday and what a delightful person she is. One, uh, I was very intimidated. She clearly knows more about cars than I do. But, uh, two, we just like,

[01:05:41] we had three things that we were going to talk about and we just kind of rolled into a bunch of other things. It was a fun conversation. So I think you're going to like that. So it's going to go Kim and then Max and then Jill. So, uh, I hope you all, uh, are looking forward to the next three episodes. I'm looking forward to sharing them with you. And, uh, thanks everybody so much for listening and I will talk to you soon.

[01:06:28] If you like the show, please take a moment to rate review and subscribe. It really does help the show to grow. Thank you for listening. Thank you.