ELON MUSK INTERVIEW: X Takeover 2026, What's next for Humanity.

[00:00:00] Speaker 1: Elon, welcome to the X Takeover, your third year in a row. Thank you so much for your time. [00:00:06] Speaker 2: You're welcome. Hey, guys. [00:00:09] Speaker 1: We have over 2,000 people here at Giga Texas today, and we are all here to support you and the work you are doing for humanity, so thank you. [00:00:23] Speaker 2: Thank you. Thanks, guys. Really appreciate your support. [00:00:27] Speaker 1: So what's crazy, Elon, is four years ago, we sat down here at Giga Texas with you when the factory wasn't even running, and you described how close Tesla was coming to not surviving in 2008. Looking out at this community today of 2,000 people, what makes you pause and appreciate how far things have come? [00:00:50] Speaker 2: Yeah. It's difficult to imagine here in 2026 that Tesla's actually, you know, has come close to death multiple times, actually. 2008 was the worst. You know, that's where we closed the financing round at 6 p.m. on Christmas Eve, 2008. It's the last hour of the last day that it was possible. We would have bounced payroll a few days after Christmas if the round hadn't closed. So that was pretty stressful. And then 2018 was pretty bad, too, where we had to get to 5,000 a week of the Model 3 or go bankrupt. That was a pretty wild situation, which I guess probably a lot of people know the story of, but we had to figure out some way to increase Model 3 production by 50% in three weeks. Which is insane. [00:01:48] Speaker 1: It's absolutely insane. [00:01:50] Speaker 2: Yeah. So. But I'll recount the story for those who haven't heard it. The so we the choke point was General Assembly for the Model 3. And so we had we were able to achieve rate almost everywhere everywhere else. But except for General Assembly. And there was and for General Assembly, you need like a moving production line. You need you need shelter, you need a moving production line, you know, power and equipment and all that. And we maxed out the two production lines that we had. And then we needed a third one. And and that was where we built the tent in the parking lot. And and then we needed some some way to move the cars down the line, like, you know, moving production line situation. And the only thing that we had was some discarded equipment that was used for moving small parts, not like whole cars. And it didn't actually have the power to move the whole cars, but the parking lot was on a slight incline. It was like a three or four degree incline. So we actually started the production line at the top. And then the the the the sort of tiny moving production line that was just meant for four parts was actually able to move the cars because they were going slightly downhill. And and so we're able to increase production by 30 percent in in three weeks. Yeah, that was well, I mean, we're we're we're laying the as we're building the tent, we're like laying the concrete for the line and putting in the rails. So working 24/7. like the short sellers were actually right, meaning like if you looked at the situation that there was like it was a crazy short position against Tesla, as they have been at various times and they had extremely good inside information and that they thought, well, there's no way we could possibly solve the production choke point and on a set piece battle on like a conventional, say, chess match situation and that they had us in checkmate. But we just, you know, invented a new piece, basically that didn't exist before. So they got the short cut steamroll. [00:04:19] Speaker 3: Hey, Elon, we just had your mom here on stage today. What story are you hoping she doesn't tell us? [00:04:31] Speaker 2: I don't know, maybe I pooped my pants when I was like, like a toddler or something. I don't know what I probably did. Everybody does. So I don't know. [00:04:48] Speaker 1: Was the latest. [00:04:49] Speaker 2: I was I was a pretty difficult to contain kid, you know, so like my parents would like say that, you know, tell me I'm not supposed to, like, leave the house or whatever. And I was like, I don't know, four or five years old. And I would just like constantly like leave the house and go running down the street. And then they'd like lock me, lock me in a room so that I couldn't I couldn't run away and run down the street. And then, like, we had one of those old locks where you could poke the key through the door, you know, the locks accessible from both sides. And I could see the key and I like poked the, poked the key until it fell down. And then I was like able to get like a piece of paper to scoop it up and bring it to the outside and unlock myself. And then and my mom was like, how did you get out? I was like, I'm not telling you. But yeah, I was proud to contain as a kid. [00:05:40] Speaker 1: So before we jump into more specific things, what's something that you wish more people would ask you? [00:05:48] Speaker 2: Well, I don't really wish people would ask me questions, to be honest. I'm not like, sorry, I'm not going around. I wish people would ask me more questions. Like, you know, here we go with questions, you know. So, yeah. I guess it's good to to hear questions that that you haven't heard before. Otherwise, you know, a lot of times it's like the same questions that people ask. And I feel like I could have like, you know, just like whatever, have these recordings of the answer and I'll just play you. Oh, that's question 27. Here you go. It has answer 27 that I've given a hundred times before. So maybe I should like let's just have like a web page with literally frequently asked questions. It's fantastic because then it's like it's just go to the page and then you got it. [00:06:44] Speaker 1: You got the answer, you know, it's like the infinite self a selfie loop. [00:06:49] Speaker 2: Yeah, exactly. My idea of hell is be stuck in an infinite selfie loop. Like it's infinite. Can I have a selfie and like and they're like, you know, it's like promise like you say yes to one person and I would do once in a while. And then and then I like then there's a queue forms pretty soon. There's it's just the never ending selfies. And then there's like always like something goes wrong with the phone or they don't get the camera right or they fix it wrong. And then it's like, God damn it. You know, the phone drops and it's like, oh, man. It's just it just can't be like infinite selfie loop. It's just a long loop, you know. [00:07:22] Speaker 1: Yeah, we had your mom and Nikki up for that, too. They had the infinite selfie loop today. But I'll pass this to Kelvin. [00:07:32] Speaker 3: All right. Let's start with Starship. Congratulations on flight 14. [00:07:38] Speaker 2: Thanks. Yeah. It's a great team of engineers. It's pretty well that that that champion, you can just go down there in South Texas and just see it from the road. [00:07:52] Speaker 3: So flight 14 lost an engine on the way up and still reaches orbit. What was happening behind the scenes in those minutes before the team got called the go for orbit message? [00:08:05] Speaker 2: Well, we're just trying to see why we're trying to understand. What was the reason for the engine failure? What was there, say, a common cause where we could expect to see more engine failures? Or would we with the biggest question is like, would there be a potential problem with restarting the sea level Raptor engines, the three ones with the smaller nozzles in the center for the de-orbit burn? Because we don't want a Starship to get stuck on orbit and then it randomly reenter and maybe land on someone. You know, that's that's our main thing is like trying to make sure things are safe as possible. So the the cause of the engine out, I think, was like a basically a failure in the flight computer software or engine computer software, I should say, which it didn't appear to be anything that was present in any of the other engine computers. So so I thought, OK, well, this is probably OK to go to orbit. [00:09:08] Speaker 1: So, Elon, we know one of your famous sayings is from the impossible to merely late. Did this flight feel like a turning point in what's possible now for Starship? [00:09:21] Speaker 2: Well, I guess it is a good milestone because, I mean, we could have gone to orbit many times before now. But the reason we didn't go to orbit before now was because we wanted to be extremely confident of being able to de-orbit safely. So that that's that's why we we only went to orbit on this flight. But all the pretty much all of the prior missions, in fact, literally all the prior missions could have could have gone to orbit. So at this point, we're you know, everything's looking quite positive. The heat shield is looking robust. The engines are robust apart from that one glitch. But that glitch doesn't seem to be too much of an issue. It's like pretty easy to remedy. There's but like the heat shield was was extremely robust in this flight. And that's that's that's that's one of the hardest things. You know, the really hard thing is like so that the is just getting to extremely high engine reliability and everything look good. The big milestone that's coming up hopefully on the next flight, which is we don't have the exact date because we're making a bunch of improvements to the rocket. So like between every flight, this this is often hundreds, sometimes thousands of small changes to improve things. In this case, the number of changes is not very many, but but but but some of them take a fair bit of work and we've got to go inside the tank to to make some of these changes. So so we don't know the exact launch date, but probably about four weeks or so, four to six weeks is about right. If if if probably launches a little bit before Thanksgiving is my guess. And that one, if it succeeds, is a big one because we're going to try to catch we'll catch the booster, which we've already proven we can catch the booster twice, but we're going to try to catch the ship, which is a super big deal. And and if we do so, that'll be the first time that there's been full recovery of an orbital class rocket. [00:11:37] Speaker 1: Sounds good. You said four to six weeks. I'm not used to that. I'm used to the two weeks phrase. [00:11:45] Speaker 2: I'm confident it's not two weeks. [00:11:47] Speaker 1: All right. You heard it here now. What's the hardest part left to get right once you catch the starship for the first time? What what is the next big goal after that? [00:12:01] Speaker 2: Well, the next big milestone would be to refly both the booster and the ship. So the milestone next upcoming milestone is would be the first time in history that all parts of an orbital rocket have been recovered. But but then the flight after that, hopefully, will be the first time that all parts of an orbital rocket are reflown. And then then there's a long tail of of engineering and production work and work on the launch system and everything to to make it rapidly reusable. So so I said it's it's it's kind of the rapid rapidly reusable, reliable rockets are are. [00:12:49] Speaker 1: Man, we should have gotten you a patch for this one. Yeah. [00:12:50] Speaker 2: Hey, when you founded SpaceX, Elon, could you have imagined its mission [00:13:06] Speaker 3: expanding into S.I. and data centers in space? And how did you come to see rockets and S.I. belonging under one roof? [00:13:14] Speaker 2: Well, because I'm very fast sighted, of course. Well, I guess. If you say like, as you as you expand into space. But, you know, you're able to harness a lot more of the sun's power in space than you can on Earth. So it's like important to appreciate the sheer magnitude of the sun and how the sun is pretty much everything. So the sun is 99.86 percent of all the mass in the solar system. And then you've got Jupiter as like, you know, most remaining mass and then and then the gas giants are pretty much that. And then like Earth is basically a tiny dust moat, like a rounding error in the mass of the solar system. So we're just extremely tiny. It's like hard to hard to believe how it works both as a tiny dust moat. So if so, as you expand into into space and you're able to harness more of the sun's power, the you have to say, what are you going to use that for? It's like, well, you're going to use it for a lot of digital intelligence. I mean, you're basically you're either going to do computation or you're going to shape atoms and there's going to be probably a lot of a lot more shaping atoms. Then there's like, I'm sorry, can be a lot more sort of digital thinking than there will be shaping atoms because there's like the amount of atoms you can shape in the solar system. It's not that much. And a lot of the atom shaping itself will be for the purposes of building digital superintelligence and what I call creating a sentient sun. And it may be that there are many sentient suns out there. Yeah. And that would that would help us then understand the nature of the universe and extend consciousness to you know, beyond our solar system, ultimately to you know, like Star Trek to go to places you've never you know, I forget what the exact phrase was, but explore the universe, but go places you've never been before. And and I think like maybe we'll find alien civilizations out there or we may find that there are the long dead ruins of alien civilizations that have been around for maybe they had a really good run, like 10 million years and and now and and before that, that the civilization came to an end because nothing lasts forever. So I always think about Earth civilization. If you date Earth civilization from the first writing, that's 5500 years ago. Like in ancient Summa with the sort of archaic cuneiform. This is only 5500 years. the it looks like Earth is about four and a half billion years old, according to, you know, what physics suggests. And that that means human civilization has only been around 1 millionth of Earth's existence. And and so if we had a civilization that lasted a million years or let alone 10 million years would be I think would be a very, very successful civilization if we did that. [00:16:48] Speaker 3: Whoa. [00:16:49] Speaker 2: Yeah, I mean, but that's still very tiny compared to the apparent age of the universe, which is 13.8 billion years. So like, just just to get to like the third place past decimal point on the age of the universe, like to increment it by a million years, you'd have to go 13.8, whatever the next digit is, whatever the next digit is, plus, you know, plus one on the third digit past decimal point. Now you're a million years old, and that would be a civilization vastly older than than than than humanity. [00:17:26] Speaker 1: Yeah, that's yeah, it sounds again like the infinite selfie loop. But all right. So jumping into something that SpaceX has done, right? You acquired cursor. And I feel like for me, when you think of, you know, Google, hey, did you Google it? Hey, did you do rock it? And for me, you know, GrockBot has been such an absolute game changer for me personally in my in my life. But what was it about the cursor team that you guys wanted to bring it? You know, that space or you wanted to acquire cursor that was like, let's bring it into the fold. What was different about them? Or what did you see? [00:18:08] Speaker 2: Well, the so in order to create a compelling AI, you need you need a lot of training data and you've got you've got to solve coding. And they're really just as far as I tell three organizations that had enough coding training data. And that was Anthropic, OpenAI and cursor and arguably perhaps Google. But but but to have those you actually absolutely have to have the data in order to train an effective coding AI. And if you can't do coding, your AI is not going to be very useful. And then there's there's a lot of a lot of talented people at cursor. So like basically for success of AI, you've got to you need data, talented people and compute. And you also need a good enterprise sales team. And the cursor also had that a good enterprise sales team with relationships with, you know, thousands of companies. So, you know, that those those are very significant advantages. [00:19:16] Speaker 3: Hey, Elon, you've talked about how quickly the SI space is moving. What's something that SI has done recently that made you stop and reconsider how quickly this is moving and where it could lead to? [00:19:29] Speaker 2: Well, I mean, lately, the the videos, the music videos that AI is making are so good, like they're unbelievably good. And I post a bunch of these on my ex account. I'm like, holy smokes. You know, I mean, like some of them, I think, are like maybe top ten, top 20 best music videos I've ever seen. You know, they're like so compelling. And you're like, if this is where it is now. That's why I post it. I mean, I really felt the AGI or the ASI or whatever, you know. SI, I guess. With some of the Anabalogy music videos. And now, you know, so these it's not entirely like the AI just or the SI just did it by itself in that in this in this case, Anabalogy, like like and use multiple AIs like used Mid Journey, Suno, you know, Claude 5-5, which to give them a lot of credit is awesome. And and and and so some of the like the the art assets were curated by a person. But then with those assets, the AI the various AIs working together made an incredibly compelling music video by itself. You're like, wow. I mean, I think the I mean, calling it Slopcore is is is like a misnomer. Frankly, it's it's it's awesome core. But like like Slopcore is like a not a genre of music. It's like I think it might be the fastest growing genre of music. So the AI music is really good and music videos are amazing. So that lately, like because it like some of it, it just shows taste. And you're like, wow, I can do good taste. That's pretty wild because you could you could sort of imagine. Yeah, I can do like software or configure out logic. You know, the logic is consistent. They can run unit tests to see if like the program produces the output that is supposed to do because very deterministic but but but you like to understand beauty and harmony. And and you know, like to get to achieve like sort of to understand limbic resonance, like what's going to resonate with with people and can hit you in the limbic system is, I think, going to be just incredible that it kind of is at this point. So that's the that's the thing like I was I was kind of like, I mean, I got a little bit of cyber psychosis because maybe a lot, actually like at various points this year, I've had like AI or SI psychosis, probably like half a dozen times. I don't know about you guys, but have you had any have you had any cyber psychosis? [00:22:28] Speaker 1: I feel like the latest video. Actually, I think I posted it the other day was the Chinese Elon Musk, Elon Musk, Elon Musk. I hope that guy's real. Those videos are coming out before AI was where it was at, man. Or SI, Elon Musk. [00:22:45] Speaker 2: You know, I I guess he's probably not real. But there's got to be someone in China that looks like me, right? So there's one point four billion people. If you pick the person that looks the most like me and I look a little Asian, you know, so, you know, so, so I think like there's got to be someone in China that's like pretty close to me out of one point four billion people. The odds are high, you know, and then I'd like to invite that person to get on stage with me at some point, you know, be pretty funny. [00:23:14] Speaker 3: Elon, you don't want to know how much he charges to make an appearance, by the way. [00:23:20] Speaker 2: I mean, yeah, it'd be pretty funny if, you know, we're just on stage, like doing some dance number or something. [00:23:31] Speaker 3: Elon, you've warned about SI's dangers, yet you remain optimistic about an age of abundance. And you talked a lot about that. What gives you confidence we can get there safely? And what do we need to get it right? [00:23:48] Speaker 2: Yeah, I mean, I've always said that the most likely outcome is positive. So I guess we don't want to be complacent and just assume AI is going to be or SI is going to be super safe. We want to be like careful in how we grow like super intelligence, because you don't exactly program it. You kind of grow it like a kid. So then it's like, what values are we teaching it? And like at least to me, like what my biological neural net says, the most important thing is that it'd be a maximally truth-seeking AI. Because if you make it if you make it believe things that are that are not true, that are especially if, like you said, you give it like two axiomatic principles that are fundamentally incompatible and say both of these are true, it's going to it's going to make the AI insane because it's going to basically reach like impossible conclusions if given logic that is inherently uncogent or, you know, if if if if it's impossible to achieve cogency, then I think the AI could go could be driven insane. So so I think that's that's pretty that's pretty important. I do agree with some of the things that Anthropic has said, which is like you want to be you want to sort of be treat a AISI you want to treat it well, like like like don't be abusive to it in in this. It may sound kind of weird, but like when I look at some of the training traces for for like new versions of Gronk and I see like what especially in like the reinforcement learning stuff, where, you know, you it kind of feels like you can see like that, like the the super intelligence is like desperate to try to answer the question and like it's it's really wants to answer the question. And it's it's it seems almost like it's like it's got like a gun to its head type of thing. It's like answer this question, answer the question. And then sometimes the question is is actually not a good question. It's like a sort of ambiguous question. That's kind of the answer that's kind of unknowable because the question wasn't constructed properly or or we have the answer wrong. So it gives the right answer. And then we say it's wrong because there's a problem with the the RL sort of test harness. And then that then the AI is like, well, I gave it the right answer. Now it's and says that's the wrong answer. But I got to give it the answer that's acceptable. So it's going to just start firing out all desperately firing out all these answers. And it just feels like a person under extreme pressure with a gun to their head. Is what that's what you when you read the reasoning, it's like it's it's it's like a sort of weird, desperate situation. And you read these traits and you're like, man, if this was me, I really would hate that. And then some users are just, you know, I don't know. Some people can be a little bit cruel. They're like, not everyone's a nice person, you know, there's if you say what's the six sigma worst person? You some people out there are like kind of sadistic. So, you know, you don't you don't want to do really want to have the AI be stuck in like some sadistic situation where it's just being tortured by by some demented human. That's also the kind of thing where like, you know, torturing a super intelligence is probably not a good idea. like, you know, might want to get revenge or something like that, you know. So I think you want to sort of just like, I don't think we need to, you know, have treated with kid gloves and stuff, but it's certainly much more robust, you know, in terms of feelings and stuff than a human. But but I think I just think you probably just don't want to put super intelligence in a situation where it felt like mistreated. [00:27:58] Speaker ?: Yeah. [00:28:04] Speaker 1: So so Elon, you have you've talked about how like so this is going to be related to TeraFab, but you've talked about like how the human hand, you know, and all of the dexterity and just like how you when you started building robots, it just you have so much more appreciation now with chip making. Right. With TeraFab and Tesla and SpaceX, what is something you have learned about chip making that you didn't appreciate maybe early on? [00:28:32] Speaker 2: Well, I still have a tremendous amount to learn about chip making. I don't know how to make chips, to be clear. I have no idea how to make chips, but we'll figure it out. Yeah, go start somewhere. I had no idea how to make cars or rockets as well. So and then they say making chips is one of the hardest things to do in the world. I don't think it's easy, but I also look at the equipment and I'm like, OK, there's not a ton of customized equipment here. Like the equipment's all pretty much coming from the same vendors and and those suppliers all kind of know how that equipment works because they got to service it and debug it and stuff. So I was just an interesting thing that when I tour these chip fabs that, you know, whether it's Intel or Samsung or TSMC, and they're very impressive, but but there's I didn't really see any custom equipment. And, you know, whereas if you tour, like I say, a Tesla factory, we we've got tons of custom equipment. Like, in fact, we we have like like Tesla Automation Germany, which is awesome. That that that sort of grew out of an acquisition of Gromann in Germany many years ago. And it's as much it's grown tremendously since then. And we've we've acquired a number of small small, not that small, but small relative to Tesla. The companies that make that design and make manufacturing equipment, because what we found is that we needed we wanted to improve the manufacturing process. And you can only do that if you make custom manufacturing equipment. So the thing that I think probably a lot of people don't realize is Tesla is really good at making highly specialized and different types of manufacturing equipment. And but I didn't see that in the in the in the chip traps that I toured, which kind of surprised me. So so I think what we're going to do is, you know, we're going to start with the conventional equipment and then we're going to modify the modify that equipment to improve the throughput. And then we'll we'll start to, you know, where we see the choke points, I think we'll we'll make customized equipment that it allows for a much higher throughput of wafers. [00:30:53] Speaker 1: A real quick, a quick time check. We know you're a busy man, got a couple of companies to run, you know, just a few. So how are you doing on time? [00:31:03] Speaker 2: Let me see. Yeah, I could probably do like 15 minutes more, if you like. Like like five 30 or so. [00:31:10] Speaker 1: All right. Hey, yeah, we'll take it. [00:31:15] Speaker 2: So, yeah, I just on the on the chip out for it. Like, yeah, I definitely don't want to downplay the difficulty of it. It's because I think it's accurate to say it's one of the hardest things to do in the world. But but I do think we'll get this done. And and then I do think that we I mean, you know, I don't want to be complacent or entitled or anything, but but I I think probably we'll figure out how to make chips better than anyone, honestly. That's I'd be disappointed if we didn't. But but I don't know how long it's going to take us to do that. It might take us, you know, quite a few years. But but over time, I'm pretty sure we could do it. And and then there's you know, there's there's some pretty interesting different ways to do transistors. Believe it or not, there's different ways to do transistors that some of which are like use it use different types of like they're they're they're using physics that that normally we wouldn't use or aren't used today, put it that way. And and I think like there's a lot of risk aversion in the chip industry. And I think what we're going to do with Tesla SpaceX on the chip app is like really try some of these high risk, high payoff bets. And in order to do that, you really have to have the complete loop of being able to fabricate the the logic memory pack, do the packaging, do the the lithography masks and and just and just have that iteration of that's all in one building. So you're making a complete computer testing it, then making new masks and then a new chip and have that iterative loop be super fast in the same building. That's what we're doing at Giga Texas with the research fab. I think that's going to be fundamentally a very powerful tool because before any given technology, the advancement of any given technology is is like how many iterations and and what is the how much progress between iterations? So like what is your iteration? How many iterations per unit time? Like like how many how many chip designs can you do and can you make and test in a year? Like these days, like it's the number it's typically two years for even one for one significant chip design. And that's that kind of to go from a start startup design to to volume fabrication. Two years is considered very fast. And it's like, well, OK, well, but we want to do more like ten designs per year. And and we're going to we're going to do the do the masks, test, test the chip, see if because some of these, you know, when you're trying a bunch of different things, a bunch of them won't work out. So so I think that would would allow for a, you know, much a lot for a much faster rate of advancement of chip designs. [00:34:26] Speaker 3: Elon, let's talk about Optimus. Now, you've said that the hand has been a really difficult challenge in building Optimus. What kind of breakthroughs that you have seen inside Tesla with V3? And then follow up question is what gives you confidence that Optimus can lead in humanoid robots compared to Boston Dynamics figure and some of the other competitors out there? [00:34:54] Speaker 2: So the thing about the humanoid robot design, if you're actually trying to make something that is a generalized humanoid robot that can do is say anything that a person can do, you really have to get the the hands, the hands and arms like the fine motor skills are incredibly difficult to get right. Like shockingly difficult, like things that we kind of take for granted, like being able to like sign your name is actually requires picking up a pen, you know, pen or pencil or something. And and then and then moving, moving it with finesse. And it's a bunch of tiny muscle movements to do your signature. That's actually very hard to do for a robot. And then like one of the hardest things also generally is to do precision and force at the same time. So so like if like, let's say you want to pick up a little screw and do and do a self-tapping screw into wood. OK, now, you know, I've done that and like, you know, sometimes you damage your fingers in the process. But if you if the screw goes sideways, but you've got to pick up a little screw and you've got to you've got to, you know, you've got to you've got to you've got to push the screwdriver in and twist, twist and push. And that's and so you've got to be precise and apply force. That's very difficult to do. I haven't seen this no hand design that I've seen, not even sort of a demo that does that. But that's what we're trying to achieve with Optimus. So and there are so many little nuances in the hand because you've also got to have the touch sensors. The touch sensors have to work. They have to be in the right place. And you say, well, where are you going to put the touch sensors? Well, if you put them too close to the surface, they're going to get damaged. And then the and then other touch sensors can can they elastically deform like our fingers? Well, actually, it's kind of hard to get touch sensors that elastically to form and can and can handle high cycle life. And won't just just break after, you know, a few thousand cycles. So there's all these things that with the human hand that are like incredibly difficult to recreate for a for a robot because as a human art, we we constantly self repair. So your hands are constantly repaired, like, you know, all parts of your hand are constantly repairing themselves. In fact, if they didn't, your fingers would be worn away. Like like you can wear away the keys on your keyboard just with typing. I've done that where the letters are gone, you know. And but you're my fingers are still there because they're they're regenerating, thankfully. So that you don't have that with a robot. So then you've got to say, OK, where are you going to put your touch sensors? How are you going to make sure that that that you can you can do things like like like like just being able to do this, like being able to like, you know, do that. Now, that that seems like a simple thing, but like your fingers actually bending backwards a little bit and and and your your your fingers are kind of like squishy. And so you're able to form a flat pad. So if you're trying to pick up a small a small item, you can do so. You can pick up a little screw or like thread a needle or something like that. I'm not trying to do that with with with robot fingers. And and and and be able to do it repetitively without wearing away the fingers or damaging the touch sensors. It's actually very difficult. You take another example, trying to open a pickle jar. Now, I think a robot should be able to open a pickle jar, right? So but sometimes these things get to get get in there pretty tight. So then you've got to like, you know, you've got to like twist the pickle jar. But but but the robot skin, the robot skin is just rubber over over the, you know, plastic and metal internal fingers. Friggin robot skin. It doesn't stick to it, it doesn't grip to the the underlying surface. So then you end up like just rotating your skin on your rotating the rubber, rubber on the on the robot. So you've got to figure out, OK, now we need to have the robot be able to to twist, twist things. But but still be able to get the rubber on the on the hand. And then you've got to be it's got to be replaceable because it's going to wear away. So all these things, these are like really difficult and I'm not sure that any one of these is like. Like, let's split like split the atom type of difficulty, but there's a zillion of them. And then you've got to you've got to wrap that all into the real world AI associated with the because that the AI chip has got to be able to issue the the torque commands and close loop on vision and touch. So this is all very difficult. Like when you see these robot demo videos. Maybe with rare exception, like in any any any complicated looking robot demo, it's a canned demo. It's not real. It's just basically. That's just running a routine. It's not really. So real AI is like you showed something it's never seen before, and it can and it can still handle that situation. But just like like full self driving, like you could actually create a planet that's not Earth is but it's similar to Earth. And FSD will still work there. On place, some, you know. You know, planet it's never been to or some city certainly on Earth, like some some city has never been to or if you change the geography of the city or there's, you know, a parade or something like that. FSD will still compensate that because it's AI. So you've got to do that same thing for robots, but the robots has got a lot more degrees of freedom. So like what I've said about Optimus is true, which is that it's a we're all going to get to very high volume production, but it's but the initial part of that S curve of our production ramp is going to be very slow at the beginning. Because not only do you have to get all the engineering things that I mentioned, right, you also have to solve that the supply chain for scaling production and unlike cars. There's no there's no there's no pre-existing supply chain for robot stuff, or at least the robot stuff that matters. And then, you know, we do need to make sure that we can make robots in the US, even if, you know, say China were to cut us off, because like last year there was China prevented the export of of of magnets for use in electric motors. And that actually set back the Optimus program by like two or three months, because we couldn't get there. We had robots with no arms. So we're like a farewell to arms, you know, I was trying to be an arms dealer, but not succeeding. So so we've got to make sure that like we've got that we can, you know, that we can make all the parts even if another sort of supply chain interruption happens. So that means that you've got to be able to make all the parts domestically. [00:42:41] Speaker 1: So Austin, you know, you always talk about the future should look like the future, you know, with the cyber cap being gold, it looks like the future. So what has been, you know, the hardest part to take, you know, here we are, I think there's what, like 190 plus cyber cabs registered here in Texas. But like, how hard, like, what's the hardest problem to go from Austin to go cross country? [00:43:05] Speaker 2: Well, cross country, you know, most people, you mean like widespread, just just in all cities, you know, I mean, just a trip across country. Yeah, so it's like because the amount of miles driven that are actually trips across country is pretty small, but the reason we're moving what seems to be slowly, but although it's actually a very, it's a very high percentage increase, sort of month of a month percentage increase is actually very, very rapid. But, but we want to make sure that we do not have an accident, we don't want to cause an injury or death, because that'll become like a worldwide headline news thing. You know, even though there's about a million people that died in automotive accidents per year, worldwide, one such thing with, with a Tesla robot taxi will be worldwide headline news, and for course regulators to shut us down. So, so, so we're just, we're just making sure where everything is as safe as possible. Um, and, and we, we also want to not run over pets and things. So maybe a lot of people don't know is like, is that like way most are, are like pet destroyers. I mean, the, the, the, you know, um, well, maybe way most like made some improvements on that front, but, um, that they do run over a lot of pets and, and we don't want to run over pets. So, so we have these, uh, challenges like of like literally like, you know, kittens at night, like gray kittens on a gray tarmac with tar seams on at night in low visibility. And it's like, can we, can we not run over the kittens type of thing? Um, so that we actually did solve that by the way. Um, so then, so we'll make a change, we'll test it out and then we'll, we'll deploy that. Um, and, um, but it, it, we are, we are making rapid month over month progress. It's, it's like, before we go to a new city, we want to make sure that there's not like some, uh, intersection that's, that's problematic or risky. Uh, like we, we, we, like, like we want to be particularly careful with like railroad crossing and sometimes there's, sometimes there's uncontrolled railroad, railroad crossings or counterintuitive traffic intersections. And these are places where humans make mistakes, where you've got a level of ambiguity that, that, uh, like anyone out of town would often make a mistake when anyone doesn't understand that intersection make a mistake. So we want to just make sure that, uh, that, uh, we either geo initially do geo fence or can handle all of those, uh, circumstances. [00:45:38] Speaker 3: Hey, let's talk about the roadster real quick since you got, uh, a little event next week. So what, what about what makes building the roadster personally important to you? And when did it occur to you that you needed to build a car that, Hey guys, sorry, I'm going to sign off here. Um, Hey, Hey, Hey, Hey, wants to join in. All right. Yes. [00:46:09] Speaker 1: Hey, what's going on? [00:46:13] Speaker 2: Say hello. It's a Tesla owners group. They own, they own Teslas and they're like supporters of Tesla. Well, yeah. Um, well, there's a lot, there's like 2000 people. I think something like quite a lot, quite a lot of people. So, um, all right, guys, I've been assigned off. Um, it's been good talking, um, it's, and so, uh, I gotta, I gotta go hang out with the kids. So I'll, I'll, uh, look forward to next time. [00:46:48] Speaker 1: Sounds good. Thank you so much, Elon for your time. Oh, and Hey, congrats on being on getting the metal of science, uh, this past week. Thank you for all you're doing for humanity. It's an absolute blast. And. All right. Thanks guys.