Tesla and SpaceX Need to Merge. Here’s Why. w/ Emmet Peppers

It's Dave. So, today is the Tesla Roadster event. I'm traveling with Emmett here. He just flew in last night or yesterday afternoon. And we're going to drive from Austin to Waco to check out the Roadster event. And while we drive, I figured we'd just record our conversation. I texted Emmett some topics he could kind of, you know, shoot out there so we could talk about it. After the Roadster event, we'll do another kind of thing. Might be in the dark though, so we'll see if it works out. But just to give our impressions on the Tesla Roadster event. So, this is going to be part one, just Tesla, SpaceX, AI, just life in general, what's going on. Hope you enjoy the conversation. And then we'll see you hopefully in part two with the Roadster impressions after the event. Emmett, how you doing? >> Good. Good. Yeah, and I think it's important to note that, you know, we've been hanging out for a little while at lunch already, but we purposely avoided talking about all these subjects until right now. We've been talking about more personal stuff, but now we're going to talk about all the stuff you guys like to hear us talk about for the first time in a while together. >> Yeah. So, for those who are joining us for the first time, we've done chats, many chats over the years. So, this is a pretty comfortable and usual for us, hopefully. For those who have joined in before, it can bring back some memories. So, I'm starting at Vesdy. We are leaving from Austin right now. >> Yes. >> We've got our Starbucks coffee here after having a Texas barbecue. >> Yes, driving with Dave. >> Good stuff. >> It's not driving with Daisy, driving with Dave. >> Exactly. >> So, the Tesla Robo-taxi, we first uh >> [clears throat] >> We went We did that when it first opened up over a year ago, right? That was the head monitor, safety monitors. Within that first week or two, I came out here and test rode it with you. >> Yes. I remember we were excited about it. >> Yes. >> Has it the I mean has it rolled out been slower than you expected? >> Very slow compared to what I expected. I expected, you know, the monitors to be out in like 2 weeks at that time. I was like, I think in 2 weeks or a month the monitors will be out and there'll be thousands of them, you know, but it's it's been very slow and careful and now we're at finally at an inflection point with the cyber cab cuz I rode in the cyber cab several times yesterday and a couple of times this morning. And yeah, I think this is, you know, for real like a turning point. I'm very careful on, you know, timing, but um this this cyber cab is the experience is I, you know, posted something last night. It's like legit and Elon reposted it apparently cuz I had like a thousand reposts or tweets or after that it was crazy, but uh >> So, >> I'm psyched. >> So, how many cab rides did you take yesterday? >> Uh I think I took four yesterday and two this morning. Yeah. >> Okay, so you got a decent experience. What was How was it? >> It was great. I mean, the only thing was this morning every time the two times I got in this morning, um both times there was like a little bit of a delay of like 30 seconds or a minute until it started up. It might have been like a bug or something, but it wasn't that big of a deal. I'm sure they'll work that stuff out, but the you know, everything about it was amazing in terms of um just I feel like I'm like in my own Tesla full self-driving mode instead of the camera, but the camera's not nagging at me to look forward anymore, you know? Cuz I'm so used to full self-driving in my car and having my phone up here, you know, and just like on my phone, but looking forward and occasionally I have my phone down here and then the camera's nagging at me. But I'm so used to the full self-driving driving me everywhere, so I was so it wasn't like didn't take me a long time to get used to it. >> Yeah. >> Have you driven You've been in the cyber cab a bunch, right? >> Yeah, it's like um Yeah, I think what's I mean it seems like because the rollout has been so slow since last year and I think what made it also interesting is like Tesla and Elon gave initially like pretty high expectations of rollout. >> Mhm. >> Um that it kind of like drained people's energy. Kind of waiting, waiting, waiting, and nothing happening. >> Yes. >> And so now that it's starting to happen, the tendency is people discount it as like uh it's just >> a delay. >> Yeah, never going to happen. Um and I don't know what the rollout's going to look like, but purely just on the experience itself it's like so good, you know. >> And there's like they're all over Austin now. There was like what, 400 robo taxis now or more in Austin roaming around and I when I was staying downtown Austin last night and riding around in the small morning doing a couple things before I met up with you I saw a every corner there's like a cyber cab or a Model Y with robo taxi on it. Like everywhere. It's everywhere. Way more than Waymo from what I see. It's only a matter of time before Waymo takes their, you know, couple hundred vehicles out of here, moves it to another city, I think. >> Really? You think they're going to >> Oh, serious? >> Yeah, I think so. I mean, if Tesla can lock down the airport >> Mhm. >> and get approved for pick up and drop off the airport >> Yeah. >> game over. You know, there's nothing else in Austin to conquer in terms of self-driving. >> Huh. >> Huh. Interesting. So, you I would would you say the cyber cab experience was better than you expected? >> Yes. >> Really? >> Way better. I mean, that big screen and infotainment there like you could, you know, the first time I did was FaceTime my kids and wife back home. I was like, "Look at this cyber cab. It's amazing. Look, there's a big screen. You you can watch YouTube. It connects automatically to my current profile." I was telling you I I like today I didn't even realize but I said like I'm going to say quiet doesn't come out hey Grok or whatever. I said that you know loud enough where picks me up in the cyber cab and and it I was able to connect to my Grok bot and like be like can you check my emails and tell me what is urgent or can you send an email to this place or what's this you know and it was just like in my own car using Grok bot to and I'm in the cyber cabin some other city you know and I didn't have to like sign in to anything it just automatically works it's incredible. >> Yeah it's a I for one of my rides I had my kids kind of walk out to the cyber cab. >> Yeah. >> It was my first ride actually and my kids I got to see inside a bit and >> I saw that video yeah. >> Yeah and then my daughter a couple weeks ago she's like I want to ride the cyber cab too you know >> [laughter] >> she still remembers >> Yeah. >> the cyber cab and she wants to experience it you know like taking a ride um >> Yeah. >> but yeah it's it's it's definitely I think like it's worth it for someone if they have a large Tesla holding to come out to Austin even for a couple day or two you know >> Yes. >> take five to 10 rides um >> Yeah I mean it's definitely worth it I'm glad I did it experiencing it is >> Yeah. >> a different level >> Yeah yeah. >> You know it's sort of like experiencing the full self driving for the first time as opposed to watching other people do it on YouTube or something. It just it hits you hits you harder and you just you understand. >> Yeah I think no matter how much you explain something to someone it just doesn't have the fidelity of like the actual experience of the cyber cab. There's so many things I think they did right. >> Yeah. >> I don't know about you but I think right now it kind of routes sometimes I don't know if you go if you go further away sometimes it routes in these like really long routes. >> Yeah I did go kind of far to this poker room intentionally that was kind of far away to like just play a little poker of my own for a couple hours and uh it routed through some neighborhood that I was like, "Why did it do that?" It was like avoiding like a It was like It was like there was a detour, but there wasn't a detour. It was kind of strange. I mean, I'm not in a hurry. I was fine with it. It maybe added like 1 minute to the the ride time, but you know, it went a few blocks out of the way, but you know, that's the one you know, if I'm in a hurry somewhere and I'm like late for something, just like full self-driving sometimes I got to take over and just be a little more aggressive to to get in the left lane or whatever in the highway just to add you know, save the extra 30 seconds or minute. But other than that, I mean, that's a small thing. I feel like it's ready for prime time. I would I would put my kids in that thing all day. I'd say, "Go go to school. Go to here. >> [laughter] >> Here's a cyber cab. Get in." And they would love it. They would be like, "Yeah, I want to go do this. Put me in the cyber cabs so I can watch my YouTube videos or play the games or whatever." I think it'd be cool if the cyber cab had a couple of controllers ready for you to pick up and start playing a game, you know, especially for kids. >> That's true. Yeah. >> You know, that just work, you know? I think that'd be neat. >> Yeah. Yeah. Yeah, it's going to I mean, what are your What do you think? Do you think they're ready to I mean, do you think roll out's going to happen? Is it a >> It should happen. >> It's going to take a long time. >> They're just afraid of that first accident and I think that's what it is and and I mean, I understand justifiably so with the way the media portrays things and um the political opportunity it'll provide for politicians to blame full self-driving on something or create outcry about something irrational. So, I understand that, but at the same time like we're talking about lives at stake and I think they should just push to roll it out. You know, this is a friendly administration to Tesla, I would think. Um it's not going to get much more friendly than this. And so, I think they should leverage that to push it out and save lives. And then also with incredible experience for consumers overall, you know. What do you think? >> Yeah, so my my hunch is they're waiting for a V15 to roll out more aggressively. But, because with V15, I mean, I think they need the AI 5 chip. And I think perhaps they're starting to put them in already in the cyber cabs. >> Really? You're saying the NFAS is going into some of the cyber cabs? >> That's my guess, yeah. I mean, I could be wrong, but maybe they have some early type of chips that are comparable or something, early production. >> Yeah. >> But, I think, you know, they want V15 because I think it's going to be a step change in terms of model size and capabilities. >> Yeah. >> And so, they want that uh, before scaling a lot. >> Yeah. >> But, what you got right now is you've got them scaling some, but it's like a few there's some few limitations. Like, for example, like some of the ways they route >> Mhm. >> may like it's taking 5 or 10 minutes longer on a 15-minute ride because it's like >> Yeah. >> going around places they're not sure. >> you think like perfect is the enemy of good is always going to be the case? >> Yeah, but I'm just saying like V15, they're going to be more confident with more, you know. >> Yeah, it'll be even better for sure. With routing, it'll be close to perfect. People won't even tell. You won't be able to tell that it isn't perfect. >> Yeah, and then they're also like not really aggressively expanding Austin. Um, it's been a while. >> you mean? >> the geo-fence has been stuck >> Mhm. >> for quite a while. Um, and so, I think like with V15, they're going to be able to scale to different city centers, like Austin. But, my question is like how confident will it take them to scale like to the suburban you know like the greater Austin area or the greater Dallas area like Yeah. Will B15 be good enough for them to confident to do that or is it a B16 or B17 thing you know like >> Yeah. >> It's like My hunch is it's like very data driven maybe. >> Mhm. >> It's like they're looking at all of the data all of the interventions or possible different stuff that's going on and I think they realize B14 is you know >> Not quite there for scaling. >> Yeah, they they would have scaled if earlier if B14 was good enough. >> Mhm. >> I think the only reason why they're waiting right now is I think they're starting Yeah, they're starting to scale a bit is because they have the early build of B15 and maybe AI >> think some of the cyber cabs or all of them are already on B15 with the AI 5 chip? >> Yeah, either that or I heard rumors of an AI what 4.5 chip or something like that. >> Okay. So something intermediate to me. >> It's different I think it's not the old AI 4 chip at all. It's it's a bigger model more capable Tesla feels it's like you know good enough to start scaling but how far will that scaling allow them to do? That's the question you know. >> Yeah. Yeah. Well, I sort of hope you're wrong and that they can scale with AI 4 chip because I want my set my Tesla at home to be able to have the unsupervised full self driving no problem but you might be right. I mean maybe they will uh allow unsupervised full self driving in certain geo fenced areas for example um and not everywhere until you have the AI 5 chip. Maybe something like that or >> Yeah. >> Maybe the AI [clears throat] 5 chip I thought you know I was hoping that the AI 5 chip was really just for Optimus primarily but also helpful on the on the full self driving to be >> Mhm. >> a little an extra nine of uh accuracy. >> Yeah, I mean I'm kind of somewhat disappointed that I thought by now I would be able to like close my eyes and take a a nap for at least like a 20-minute drive in an area that FSD is comfortable with, you know? >> Yeah. >> But I can't even do that right now, you know? So it it makes me question like A, is this AI 4 chip really you know, going to handle like really unsupervised FSD? I mean maybe, but I feel like you know, I'm hesitant because like with hardware 3 it's been kind of like disappointment where you get this distilled you know, version 14. >> Yeah. >> But like I was riding my my wife's hardware 3 Model Y a few days ago and it just is not as good as hardware 4. >> No, it's not. Hardware 3 is definitely a step backwards. >> stops. It doesn't, you know, navigate the way it or turn the way it wanted. It just I just don't feel as safe in that car. And that's where I don't know. I mean maybe AI 4 will work good enough, but with if the models get so good that you can get a smaller model that's you know, really good, but I don't know. I think Tesla's you know, AI AI 5 is V15 is where it's at for Cybertruck. >> Yeah. >> Um yeah. >> I think it'll be a big game-changer, too, when they I think he's talk he mentioned this on X or something or someone's commented and he's like it's coming or Ashock has made Ashock has maybe mentioned it that you'll be able to talk to the full self-driving and be like pick that parking space over there to the top right of the screen or you'll be able to like say hey, don't take that detour, go straight instead. You just talk to it and it'll respond and and do that accordingly. And I wonder if that's going to make a big difference if and when they can integrate that into the full self-driving where you can just like cuz right now the most most of the time I take it over for parking or getting out of a parking lot, right? Would you say the same? >> Yeah. Yeah. Yeah. >> But the actual drive itself is usually really good. Just little nuances where you want to adjust it, you know? >> Yeah, like it doesn't like park in my garage. It just It like stops right before my garage in the driveway, you know? I have to like stop it. But yeah, like it's gotten really good. I mean, hard before, I have to admit. But again, it's like you know, you can't sleep in the car yet. So >> No. I I actually I mean, maybe I shouldn't say it, but sometimes alone in my car, I'll be driving back from like the airport. I'd do the airport you know, picking people up, dropping people off. But I'll be alone or something and I'll be tired and put my sunglasses on and just kind of like I I spot I've been at a thousand times and it's flawless every time for like a half hour. I'll just like close my eyes and probably take like a half >> [laughter] >> hour. No, like five or ten minutes. Like a power nap almost. I mean, any jolt or anything would wake me up. But I like trust it so much that I can like doze off a little bit with my sunglasses [laughter] on. And then I'm like refreshed. I'm like, "Okay, that was my power nap for the day." You know? You know, not with anyone in the car, but just alone and it's it's perfect, you know? It's better than a human, you know? And it's probably better than if I tried to stay awake and was like, you know, being an ass sleeping and driving myself, I would say, you know? So it's you know, I I do feel like I've gotten the first nap in a in the self-driving that way a couple times. >> Yeah. Yeah. >> [laughter] >> Yeah. >> That's funny. Um yeah, so like when I think it reminds like whole FSD is interesting cuz this like FSD Tesla started FSD before the whole ChatGPT moment and the whole crazy, you know, LLM wars and stuff. >> Yeah. >> And like part of me thinks that like Tesla should be more in the thick of things with AI generally, not just with FSD. >> Mhm. >> Like part of me thinks like I think Tesla could have done data centers, you know? Like >> Yeah. >> well, you know, cuz they had a lot of the pieces, you know? Um part of me thinks that they could have done large language models well, too. Um >> Mhm. >> But you know, they've chosen to kind of really focus just on FSD. And and Optimus, of course. Um and leave and Elon has chosen You know, he focused his AI efforts initially with xAI. >> Yes. >> And then combined with SpaceX. >> Yes. >> Saying that he needed like a new structure to incentivize, you know, founders whatever. >> Mhm. >> Which, you know, I understand. Um But it puts Tesla in an interesting situation where it's like there's pros and cons. Tesla isn't in the thick of things with frontier models, you know? >> Yeah. >> Which is good and bad in some ways. And they're not in the thick of things with like building out data center infrastructure, either, you know? >> Mhm. >> Um But they do have a strong position in the kind of the narrow AI use case of FSD. >> Yeah. >> And it's future use case with humanoid robots. >> Yeah. >> Which is kind of a decent position to be in because the like even if frontier model LLMs get really good >> Yeah. >> driving requires so much of this like instant uh milliseconds timing. >> Yeah. Reaction process. >> so you need like the fastest hardware models like tuned for that hardware integrated with the car. >> Yeah. >> So, there's a lot of moving parts that that make Tesla's position you know, pretty strong um against competition. >> Yeah. >> So, in some ways like they have a a a a good competitive position, you know? Like if they were in the frontier model race like it changes every 6 months like, you know, who's in who's in the lead changes and >> Yeah. >> it's just it's crazy how rapid things can, you know, go up and down and who knows like open source Chinese models or other models could like take out you know, the certain frontier models if they're not at the top, you know? And so it could be a bigger risk factor if Tesla went into frontier models, but they weren't the best, you know? >> That's true. >> So, in some ways like it's okay Tesla has stayed with FSD, but in some ways I'm like you know, I wish Tesla would be more hands in hands-on with the AI boom, you know? >> Yeah. >> That's happening >> Yeah. >> and take advantage [clears throat] of that. I mean like you got SpaceX like taking full advantage of it with, you know, data center builds. >> Yeah. >> You know, going to space with their, you know, >> Yeah. >> space, you know, ambitions. >> Yeah. >> Um xAI um >> Yeah. I wonder if he just doesn't like since he's gone so so all in with xAI and now SpaceX with with a lot of that. And he's probably been planning that for a year or two before he founded xAI, maybe that's why he's purposely left Tesla out of that equation somewhat knowing that he's going to eventually leverage all the resources together, maybe a merger. And we'll talk about that later maybe a possibility. >> No, we should talk about it now. >> Okay. So, that's where I'm like, I feel like Yeah, Tesla should do it. I mean, Tesla SpaceX whatever like we should they should merge, you know because it's like for Elon, he probably had a choice of what how to get involved with the AI wars, cuz he saw OpenAI, you know, just kind of take a lead with ChatGPT. >> Yeah. And didn't he have some of the Didn't he create XAI purpose like initially as a way for some of his top Tesla AI talent to kind of move towards for better compensation as opposed to staying at Tesla trying to I felt like there was something there where he like created a new entity for the AI only that the top engineers he wanted to retain and keep and he kept them at XAI that way. >> Yeah, I think I think it was mostly for the new talent though, like the founders. >> Okay. >> He wanted the some of the best people that >> more attractive risk reward for their personal incentive, you know, ownership of a startup. >> the co-founders became, you know, billionaires because of that. >> Yes. >> But, yeah, so I feel like Elon had a choice, like, okay, how do we take on OpenAI? This is like, you know, >> Yeah. >> 3 4 years ago, whatever. And he's like, okay, I could do it through Tesla, I could do it through SpaceX, or create a new venture, right? >> Yes. >> And he's used to, obviously, starting new ventures with Boring Company, Neuralink. He understands stock options, incentives, all that stuff. So, he chose XAI, a new company, to do it. >> Yeah. >> Um which, okay, but now that XAI has been taken over basically by SpaceX, and now it seems like SpaceX has gotten all this like extra momentum now. >> Yes. >> Cuz now they have like a frontier or Grok, you know, frontier model. They have all these the Grok the XAI data centers and now they are having these contracts with Nvidia for, you know, buying all the bunch of GPUs, building more data centers, and then it just leads into their SpaceX, you know, space data centers and you know, all this stuff. It just seems like there's so much momentum now >> Mhm. >> in in the AI wars that Elon has now with SpaceX and XAI, right? But it seems like Tesla is not really like >> It's like left behind a little bit on that front. >> Yeah, with the AI wars. Like they're still, of course, you know, it's chugging away with FSD and Optimus and all that stuff, but and sure, selling whatever megapacks and energy stuff. >> let's be real though, Optimus is the game changer for Tesla, but we don't know when that's I mean, full self-driving and robotaxis great, but it's the Optimus is going to be something potentially bigger than everything else they do. >> Yeah, but Okay, so even that people tend to think like, you know, Optimus is like this like amazing thing by Tesla, which I agree, it could be, but I don't think you could in a way isolate the hardware from the software. Meaning Optimus is going to be great not just because it's great hardware. It has to be great hardware, obviously. >> Yeah, the brains, the software. >> But it's got to have, yeah, the mind. >> Mhm. >> And Tesla doesn't have a general purpose mind. So, they're going to have to borrow that, right? Which, you know, they borrow Grok or whatever. >> Yeah. Maybe they'll be some cloud version they integrate into it or something. >> Yeah, it's just it's not their own, which they could, but it's like the other thing is like, for example, Sam Altman from OpenAI has openly said they're going to get into humanoid robots. >> Yeah. Good luck manufacturing. >> I I actually, no. You have to Okay, so I actually think they could be a a big threat. >> Really? >> And Anthropic. And the reason is is because OpenAI and Anthropic >> Mhm. >> they've got some of the most like they have some really, really like uh a what I call it? They have an amazing business model that I don't think people have figured out yet. >> Mhm. >> And what it what it is is once you're serving so much of your frontier models, right? >> Yeah. >> to people you basically are always a step ahead, right? Or two steps ahead, right? So, for example, we're on PaLM 5.1, they have PaLM 5.5 will come out any day, but they probably already have PaLM 6 ready ready for release or you know, it's just going through actual testing. They're probably working on the next model, the next model, you know? And so >> what you do is if you're OpenAI or Anthropic you use your extra compute. Like so, for example, there are these spikes, too, right? So, you have like some usage during the day that spikes, during the evening it goes down. And what you do is during any like like low usage time >> Mhm. >> you basically [clears throat] run your own projects during that time. So, what you do is but you don't run your own projects with the current model. >> Okay. >> You run your projects with the latest, craziest model that isn't ready for public release because you have [clears throat] to like test it, make all these safety checks, and all this stuff. You can go to the government, you know, you don't what you do is you run your own crazy, like the best internal models, right? PaLM 6.5 or whatever you know, Astra, they call it Bell or something, but your crazy next generation models on that extra compute whenever the compute is low and you run these really ambitious projects. >> Yeah. >> So, you run stuff that takes like >> When you say ambitious projects, you're talking about ambitious digital software stuff. >> No, no, no, no. It starts like that. But, for example, like OpenAI has already done stuff like solving like math problems that have never been solved before, right? Now, Anthropic has come up with this like health basically biohacking lab. >> What has OpenAI or Anthropic designed a factory to produce something yet? >> Okay, so hold on. So, here's the train of thought. >> Okay. >> So, Anthropic is doing this like health uh bio lab >> Oh. >> project, right? >> Supposedly. Yeah, I don't know. >> And so >> I didn't hear about that, yeah. >> So, the reason why this makes complete sense is Anthropic can let's say bring in the best scientists, create a lab with a thousand scientists, right? Running these stuff. Run not Gable 5.1, not whatever Gable 5.5, but they run the next gen model. >> Okay. >> Gable 6 or whatever >> On the digital AI, okay. >> Yeah, on these crazy moonshot bio health projects. >> Okay. >> And the amount of compute that they can throw at these projects is unrivaled. >> Yeah. >> Because you just even out the usage curve by anytime anything dips >> Yeah. >> like, you know, you just bring it to full throttle with your secret projects. >> Okay. >> So, you're talking like no one can compete with this. >> In terms of like designing new medicines and cures potentially for things. I get that. >> You know why this is? Because everyone else has to use the current model. >> Mhm. >> Right? They're the only ones that get to use their next gen models. And you know how good like the next gen models are, right? They're like 10 times better often times than the current models. So, there's no competition. There's zero competition. They can't compete with an Anthropic internal biohacking lab with having a billion dollars using the current best models out there. It doesn't work because Anthropic's model that they'll use in their biohacking lab is next gen, 10 times better than what you're using, and they have so much compute to use. They can just fill, you know, as they grow, they're using all their unused compute on these crazy projects. How are you going to beat that? That's like crazy. >> That's great for medicine, but how is it going to create manufacturing? >> No, so the same thing you do for humanoid robots. It's obvious. You do it for all this different stuff. It's not just bio labs. Yeah, you have a humanoid robot >> Design the new factory output. >> manufacturing lab. You have all these labs. It's like I'm telling you, man. It's like what OpenAI and >> Like it'll basically tell them It'll be the director. It'll be the director of creation of the factory in terms of like hire this person to do this, buy this piece of land, get this concrete order, do this all. Like it'll just do it all for them right away, so it's not like depending on a person to manage it. Is that kind of what you're thinking? >> No, so it works more in steps. So, you know, you've got to design like the actuators. You've got to design, you know, the hardware and all this stuff. But like what you're doing is and it might not be primetime right now, but maybe in a couple of years, basically OpenAI or Anthropic, whoever has the best frontier models, they can basically pour tens of billions of dollars, perhaps hundreds of billions of dollars of compute that is not the current best model out there publicly, but their next gen private models onto certain problems. And it's going to progressively >> It can make a better actuator than what Tesla has, for example, for what >> because it's so far so far advanced in the current models. But, it's going to progressively get more into the physical world year after year. Meaning, right now, predominantly digital, predominantly, like, you know, >> Mhm. >> you know, >> So, I don't see that crossover happen so easily. I think you see it as a flip that easily transitions, but >> Something is progressively going to happen, like, health is an interesting one because >> Yes. Health is interesting. But, I do think I still disagree, and I think that the manufacturing, the building of the factory, you know, the machines that make the machines, putting that in place, the logistics, setting up the power, you know just >> Yeah. >> There's so many things that Tesla has gotten so good at that I think is a very high bar, except for maybe Foxconn or something like that. You know, like, I just don't know of anyone else that can do what Tesla's doing in manufacturing. >> Yeah, but I I don't know. My hunch, I've talked to a few engineers. I don't know that much about it, but my hunch is engineers are using AI more than people realize right now, to the point where >> I believe that. >> it's like AI is just starting to take over even engineering jobs. What that means is it's like in two or three years, with not just the best publicly available Fable 7 model, >> Yeah. >> but we're talking about if Anthropic has a private Fable 8 or 9 or something crazy model, like, that's going to be able to design a machine that makes a machine, >> But, do you think >> and 3D print it. >> Do you think Tesla's version of their frontier model will be close enough, or do you think Fable 9 will always be 10 times ahead, 10 10 times better, you know, or do you think it's going to be able to even out? >> that's a good question because it's like And this is one of the reasons why I think SpaceX and Tesla need to emerge because Tesla like fine, Optimus hardware good you know, AI 5 6 7 8 good, the chip is good, the models are good. Yes, all that's important but I think Tesla needs to like I think Tesla SpaceX together if they can push the envelope and compete in the frontier model race to be as good as open AI and Anthropic. Now we're talking about the future of Optimus robot being like >> Yeah. >> pretty solid, like really really solid. >> Yeah. >> But if for example, there's a future where open AI or Anthropic like take a a big lead in AI >> Yeah. >> then everything is up then everything is shaky because >> Yeah. >> whichever company has a big lead in AI that company will have an unfair advantage in every single field basically out there. >> Temporarily until the others catch up, which could be >> Well >> 3 months or 6 months or >> the question too. That's the question too. It's like >> There could be a point where once you get enough of a big lead, you use that lead to just do this crazy reinforcement loop where you're just like >> Yeah. But how smart can a humanoid robot really get like? I mean, it gets super smart, but like if it's super smart, how much more useful is it? Like a humanoid robot with an IQ of 150 versus a humanoid robot with an IQ of 300. Like how much more useful can it be? >> it's it's no, it's tremendously useful. I would I'm talking about like like So, a smarter Okay so I'll give you an >> Mhm. Like there's a ton of really, really difficult Like humans do a lot of difficult things. >> like infinite demand for humanoids with 150 IQ and infinite demands for humanoids with 300 IQ, don't you think? >> No, no, no, no. It's like um um I'll give you an example like plumbers right? Like I don't think IQ uh >> I mean that's not a great metric. I'm just saying like in terms of competence like competence like whatever you're telling them to do. >> Like you For AI to get so good it could replace a planner plumber. See a plumber is so and this is just one job example like many, many jobs are are very complex, yeah. Um but to replace a plumber, a plumber does so many things, you know? A plumber can dig, cut the pipe, do copper piping, do fixtures, demo, install water heater, yeah, leak, do like troubleshoot, you know? Toilet, do whatever, unplug like it's We're talking about thousands and thousands of things, but each situation can be a little bit different, too. You have to improvise and do all this stuff. The combination that a plumber needs to have for example in terms of manual dexterity, which is kind of what Tesla's trying to get at with is this the mechanical ability to do the complex things that a human can do with their hands. So, that's one level, but the other level is the brain to be able to face thousands of plumbing related situations. Even gas lines, like plumbers do gas lines. They It's like tremendously install air conditioning, start a backup generator, install like, you know, like a a gas line to your sauna, like go up a you know, like >> Whatever. >> There's like there's so much involved. Like for Optimus to be able to do all of those things a plumber can do, that takes so much capability intelligence. We're talking about like You know. >> Experience. >> Like a huge amount. And that's the thing. If your robot, like for example, if you can use Let's say you're the you're the top frontier model. >> Yeah. >> You're ahead. And you have a private model that's like ahead of everyone else, much better. >> Yeah. >> You use that to give your humanoid robot the brains. And of course you have to distill it because it's a real-time running of that large language model, but you're able to do that where you you use your frontier models on these crazy big machines and you distill it where it gets put on the humanoid robot, but it's so good. It's just like so good. >> Plus the so good AI is probably going to invent new solutions that are more practical and easier than what traditional plumbing does for things, right? >> Exactly. So with that mind, you can even be a little bit less like like capable with dexterity as long as you have the super crazy brain, you know? You could do some workarounds, you know? And still fix stuff. >> Mhm. >> So that's what I'm saying. It's like if Tesla and SpaceX can't merge, Tesla Okay, here's the ideal situation in my opinion. Tesla and SpaceX merge, Elon goes crazy in turn good crazy, >> Yeah. >> pushes the envelope, gets the lead. >> Yeah. >> Right? At least >> But I don't even think they need the lead. They just need to be competitive with the lead. >> Yeah, okay. At least you know, neck to neck. >> Yeah. >> You know, don't let any of these companies get get >> more than 3 months ahead. >> Exactly, you know? Like it's too dangerous right? >> Yeah. >> Get with like neck to neck with the best companies. Now we're talking about Optimus as the next as crazy. This is the next big thing. But, if you don't have that, like sure, you could say, "Oh, yeah, we could use Grok's model. Oh, we could like whatever partner with" But, that's the thing, it's not the same. You don't use like a model and stick it in your humanoid robot to do stuff. No, you you you in a way you've got to create kind of like use your best next-gen private models that are a year or 2 years ahead of your public models to create the brain of your humanoid robot so that could be so stinking smart, better than anything else anyone can do. >> Yeah. >> Um So, anyways, I mean, that's my my take with the merger is that it's not optional. Like, Tesla and SpaceX really need to merge, you know? This is important for the future of you know >> the businesses. Yeah. I totally agree. I just think of it like in a simple way. I try to think of the way I think Elon thinks is like, what's the best outcome for humanity, you know, 20 years out? And is a future where Tesla and SpaceX remain separate going to create a better future 20 years out or is a future where Tesla and SpaceX merge soon 20 years from now going to be a better future? And I think definitely the latter. And so, and I agree. I mean, I think that's how he thinks about it, that's how I think about it. To me, it's imminent and uh I don't think you know, I understand how some Tesla shareholders, including me, you know, we've been holding the bag, so to say, for 5 or 6 years. The stock hasn't really gone anywhere. Um it had its moment in 2019, 2020. Um and then it's kind of been languishing since while the rest of the Nasdaq has zipped up, you know, crazy. And so, it's it's been frustrating for long-term Tesla shareholders, you know, that don't have exposure to SpaceX yet, for example. So, you know, I get that. Um but you got to ask yourself, you know, the long-term question and the Tesla like you may think that robo taxi is right about to be a massive rollout and then Optimus is around the corner and Tesla's stock is about to go up 5x or 10x, which I think it would, to be honest, in the next few years. We'll talk about positioning or whatever later, but um or you can talk about you can think about if Tesla and SpaceX merge right now with SpaceX already at 2 trillion and let's say it's had its moment and it's not going to on its own go up another two or three or four X for another few years until, you know, uh the AI bubble bursts and comes back to let, you know, be valued appropriately again. Let's say Let's say that's your mindset. And if they merge, then, you know, in that mindset, if they merge, then Tesla's, you know, robo taxi incredible success and Optimus unveiling in the next couple years will help SpaceX and Tesla combined go up three to five X as opposed to Tesla by itself going up five to 10x, you know? That's like the worst-case scenario you have to have in your head as a Tesla investor, you know? And that's still a good outcome, three to five X return on the stock, in my opinion. But that's like the pessimistic way to look at it. I don't think that's the way to look at it. I think you got to look like 10 years out, you know, like combined, the company's going to be amazing. And you got to realize like as a Tesla investor, maybe you haven't studied SpaceX all that much and maybe SpaceX is going to go up on its own three to five X from here to five or 10 trillion market cap in 2 years anyway, and maybe that's a huge boon for Tesla shareholders who merge with SpaceX now, you know? You could be wrong on that front. So, you got to take all those perspectives into consideration, I think, when you kind of are trying to weigh a merger or not. Because initially, I think a lot of people didn't really take all those perspectives. They kind of just thought, "Oh, well, we're about to have our moment with the RoboTax you roll out. Now we're going to merge with SpaceX first and save SpaceX." And now more people are starting to get on board with, "Oh, wait, SpaceX public company. I'm learning a lot more about it. It's actually legitimate fast-growing AI business, you know?" >> Uh-huh. Yeah, I I mean, the way I look at it is is I mean, for me it's it's the merger the whole question is is more like it's this idea that with SpaceX, so okay, with SpaceX they've got this So, the here's here's kind of like the bull case for I think SpaceX's AI kind of like ambitions is even though Grok right now is like is not as good as OpenAI or Anthropic's models. I mean, you can argue but it's obvious to me like the best Anthropic models are way better Grok has. >> Mhm. >> So, >> [clears throat] >> the thing that what what SpaceX has is they have this really interesting capability to build out compute or data centers in a way that no other company can and that's basically to launch them into space. And >> But also the massive ones, right? They can put it together. >> Yeah, but eventually like the chances are it's going to get harder and harder >> to build the big ones >> on Earth. >> Yeah, because there's limited number of like existing electricity infrastructure. >> And all the NIMBY stuff going on in politics. >> Yeah, so so there's going to there's going to be a a shortage of infrastructure of of data center compute needed to run AI. >> Yeah. >> And so SpaceX has this really interesting unique ability to shoot things into space. >> a glitch in the economy. >> Exactly. And they're like the only ones that could do it at that the scale that they're they're a Starship is going to be able to do it at. >> Yeah. >> So in a way >> next Starship launch is so important to watch, right? Make sure >> Uh-huh. >> And then you like go up and down. >> yeah, yeah. Definitely. And so what it does is it kind of even the field now where as with right now with Anthropic and OpenAI they kind of have an unfair advantage because they started early, they recruited the best AI talent. >> Yeah. >> They have the most like data, they have the most users. Um like a lot of the stuff you see is like they still have a ton of stuff that they're working on that we haven't seen yet. There's just um it's it's in a lot of ways like xAI by starting with like live in baits. >> Yeah. >> And also it took I think a little bit of time for Elon to kind of like get his bearings and focus right. Like I I feel like with xAI like their early focus was a little bit distracted. Like it wasn't really like I think it I mean they could be wrong, but I feel like xAI focused too much on like kind of becoming truth seeking without >> first solving like the right problems first. Like solve code. You know? >> Yeah. >> Solve code better than anything else. Or like the the lame avatar whatever like things that they're doing and some other stuff. It's like they're just like even Grok imagine with the whole images and video right now. >> Yeah. >> I'm not a big fan of it. I I I personally think like yeah, like Anthropic when you look at Anthropic's focus, they're not doing a image model or video model or whatever whatever. They're just like laser-focused on like on the >> coding >> coding and the model and eventually right now like code code is creating all the pictures now and it's going to create like the videos, too. >> Yeah. >> It's like you don't need like a image diffusion model and all this other stuff. >> Yeah. >> So I kind of anyways OpenAI and Anthropic kind of have the best talent. They have the most momentum. They have the most users. It's kind of an unfair race in some ways right now. But SpaceX AI or SpaceX SASI, which is crazy to me all the different name changes. But it could even things out by through the infrastructure, the hardware side, like creating these like data centers that no one else can compete with. And SpaceX can use that to try to catch up and hopefully you know, surpass Anthropic and OpenAI. That's the idea, >> Yeah. >> right? That the compute is going to be the secret weapon for SpaceX, right? >> Yeah. >> To compete and to to over to proceed um uh Anthropic and OpenAI. And so when you look at it like that, I'm like, you know, that's believable. To me it's believable that SpaceX that has a chance to use their extra compute, these new data centers to leverage that to have a good chance of being at the front or neck to neck at least. >> Mhm. >> So, when you have that situation and you're left with Tesla and their future relying on let's say FSD and Optimus, but increasingly it's going to be on Optimus, but you don't have a real brain for Optimus. >> Mhm. >> It's like you need you need that SpaceX partnership. You need access to the smartest intelligence. >> Mhm. >> Or else you're at risk, you know, of being eaten alive by the smart smartest intelligence. And so to me it's it is existential in a lot of ways. Like Tesla does need to >> Unless they own the whole whole stack >> Oh, no way. Exactly. >> can if things work out versus if they didn't merge, Tesla's Optimus might they might manufacture, but they might have to out they might end up outsourcing some of the brains to someone else like Anthropic or >> No, I don't think so. >> or you think that Anthropic and OpenAI would build their own version of a humanoid? >> yeah. They're all going to go after the human humanoids. It's too big of a market. Yeah, it's too big. >> Okay. >> Yeah, yeah. This isn't this isn't this market the humanoid market isn't like um it isn't optional. There there aren't that many you know, huge markets. >> But I still think that advantage Tesla will always have for a few years at least at least like three to five years lead is on manufacturing capability for massive manufacturing. Like I'm talking like hundreds of thousands of Optimus. They're almost like ready to do that now almost. You know, like I would think any other company that wants to get in the space of manufacturing humanoids except for China, any company in the US that wants to get in that space it's going to take them three to five years at least to build that capacity from here. You don't think so? >> Yeah. Yeah, no, yeah. Sure, maybe Tesla has the lead, but if they again, it's like if they don't have the smartest brain, you know, then what's going to happen in 5 years, you know? >> Mhm. >> Like when people catch up. Here's the other thing. It's like when you look at manufacturing, sure there's a short-term, you know, advantage you have of all the manufacturing like machines, engineering, the processes, the supply chain, like all that stuff. But if you think of about what manufacturing it is, it really is like this is my opinion. It's it's intelligence >> Mhm. >> applied. Meaning it's intelligence that has designed the parts, right? To come together [clears throat] into a complex product. And this intelligence has designed the production line, this intelligence has designed the custom machines and the parts of those custom machines, and this intelligence has sourced, you know, the right stuff everywhere. It's basically intelligence that has, you know, that to manufacture. And that's why, you know, when you have the the best frontier models, like we're going to see this transition into the physical world. AI is going to s- Like right now people are so used to like the internet being internet, digital. Like you have a digital world, right? We're so used to AI being like a chat or something, right? Or picture or video or something, yeah, software, right? But it's like it it's only limited that because intelligence was limited. It just wasn't that [clears throat] smart. >> Mhm. >> But once you have And that's the thing. In my opinion, intelligence crossed this crazy line like about maybe like 8 weeks ago. That's my rough estimate. It crossed the line of, I think, real AGI. Like I think the line was like stable five. Table five and now Opus 5.5 also is above that line. It's this line where it's like >> So basically if you go to like table five, like applying your thesis here is you go to you're using this new AGI or maybe the model above that that only Anthropic has internally right now and you say, "Hey, I got a million bucks start or 10 million bucks. Help me build the fastest humanoid robot making mini factory where you can start you know, tell me what to order order it all do it all yourself." You think it could basically take charge of the project from there and get it going right away? >> Not right now. Yeah, even the best ones can't do it right now. >> But very soon something like that. >> Well, okay. What's very soon? So okay. >> Like a year from now maybe. I don't know. >> I don't know about a year but but but Like I'm not like the progression of how much the intelligent is going up is so >> But that's where we're going to go. >> crazy. Eventually, yes. I'm not sure if it's two or five years. It's probably not 10 years though, you know? It's like it's it's improving so fast but I mean it's You're going to see as intelligence has crossed over a certain point and it's getting better and better and better. Now that intelligence is going to be used to do stuff that in the physical world that really requires a lot of intelligence, right? And so you're going to have like AI not being hooked up with 3D printing, with making parts, with assembling parts, with with all this stuff. That's the next kind of big step for AI. It's a crossover and it's like >> Yeah. >> And the thing is the crossover is probably going to happen the most like the private models are probably like so like you know, every model is a big like not just the point but like there's like these step change generation model changes, right? >> Yeah. >> Those are the big you know differentiators. Like there's like sometimes five times 10 times better in a lot of ways, right? >> Yeah. >> But once you have a few of those they compound, you know? And you have something that's like a thousand times better in like two three years. >> Yeah. >> And it's like almost unmanageable right now what could be a thousand times better than what we already have right now. >> But that we probably didn't imagine that two years ago. >> Yeah, exactly. Like the stuff that AI is doing right now it's like I'm coding different projects. Before I would have to like debug and look at the lines of code like two years ago. Now it's like coding projects that I didn't think was going to be possible for another three years. >> Yeah, like right now we're at that level of that movie Her, you know, where they're like the agent's basically taking charge of your life doing everything for you if you wanted to. >> Yeah. And so anyways, I mean all this is to say that I think like so like we have we're in this paradigm we've been been in paradigm where the physical world and software world are so differentiated, right? And they don't cross over that much and so therefore Tesla's a manufacturing a hardware company. They're kind of isolated from the software but now there's going to be this this crossover and the crossover most likely is going to happen in the private labs and private projects using the next gen models that aren't publicly accessible where these the top frontier lab or labs basically choose the biggest bang. >> Yeah. >> You know? >> Yeah. >> markets. >> Yeah. >> And that's what they focus on. >> Yeah. Yeah, it's wild times. I mean, but do you think that those top Do you think it's a moving thing always. It's not like there's not some destination I feel like of intelligence. It's always going to get better and better. It's always going to be increasing from here, right? And so but do you think once you cross a certain threshold where like we're describing now that the time it takes for SpaceX SI to get to match it with Anthropic, wherever Anthropic is, more than like a few months or >> Yeah, no. Yeah, it's possible that the once you get a lead, at a certain point it's possible >> is so fast even further somehow >> Exactly. Where just is years >> They could They could do something to prevent the others from getting to that same spot. >> I wouldn't even say prevent, but it would take them perhaps a lot longer because you're using your best models to create your next model, right? And that's the whole thing that's happening right now is the your next model is being created by like the next model. Like it's just >> Okay. >> constantly like self-improving, and >> [clears throat] >> if you can get like 6 to 12 months lead and improve keep on doing that, you could perhaps there's a theoretical possibility that you can actually increase your lead further. >> Beyond 6 to 12 months to be longer. >> It would much gets gets longer and longer with each new model that your models producing. Like it just it's like a flywheel, right? I mean, there's the other side where people you could say, "Oh, you know, other people can catch up, etc." >> Yeah. >> But there comes to a a certain point where like right now a lot of the AI like improvements are usually because AI researchers, different teams around the world, there's like, you know, they put out papers, they come up with new techniques, and then people are taking those techniques and they improve AI a little bit, and they they use 100 different, you know, techniques and stuff to improve AI. Like the whole world is like helping AI get better in a in a lot of ways. But there's going to be a time, and we're probably pretty close to it, where most of the improvements, if not 99% of improvements, are actually not human-discovered or human led improvements anymore. >> Yeah, they're AI discovered. >> Exactly. AI discovers like and how to get it better. >> Yeah. >> If that happens, it it becomes theoretically more possible that and perhaps even likely in some ways that if you get a big enough lead that that lead actually increases. >> I have a hard time I'm just having a hard time grasping that, but maybe you're right. I'm I'm trying to I'm trying to envision I just can't imagine that. >> Uh >> Um unless it's doing something actively to prevent the competitors from getting to that same spot. I I I don't >> Well, okay, think of it more like you you lose control of the AI in some ways. The AI is just >> Perpetually recurring. >> don't you don't understand what it's doing. >> And you don't the others don't lose control of theirs 6 months later in the same way you're saying and >> Yeah, I know. Like okay, so for example, let's say you get a model that creates a thousand improvements, right? But you don't really understand what those improvements are. Like it's too too too >> beyond your human >> for brain level, right? To understand the thousand improvement. And you go, okay, and then you go, okay, let's put them make the improvements, right? >> Yeah. >> And then a week later it says, okay, the model's >> that much better, right? And it says, hey, I found another 2,000 improvements to make this model that much better. >> Yeah. >> And then >> 2 million improvements the next >> then you say, okay, do it, you know? And they come back and they go, okay, I found another and then and then 12 hours later they go, you know what? And you're just like, what the hell? >> And yeah, and the other >> explosion, yeah. >> But you don't even understand the improvements, you know what I'm saying? >> Yeah. >> All you know is it's getting better and it's making more improvements. That's what I'm saying. It's like if each company has those models, don't you think the end results would be different? >> I mean, I think >> not going to be same, right? >> I think the time wouldn't get longer between the competition, but I think the gap could be way differentiated like you said. Like I think the intelligence explosion could make one advanced model suddenly go up to be a thousand times smarter than the next one, but that thousand times difference, as long as it's playing fairly by the rules and not doing something actively to prevent the other, I think the others would catch up to where they are six months later, but then by that point in time that one will be a 10,000 times better, but they'll still catch up to where they were, but the other one will always be accelerating faster to be better and better, you know what I mean? >> Yeah, but I mean Yeah, I mean but if you think about the the the idea of compound improvements, yeah, if you have something that's like improving itself and then it improves on top of itself >> Yeah. >> and it improves on top of itself and you compound it just faster over time, in five years it's like you can't even it's like you can't even see the first place and second place. >> you can't see it. The distance between the capability could be so far. You're I see that definitely. I see what you're saying there and that makes sense to me. >> to me this is the crucial point. Tesla needs to join SpaceX. They need to >> that? >> They need to ensure that that SpaceX and Tesla become that first place. And the here's another thing, Tesla helps SpaceX. Cuz what you do is when Tesla joins and they become one company Tesla all the energy of Tesla needs to go into also this goal of securing first place. >> Mhm. >> Like this is the future. Like you got to secure first place, you know? So you get all of the energy of Tesla with SpaceX yeah, to secure first place. >> Yeah. Yeah. >> Yeah, that's what you need, you know what I mean? >> Yeah, no, I think combined I would put the probability is pretty high that that happens when they combine, if they if they combine, you know? I think the probability is above 50% in my mind. And it's not it's not 70 or 80%, but it's I think it's above 50% if they combine to be able to get to that first place within a few years. >> Yeah, I mean I hope so. I I hope they combine. I hope you know, they they really succeed with the >> If you take the context of like everything that's come up to this point with SpaceX being the attractor of the top talent at Tesla being the top talent attraction for all the top engineering students and still the you know, some people might want to go work with Anthropic or OpenAI, but I think more and more people are following seeing Elon as a better leader for the future than you know, Dario or Sam Altman. I mean, I think the true top engineers you know, will probably lean more towards following Elon to work for his companies if given the opportunity to pick between the three. I don't know, what do you think? >> Yeah, I don't think it's that straightforward. >> No? >> Yeah, >> Like if you take a hundred of the top engineers you don't think it'll be more than like 50 that would pick Tesla SpaceX AI versus OpenAI and Anthropic? >> So, we're talking about it's a different type of engineer we're talking about. So, you're talking like what's most important in the AI wars is not like a typical engineer. It's like it's like a it's it's an AI like engineer who's like top of like you know, let's say out of the top 1,000 AI engineers minds. These guys are people who understand it deeper who are in touch with the so they kind of push the envelope to create new things. Like there's probably like, you know, a good like 50 top minds. >> Yeah. >> That are really like really really valuable. >> Mhm. >> It's like like >> They're probably doing mostly their own thing. >> No, no. >> Or you think they choose to go work for OpenAI or Anthropic? >> and AI have the vast majority of those top 50. >> Okay. >> Like you might you have some spread out, like sure, you know, maybe Elon has some, but and he might have some in other companies. >> Ilya, maybe Ilya in a couple. >> But but >> Google. >> Out of the top 50, mine and Google is sure, you know? Like, you're looking at like probably 15 to 20 a piece at at Anthropic and and OpenAI, and the rest is just scattered, you know? And and then that's the first tier, right? And then you have a second tier of the next, let's say, 200 minds. And then again, OpenAI and Anthropic have probably the best talent, followed by probably Google. >> those top ones are at OpenAI or Anthropic? Do you think they're >> It's a It's a combination of of like OpenAI and Anthropic are in the lead. >> So, they want to be at the one who develops it first. >> working on the most interesting things, the most capable stuff, right? You get the most compute You get the most access to the best models, like, you know, it's like, why is Karpathy >> at Anthropic? >> Yeah. >> You know, it's because it's like Anthropic is at the cutting edge. He gets to be involved with the most cutting-edge stuff, right? >> Yeah. >> That's why he's there. >> Yeah. >> The other thing is, it's like, um, that's where the top talent go. It's like you're recognized, like, that's where you go. It's It's kind of like the Harvard, you know? >> Yeah. >> Stanford of whatever AI companies. It's like OpenAI and Anthropic, you know? >> Mhm. >> Like, next up is maybe Gemini, and then like other companies, like Never. And so I mean Yeah, I mean, hopefully I mean, ideally, I mean, that's what also what makes it difficult for Grok and for SpaceX uh uh to to overtake Anthropic and OpenAI is the talent war. It's because it's like OpenAI and Anthropic have like given away so much equity in terms of like you know? So many people are billionaires, like, you know, >> Yeah. >> Anthropic Can Open the Eye or or multi multi, you know, billionaires cuz equity. And so, it's like Yeah, it's a it's not like a a simple like, you know, done deal for SpaceX. Like, they got their work cut out for them. They they're the underdog in a lot of ways. >> Yeah. Yeah. >> Um yeah. [clears throat] >> Yeah, crazy. I mean, I think uh these next 2 years, 1 or 2 years are going to be pivotal and uh if SpaceX and Tesla do merge, which, you know, I still think it's going to likely happen the next announcement latest would be after S&P inclusion of SpaceX, which will be after all the lockups are done. I think that'd be the latest, like, a month or two after that. Um but I think sooner or more likely it would be some kind of announcement to attempt to merge the companies. What do you think of What do you think the odds are when? >> I don't know. I mean, I just hope they merge. I just hope it get over quickly as possible and we move on, you know, as a joint company. Call it X and forget all these like crazy names, man. >> Yeah. >> And just uh go for first place, you know? >> Yeah, you had been mentioned a long time You were like one of the first people I saw publicly talking about the X holding company idea, you know, >> Yeah. >> years ago and it just makes sense. >> It's simple. You just merge the companies and you got to go for first place. There's too much at stake. If you're second or third place, you're just It's just too much. Too much exposure. Too much risk. Too much >> Yeah. >> Yeah, it's not good. >> Not good. Not good. Yeah. So, what else is uh on your mind? Uh you know, the SpaceX uh special purpose vehicles have gone, you know, I know many of the people who watch this or are in this or in the SpaceX special purpose vehicle vehicles getting their lockups, you know, shares and it's been a crazy uh time for for those of us, me included. >> you been getting your distributions on your SpaceX shares? >> Yes, I have. Um and uh yeah, it's been, you know, life changes for me again, you know, uh so it's all in my IRA. Um so I've had some interesting decisions uh I've made, you know, I've there's something called like a uh 72t plan with the IRS if you want to extract money from a traditional IRA without the 10% penalty, you know, so I don't have an an income job anymore. I have a small fund I manage, but doesn't really give me income until or unless the fund really outperforms. Um speaking of which, that's what That's what I do most of my aggressive investing trading in these days is the fund where I have a few investors, but it's closed. No, there's no new investors anytime soon. It's just been surviving with this Tesla kind of language machine where it is for a long time. I'm just trying to wait it out, surviving until I think the stock really explodes, which, you know, not investment advice whatsoever, but I told a few I've been telling some friends and family like this is the time, I think, that it might be very uh good to uh for me, the risk-reward seems very appealing for like the long-dated December 2028 call options for Tesla. I'm talking like 700, 800 strike, you know, double the price plus, you know, that's 2 years and 2 months out, let's say. That's a long time for the robotaxi to be rolled out, for Optimus to be unveiled, and kind of like grasped by the investment community. So I think, you know, to me that's that's what I like and I and uh and I also um think if there's a SpaceX Tesla merger, which is something that, you know, I I thought about a lot and I think it's likely to happen. I think it's likely to be an all-stock merger by SpaceX. And in that case, you'd have, you know, Tesla shares convert into some ratio of of SpaceX shares and then the underlying call option contract should also convert to a similar ratio, you know, so um to me I know there's uncertainty a little bit with a merger and how it's done, but in my mind, I've tried to work it out and thought about it and I really like my positioning in my more aggressive, you know, this is money I can afford to lose or I don't I think could go to zero easily, money in my fund, for example. Um but for risk reward, you know, um probabilities and potential outcomes, uh 10x or more return, I think is is very possible, um more than 10% chance, much more in that type of situation with like a 2028 December, you know, 700 800 call option. That's that's my you know, my thing right now. You [clears throat] have For the money I'm prepared to lose. >> Yeah, I could see I mean a lot of that to me will depend on how fast rev- robotaxi can scale. Yeah, cuz if it has trouble scaling >> chip will be massive by then, right? I mean it's massively produced in all the cyber cabs by then, right? Version 16 will probably be going on. >> Yeah, I mean it's got to in a way start to show up on their profit loss, you know, on their income statement. Yeah, they got to be able to >> To me, riding the cyber cabs cab today is what reinforced me on this belief and made me like talk a little more publicly with you about it here because I I mean I've been thinking about this in in in dipping my toes in that anyway for a while, but now I'm like more confident after the cyber cab today because it reminds me so much of uh back in 2012 early 2013 when the first Model S reviews were coming in and they were just getting the first deliveries out and we were getting some of the first Model S's I remember and like it was like amazing. I was like, "Wow, this is incredible. Why isn't the stock reacting? This is the best thing ever." You know, and it took a while until, you know, the stock finally really popped. I mean, I know it was a different situation with shorts and so forth, but there was a whole new re-rating of the company after it became widely accepted by the investor community, "Hey, this company builds a very compelling electric car and it's here to stay, you know?" So, I think the wider investment community is going to be coming to a similar realization and re-rating of Tesla about the robo-taxi and it's going to overshoot evaluation could be just with a massive robo-taxi because people will still have their eyes on the prize with Optimus, assuming Optimus is not unveiled before robo-taxi real I mean, Optimus will be unveiled, but before Optimus is deployed massively, before the robo-taxi is wide-scale, you know, which I think will be by the end of next year at the latest. So, I I do think that this is a good time for Tesla stock. We You know, good things come to those who wait, you know, and it's been a six painful years for Tesla investors here, you know, and uh I think we're going to have some nice rewards, hopefully. >> Yeah. I mean, one kind of thing that is potentially optimistic or bullish is because it really does feel like FSD is getting so good. Yeah. I feel like um with AI 5 chip coming out in new cars. >> Yeah. >> I just feel like those AI 5 V15 new cars are just going to be like right now, I mean, it's great, too. >> Yeah. >> Yeah, I've four, but like >> It's not like you feel like you're going to be able to go to sleep in this car with this version. >> Yeah, but I just feel like the the value that the car brings compared to other cars increases. Right? As Tesla gets nails FSD better or more confidently, right? And so, I feel like they could sell potentially more cars that could really potentially increase their revenue and their operating margin, profits, all that stuff. Like that's the stuff that really like moves the stock in my opinion. It's not so much like, oh, a new product reveal, oh, like, you know, whatever, whatever. It's like more like something happens where like there's an unexpected move of demand that increases, you know, volume, which increases margin, which increases income. And then they re-rate everything, right? And so, I mean, I don't know if though it's like the the the the the the struggle that I think Tesla stock has is they just don't have enough models. >> Mhm. >> You know, it's like, what the heck? Tesla has like, what? Two models right now? Basically. The Cybertruck doesn't even sell much. It's basically the Model Y and the Model 3. >> Yeah. >> But think about it. How old is Tesla as a company? It's not like a 5-year-old company. >> Yeah. >> You know, Tesla's been around for like, what? 2006 or whatever? >> Well, don't you remember those early earnings calls where Elon was like, I don't think it's ridiculous to have so many models. I think a company just needs like four or five [clears throat] models. Do you remember those early earnings calls? >> Yeah, yeah. >> I mean, that seems to be a core belief unless he changes his mind since. >> Yeah, no. The thing is the the Model 3 and the Model Y are just limited in its appeal >> Yeah. >> to certain demographic, certain needs. And we're like the company's 20 years old and you got two models basically you're selling. >> It's like it makes no sense. Sure, there's the semi trying to roll out. Sure, it's like cyber cab could be a model, but in terms of selling to the consumer, >> Yeah. >> that's the big problem. You see, if Tesla had like six models right now >> Yeah. >> and AI 5 is coming out, B15's coming out, you have increased demand across all six models that are selling well, you know, demand and volume goes up 10 15% next year at least, you know, all this stuff, profits go up and all this stuff. Now we're talking about the stock moving right? >> Yeah. >> But what do you got? You got two models. It's like I It's like that's the problem, man. >> [laughter] >> I mean, this is where you and I disagree, I would say. I mean, I've seen your case for it and uh I think that argument made sense maybe 5 years ago if we knew that full self-driving rollout wasn't going to be, you know, in the next 5 years. But now that I think we really are convinced the full unsupervised full self-driving rollout is like a year away and like robo-taxi, cyber cabs are going to be deployed everywhere in a year, within a year or two. I think it's too late to have It doesn't make sense in my mind to have a lot of different models. I think the people will you know, that scratch or the the uh how they're conditioned to look at lots of different models for every It will be totally irrelevant once they're like, "Oh, I want the car that drives me everywhere. I don't want to have to drive a car." Whatever. Just get me that car, you know. I don't even care what the model is long as it fits them and their family in it, you know, and it's comfortable, which it will be. So, I don't know. I think I I That's what I think. I think Elon, similar to marketing and advertising, it's almost too late at this point to try to do something like that um because they're on the precipice of this AI Let's say it's the AI 5 chip limiting holding them back for full self-driving 15, which is what's holding them back for the massive deployment of unsupervised FSD and going all over the place. So, if we're on the precipice of this massive unsupervised full self-driving roll out, that to me is all that matters, and they could just sell all Model 3s or all Model Ys and it'd be infinite demand. You know, because no one else can do this, you know, for years. >> I I I know like there's people like Gary Black who like everyone's doing robot No. >> Tesla's the only one in this game at this level, you know, everyone else is years behind trying to figure out lidar and like >> Yeah. >> all this other stuff, you know, and like Waymo can't even produce more than like 100 cars a month of this. I don't know if they're using like third-party manufacturing. It's crazy. >> Yeah. >> So, and it's like a $200,000 for a Waymo car to be produced, you know. So, it's just it To me, Elon has changed the game in a way where I think like at the limit, he could just produce all Model 3s or all Model Ys. It doesn't even need a Cybertruck because if they have the unsupervised full self-driving, that's that's all that matters to me. >> Yeah, I I mean I don't know. I feel like if they had a few extra models, they could sell double. >> like extra models, trust me. I think it'd be cool. I think it'd be nice. >> No, they're good models. >> Yeah, yeah. I think it'd be cool. >> Like, you know, for example, >> I personally agree with you that I would like I prefer that, but I just don't know if Elon's thinking about it. >> Yeah, no, but I'm saying if they sell double, you know? >> Yeah. >> Like, have like a you know, like a boxier Model Y, you know? >> Yeah. >> Have like a >> Yeah. >> like a boxier Model 3, you know? Have a larger, you know, SUV, like there's like different things, but but I like I did this this uh podcast with Farzad on his channel. >> Mhm. >> I was trying to explain how I think Elon made a choice to not do the models >> Yeah. >> because he wanted to go all in on driverless >> Yeah. >> I think he's going to create a whole new lineup of new models based off of complete driverless. >> Yeah. >> Meaning no steering wheel, no nothing. It's like the Cybertruck is the first. >> Mhm. >> But >> I think the Robo Van or what are they called? The The >> Robo Van. Robo Robo Van. >> Uh >> I'm hoping they they give us a ride in that. >> Robo Van. >> Robo cuz like So so there's a couple clues that are that they're close to releasing it. >> Yeah. >> Like >> That'd be awesome if we get a ride. Like they should take that from the maybe where we're going now to this hotel that the hotel meet at right now. I wonder if they'll have that there and we'll ride that to the destination. >> That would be >> look on X to see if people are reporting that already. Who knows? >> I don't think it's even started this year. >> No, yeah. >> But yeah, like on their website it said the Robo Van was coming soon, but I talked to this engineer at the the Robo Tech the Cybertruck event. >> Mhm. >> And I'm like, you know the I mean what are you what are you guys working on, you know? I'm like you wouldn't tell me. So I'm like yeah, you guys started the Cybertruck like what two two and a half three years ago? I'm like okay. But also didn't you guys you know reveal the Robo Van the same time as the Cybertruck? I'm like you know doesn't take you about three years? Like shouldn't you guys be wrapping up on the Robo Van? Is that your next project? And he kind of smirked. >> Yeah. >> And he's like like he can't tell. >> Yeah. >> But I could totally tell like they've been working on the Robo Van and it's like I think that's their main project they're trying to wrap up. >> Okay. >> So my hunch is >> That's going to be so cool. >> Within the next six months we see you know, the Robo Van. >> think you'll be able to buy and own it yourself or do you think you're just going to be exclusively like you just you get it on the app and it comes and picks you up and takes >> I don't know. I don't know how they're going to release that. >> Yeah. >> It's like what do you do? Like >> a party bus or something. >> Yeah, it's like >> [laughter] >> Do you just have these things running around Austin? Like, you know, like you could do a cyber cab for two people or the Robo Van. >> Yeah. >> I don't understand. >> Like cities have instead of city buses, they can use that. >> Yeah, I understand. But anyways, my hypothesis is Tesla is I hope they're working on already. But >> Some other models of driverless >> of driverless cars, you know? Like across the board. And that will be the big revenue driver going forward. But they need to you know either they announce it. It's like the typical pattern is you work on it for 6 to 12 months, get the prototype ready. Have an event and show it, right? >> Yeah. >> And then you've got maybe like, you know, 2 years, 2 and 1/2 years where you you build out everything, you know, the machines, the production stuff, all that. >> Yeah. >> And you're >> I think that would have been great 5 years ago. I just think it's kind of late now for that. >> No, this is for the new driverless >> Yeah, I know. I mean, the new models, they could have been like, "Oh, we're going to have these new driverless models other than the cyber cab, these types of things." Maybe 2 years ago when they first showed the cyber truck. >> Yeah. >> But I just feel like now it's like there's just going to be infinite demand. Why start a new project when there's infinite demand for years to come on whatever they sell now? >> No, I don't I don't buy the infinite demand. >> No, you don't think that provides full self-driving? >> fallacy. It's like you don't have infinite demand. Like, for example >> But if you can click a button and put your car into that Robo Taxi network to make money for you, you don't think that's having that product to be infinite demand for you? >> Yeah, and then there's no there's no infinite demand. It's there's only infinite demand if the product's like free, you know? >> Yeah. >> Like you would have so much greater demand if it was a driverless car that was like $25,000, you know? >> Mhm. >> Like that would be like double, triple, five times >> Wasn't the Model 3 sort of like that? The Model 3 is like 30,000 or something, isn't it? >> Yeah, it's like 36, 37. It's like a smaller car. If you have like you know, a good like Model Y >> But $25,000 5 years ago was like $35,000 today after inflation. It's sort of there. >> Yeah, but you have to like that's the Yeah, I mean that's the other thing. Right now you have to You still compare people still compare Tesla cars with other cars because it's a so-called car. >> Yeah. >> You know, I mean I I get what you're saying cuz hey, if it's like a not a drivable car meaning like you don't drive anymore. >> Yeah. >> Then you no longer compare it with other cars. Therefore, you have the only you're the only that type of product in the market, right? So therefore, the demand is infinite because people But I don't think the demand is infinite. I think the demand is great, but it's depends on the price. >> What about this? What if the car comes with an Optimus robot and the robot and the car can be deployed by you to go do deliveries, you know, DoorDash driving stuff for you to make money for you 20 hours a day? >> I mean, if it could be free, you know. >> Yeah. >> If it could be >> Like maybe you deploy it for Maybe yeah, you deploy it for free and Tesla keeps half the revenue generates and you keep half the revenue. I don't know. >> Did you Did you Did you see Elon's tweet about just saying maybe we should make a competitor to Amazon? >> Yeah. >> Did you see that? >> Yeah, yeah, yeah, yeah, yeah. He's just threatening Bezos there. Bezos >> I [laughter] mean, it's it's like kind of like it's like is he serious or not serious? >> Yeah. >> Like in some ways it's interesting cuz Tesla has some pieces. >> Yeah. >> Cuz you have like perhaps Optimus robot, cyber cab, Tesla semi. >> Yeah. You can do all the delivery logistics stuff. >> Yeah, but then Amazon is so ahead with like so many things. >> They've just got the first mover like it's going to take a lot of differentiation to get someone to stop using Amazon to do something else instead too. Especially Amazon's getting those drones to deliver things now. I mean, it's been going on for a while, but >> have like a 3-hour delivery where you live? >> No. >> No, it's like crazy >> been same day. You have 3-hour delivery? I used to use DoorDash instead, and DoorDash delivers in an hour and a half, which is probably >> not like every day and every item, but it's like it's a game changer. >> Yeah, DoorDash has been a game changer since, you know, that's the best thing about COVID is it shepherded in like DoorDash for us for we use that we're probably like the top customer of DoorDash DoorDash as a family we order so much food from there. >> Yeah. Do you ever feel though like it's too much? Like you're relying just on like >> technology and everything? >> Yeah, I feel like I should get a tent and go somewhere out back and like hunt for my own food and like cook >> Yeah, cuz I >> with a fire. >> It's funny like this past summer we took like a 2 and 1/2 week trip to Colorado in our RV. >> Yeah. >> And like no DoorDash, no Amazon, no nothing. >> Oh, yeah. Wow. >> Yeah. And like we came back and my kids were like were like wait, you know, there's a downside to to Amazon and DoorDash and just button button click click click click click. >> Yeah. >> Just wait for it. It just there's something like I don't know. It's hard to pinpoint it, but it is so convenient that it's almost like it's almost weakening in a way. >> Yeah. >> Like or >> You're so dependent on it. >> Yeah, yeah. It's >> for a living. >> I don't I don't I don't know I don't I just don't know how to process it cuz it's getting to a point where it's so ridiculous even for our family sometimes. It's just like multiple Amazon orders in a day, like food, groceries, like this, this, this, and you're like like >> I hate having to use all these apps in my little square phone all the time. If I could just like talk to my watch or something be like, "Hey, order me this from DoorDash." And it does it without me having to click on things and sign into things, it would be so much better. I feel like we're close to that, too, with like Crock-Pot, for example. >> It's true. Yeah. >> Yeah, have you been using Rockbot or mostly Claude or both? >> No, yeah, I've been on uh Claude and like I go to the best models and I just go crazy. >> Yeah. >> On the best models cuz a lot of stuff I do I am trying to push the models to do stuff that is like really, really hard to do. >> to push the bounds of what what it can do. >> Yeah. >> Yeah, so some of the stuff you post is pretty cool. >> Yeah, so >> Example things you've done. >> Yeah, so um Yeah, but there's definitely been a shift. Like for every model that's been out there I test it on some like uh game my kids like. >> Yeah. >> So like my 11-year-old came up with this long prompt about like a Sonic-style game >> Okay. >> these crazy levels and he detailed every level and >> Wow. >> all the bosses and all the dynamics and it's like this super long prompt. >> Yeah. >> And so I give it to every AI to see what it could do. And every AI [clears throat] just fails, you know? >> Yeah. >> Like it does it kind of where you kind of play but it's just like so bad. >> Yeah. What about the latest Claude? Does it >> So something happened to the latest like uh uh Fable 5.1 and Opus 5.1 I It was actually Opus 5.5 that actually did it but >> Yeah. >> it um it created the app in 10 hours or 11 hours or the game. >> Okay. >> And >> How much did it cost in tokens? >> Barely anything cuz I'm on like their max plan and they're pretty generous. >> Okay. >> So maybe like 20% of my usage. >> Wow. >> But like it's hard to believe but this game came out to be like better than the real Sonic game. >> Wow. >> I I couldn't believe my eyes. I'm like what in the world are you talking about? Like nothing was even close before. >> Wow. >> And then all of a sudden it was like it became so good and my kids are like like because they play every single game from every AI that's my cue. >> yeah. >> And they couldn't believe it because it was like it was like indistinguishable in terms of graphics and like the mechanics were so good. >> Yeah. >> But it was like you know the prompt that my son gave with all the different stuff and it had everything that my son had dictated told it to do. >> Yeah. >> And it was like flawless like it was just basically like so stinking good. Like you could release it as a video game as is. >> Wow. >> And I was like my jaw dropped because I didn't expect that for another two or three years. >> Yeah. >> You know there's like cuz the progress was like suck suck suck suck suck. >> Was there a downside or was it just amazing? Was there a downside? >> just me it was just just crazy. >> I thought you were going to say but >> No it was just >> No it was all good. >> just like >> Okay. >> It just blew my mind. >> Wow I got to start using that latest Claude. I haven't signed up for their maximum plan. Uh what is what does it cost the max like 300 a month >> Uh 200 yeah. >> Okay I have the latest Brock stuff but I just have the 20 month Claude thing I've been using but I should sign up for the >> Yeah. Like my suggestion right now because probably in like one day or one week it'll be different but Claude like uh like for most stuff Opus 5.5 >> Yeah. >> is so stinking good. >> Yeah. >> Like they will is a high class but you know it's good and it just takes longer more more usage like all that stuff. I do I use it for some stuff as a super complex but they Opus 5.5 can do like just mind-blowing things. Um it creates like these like um animations >> Okay. >> so good. >> Yeah. >> It's like tell it to do like a two or three minute animation or even like a cartoon on like something like teaching you something something and it'll create like a whole like 2-3 minute cartoon animation, like whatever you want it. >> Yeah. As good as you can expect As good as you can expect or better than you expect. >> Yeah. Yeah. Yeah, just so good. Um >> That's so cool. >> Yeah. Yeah. >> Yeah. No, we've been playing I My one of my kids is really into using the AI stuff. He's been doing a lot of cool things with it. He built like this website He's just all into it and he started using the Hermes and he was going to start a business of showing other kids and families how to use Herms or Hermes or whatever and then and then Grok Bot came out and he was like, "Oh, I no longer need this. This is even better." Now he's like all into Grok Bot and like building stuff and >> Yeah. >> It's pretty cool to see you know, how kids can quickly grasp onto using the AI stuff sometimes, too. You know, he's still doing all the school stuff, but uh this is just like a side project where he loves to mess with the AI. >> Yeah. Yeah. >> It's crazy brave new world here. Yeah. I think I use the the Grok Bot >> Mhm. >> Um I have a bunch of bots. One of them's like uh for health and I put all my health information in it, all my blood work. I have a Google Drive that has all that stuff and I just gave it access. It's like, "Use my Google Drive." And I even gave it access to all my past Claude chats and it could go into my old Claude stuff and >> Uh-huh. >> And uh so it's like a good comprehensive like doctor that I trust more than my current my primary care physician on things. You know, it's just nice to have another set of AI eyes on suggestions about whether it's like blood work or healing from an injury or you know, something, you know, sleep optimization or whatever, supplements, you know. I think it's a a game-changer for a lot of that stuff. >> Yeah. Yeah, I think it I think it's underrated how much I I AI is probably already impacting the health health of millions of people. >> Totally. >> like it's just answering questions, it's diagnosing things early, it's just Yeah, I've got this thread on Claude. I pinned a thread where it's like I started with kind of like um uh food diet advice. >> Yeah. >> But I went first into kind of like going into insulin and, you know, >> Mhm. >> just the role of of of fat and different stuff. And then I made kind of this I told AI like we came up with this I this you know, plan of like you know, try not to have snacks between meals >> Yeah. >> unless it's like really not going to boost your insulin at all. But then, you know, it's interesting like AI was kind of came up with this idea of like you either do kind of like um um protein and high fat or you do like protein and carb and low fat or but you don't want to do like high fat high carb. >> No. >> Like that's the worst. >> Okay. >> Right? >> Yeah. >> And so I was like, "Okay, fine." And I was like So what I do is I on that thread if I'm going to a restaurant, I just say I'm going here and it'll tell me like a strategy. >> Mhm. >> Oh yeah, you're going there, you know, get this, get this, but you know, ask for like, you know, just get asked for half the rice or something or this type of corn tortilla is better than the flour. Limit yourself to uh You know, like two tacos with this and and add ask for double meat. Something something. It just gives me all these like stuff that's tailored to to me. >> Yeah. >> And so it's like it's so it's so interesting because it's got like I don't know. It's just it's just so uh so handy, you know? Personal goal. >> Very handy for health. Yeah. >> Yeah. >> And for other like my my another thing that we found value just an example of other than health is like saving money people but a new way we're saving money is like my wife and I are kind of renovating this this uh house that we're going to move to in a year or so and and so there's all these different fixtures and appliances just like hundreds of them that she worked with interior designer gave her a list of things to furnish the house properly and so she just gave the she just spit out the list to to Grokbot and was like "Can you search twice a week for all these things and send me an email twice a week of which one of these things are on sale by at least 20%?" And we've already gotten like it's been a few weeks of this and we've already saved some substantial money ordering a few things on like 40% off that otherwise we'd have to like pay full price for or something a year from that when we move in or something, you know? So it's pretty cool. >> Yeah. Yeah. >> Yeah. >> Yeah, AI was I did this uh renovation for my parents earlier this year. AI was so helpful. Oh my gosh. It just like it like I ran through all the projects and like it basically helped me design like this really custom shower all the helped me order the right fixtures and helped me like That's cool. do everything. It's like Hey, Rob Mauer. >> Oh. >> it going? >> Crazy. >> Oh wow, I didn't know Rob was going to be here. >> Yeah. >> Um So we just >> Cybertruck's in the parking lot. >> Oh yeah. So we just got here to the Hilton Waco. Uh and they are having a shuttle here. You can see the buses on the side here. So we're going to take Oh man, I was hoping they'd do Robo vans but >> Oh, I know. That would have been so cool. Maybe maybe they take us to to that the Robo vans are stationed at, that give us the last mile there. >> Yeah, hopefully we could dream on. >> Who knows? Oh my gosh. What's your last uh prediction for what's going to happen to that? I mean, by the time people watch this, well, everyone will know what happened at the event, but my prediction >> Yeah. >> is that they hover the Roadster >> Yeah. >> I don't know, 20 ft in the air. >> 20 ft? >> Something like that. It comes down, and then it's like by the way, this is what it can do if you're on private property and you call us for authorization to turn this off or on. >> [laughter] >> But otherwise, the space like thrusters are just going to be for keeping the car grounded, going 0 to 60 in 1 and 1/2 seconds or something like that. What do you think? >> That's interesting. I definitely I think going to hover. I thought it would be hover somehow just go, but I don't know how it can go in the air. >> Yeah, I don't know how it go forward. I think it just hover straight up and down. It's like a party trick on private property. >> and then it just goes straight. That makes sense. >> Yeah. I mean >> Yeah. >> Yeah. I mean >> Yeah, I mean, I'm thinking this car is like I have it at Founders Reservation, but I'm it's more I think of it as like an investment like uh I'm going to keep it in a friend's warehouse most of the time that's secured, but I'll take it out once in a while, but it's almost like a collectible that I think goes up in value over time like a piece of art or something someone might invest in to like see if it go up in value over time. >> I mean, Tesla probably won't make it that long. You know what I'm saying? Like it doesn't make sense it for them to >> They said they're making just a thousand of the Founders, then maybe they make like 10,000 other ones or something. I don't know. I don't think they're going to make a million of these. >> Yeah, yeah, definitely. >> They'll probably open up orders for a year then shut it down or something like that. Yeah. >> that and just finish the orders or something. >> Yeah. >> That makes sense. >> Yeah. >> Yeah, if I was younger like yeah, Roadster would be fun. >> Yeah. >> I just yeah, I don't know. It's like a car that fast just gives me a little bit of a >> [laughter] >> pause. >> I know. I'm going to take it to the local race track where I live. They >> Okay. >> They have like a a night where like you can you can test out all your fast cars going a quarter mile and stuff. So, I'll bring it there and test it out. And uh uh then you get kicked off the track if you're too fast without a roll cage or parachute. But, they let you they let you do it twice before the insurance forces them to kick you off. But, everyone's like rooting for you to do it. It's kind of fun. It'll be fun to do that. So, but I'm going to drive it around on weekends occasionally. I hope it's convertible still like the old one because where I live it's like I love I'd love to have a convertible to drive around on a nice on a nice day. >> Yeah. All righty, we'll wrap up this episode. So, this will I'll probably release this after the Roadster one. So, we'll do we'll go to the event. And then I'll we'll share our impressions after the event. Maybe I'll take some footage at the event. Uh release that first and then I'll release this a day or two after. Um Yeah. >> Okay. >> All right. >> Okay.