特斯拉的自动驾驶数据收集正在快速增长（周环比增长17%），但面临监管阻力以及Optimus近期投资回报的不确定性。

Cybercab event yesterday. The question is how impressed were you or was it a complete nothingburger? Yeah, I think it was interesting. Um it was hard to know what to expect as far as like there's this whole media embargo, what's going on? Um I don't understand what the deal with that was,

you know, the thousand-foot view and I I I certainly wasn't there was that they really wanted to just highlight get people in the cybercabs and get them as much experience riding in the service as possible. And so, the presentation wasn't a whole lot to that. It seems like it was almost just like a celebration event for the Tesla team that's been working on this.

I'm really curious to get to hear more from all of the people who got to talk to the engineers about some of the details that um they were able to learn on the back end of things there.

But, my biggest high-level takeaways that I haven't heard a whole lot of other people talking about um except for, you know, a relatively small number is that 17% week-over-week number that we can back into in growth in the uh unsupervised self-driving miles.

And the fact that they called it unsupervised self-driving miles to me says that they're counting you know, these are deadhead miles in a large number of instances, but I think that's okay at this stage right now. So, they don't have necessarily paid rides for those miles,

but they are able to operate the Tesla Model Ys and presumably quite a few cyber cabs now in Austin completely unsupervised, no one in the car, and they're able to gather a bunch of data.

And so, growing from where we were at, you know, in the low hundreds of thousands of miles on the Q2 earnings call uh a few however long that was ago, 6 weeks now, I think. Um and then passing 1 million miles as of yesterday is when we know it happened. And that means that we're we're just growing incredibly fast.

And 17% week-over-week, if you were to annualize that, and I'm not suggesting that this is the number that they're actually going to hit, but if you annualize 17% week-over-week, you get to a 2,600x from the beginning of the year to the end of the year on an annualized rate.

Um and so, that's just an incredible growth rate right now. Um that means that I think it's very reasonable to assume that we will exit 2026 with over 5 million miles, potentially as much as 10 million miles if we can sustain the 17% week-over-week.

Um and so, that's just absolutely huge growth in miles, and it means that Tesla's going to be able to validate new releases of the FSD software in real time on real roads at a much higher rate. And that means that we can continue to sustain development of new FSD models to solve like we know that there's going to be corner cases that make this a hard roll out to really be able to deliver you know, millions of miles in cities all around the United States.

Yeah, and I think I Yeah, I think bigger than that, too, but the stock did get dinged because of that NHTSA news of the 2 1/2 thousand limit. And that is they'll overcome that. The question is how long will it take? Will it take 30 days or is it completely just FUD? Or could it be a longer-term problem?

And I think what Tesla should do, if I was them, I'd focus on markets that are acceptable or accepting of, you know, safety and autonomy. Like there's places like um what's the name of the place with the famous Japanese rose festival or the spring blossom festival or something. They have an aging population. They have no cab drivers anymore. And nobody can get around because the young people don't want to drive cabs. You know, just drop a couple couple of thousand there in Japan. There's sure there's places in Europe that'd be arms wide open to it, too. And then there's people in need all around the world that need mobility. So there's there's so many use cases.

But when I hear about the government again putting sand in the gears, I get so frustrated. You're taking on NHTSA. Yeah, it is somewhat concerning and especially when you think about the the town that the cyber cab is going into and the number of entrenched players that are there, you know, whether it's taxi unions, which we've seen be very influential in places like New York and when Uber was trying to penetrate there. Like these things do happen and it is hard to maneuver around them.

Uh but now you have companies like Uber that are also entrenched and they have their interests. So, I think um yeah, people inside of the Tesla community shouldn't underestimate the amount of resistance that there can be on the regulatory front because it will be well-funded and well-supported by competitors and the opposition.

Um so, yeah, I would say that it it is somewhat concerning. That said, uh to your point, there's you know, places that they can go, jurisdictions where I think they can prove things out.

And ultimately, I think it's a matter of getting the um groundswell of political support from people. You know, what we saw at the Cybertruck launch event was that the Cybertruck was significantly cheaper than both the robotaxi model wise and Uber.

And so, if you can get the message out to people that, "Hey, like these big, rich companies like Uber are trying to steal your cheap transportation from you, then they might be able to turn the tides on that front."

this is kind of a Tesla SpaceX story.

How long will it take Tesla SpaceX to get terrafab up? when would you, if you had to put a line in the sand, when would you think terrafab would spit out their first chip? And they're starting with the mini fab first, and then the big one afterwards.

And so, yeah, as Elon's looking at this whole system of things that needs to be solved and Terafab being one thing that can do all of those things in one place and really start to eliminate some of those bottlenecks, I think that packaging might be the place that he starts.

And so, we may see that that mini fab um will begin with packaging and then expand into wafer fab so they can do fab and packaging together in the same place.

Um, and I would not be surprised to see something in the next 1 to 2 years actually come out of that facility um But probably just on the packaging side and that would I think it's the lowest hanging fruit for them to tackle.

And it's also one of the highest because packaging, while it is hard and it is complicated, it's not as hard and complicated as, you know, a 2 nanometer process um for TSMC's most advanced wafer fab or what goes into doing the the the base silicon fabrication for logic either for for memory.

What happens if Terafab and Elon Musk or some other entity figures a way around the the simplification of chip creation without all of these complex tools and processes and vendors in the mix?

And this is even true of you know what Elon does with Tesla and maybe to a lesser extent with SpaceX.

that's still I think they're, you know, more than likely just fab capacity is going to be one of those things where it looks a lot more like Tesla's 4680 production where they, you know, they're buying as many batteries as they can get from as many people as they can get and building their own.

A little bit of good news regarding Tesla, though, it's coming off the floor. It's rebounding. And the price of Tesla right now is higher than it was on Wednesday.

You know, we have now Cybercab running live. We've got Optimus. I know two people that I've met personally, that I've seen it. They One says it's creepy. The other one says it's absolutely sublime. The semi-truck is ramping. They have that that event, which probably be another nothingburger, but the point is they've already have some like 1,600 orders for the semi-truck. So, everything they make for the next year or so will be sold. I think the capacity is 50,000 in Nevada, but I don't know if batteries will be a bottleneck for these things because they suck up nine times the amount of batteries a Model Y takes.

Uh and then, of course, Cybercabs. The big news yesterday for me, the most exciting thing wasn't the growing 17% in terms of robotaxi miles. What was exciting is the fact that people can buy their own fleets. So, if they can build the Cybercab at scale for $15,000, which they say that was the target at scale to get to maybe 18,000, sell them for 30, 35,000, they can sell these things all day long. They don't even have to bother selling Model Ys and Model 3s and Cybertrucks, or maybe even semi trucks anymore, because people like me will queue up and buy them.

So, there'll be a lot of pressure from Trump and the administration to get that thing through in the next 30 days. But the question people have is what's the better value now? Like, SpaceX has so much goodness coming, so much incredible growth, so much high margin growth potential. And then they both have so much stuff that they're joined at the hip on, like humanoid Optimus, which would be very much required for SpaceX. SpaceX is buying a ton of Megapacks. Uh if and when SpaceX >> Megapacks, yeah. >> Yeah, and that's all It's all joined at the hip. When SpaceX buys Tesla or merge with Tesla, there'll be a premium that'll be the 40 to 60% I reckon, which is normal, or else shareholders will not accept it. And that's the call option on the future. So, you're already seeing much more tight activity between the Tesla-SpaceX pair chart. Uh it's kind of like ratcheted onto each other right now.

>> Ooh. Yeah, I'd probably go 75% SpaceX, 25% Tesla today. But part of that is just driven by the fact that like right now I probably have 90% Tesla and 10% SpaceX, if if even that. Um and so I haven't, yeah, like I'm a really boring investor, especially compared to someone like you.

Well, let me come back to Tesla for just a second. I think I'm extremely excited about Cybertruck and the regulatory uncertainty is the big overhang there. Um I am very bearish on near-term timelines for Optimus. And I know that people are really excited about what they're seeing, but what they're seeing are still demo units and demo tasks. And in my mind those are still prototypes. Like, in order for Optimus not to be a giant money-losing project, every Optimus that's manufactured has to produce a positive ROI. In order for that to happen, that Optimus needs to be able to do useful physical work for hundreds and thousands of hours at a time with little to no maintenance. And that's the thing that you're never going to get any visibility into with any single demo. That, you know, until you put this thing in a factory, you spend however much it costs to buy your your robot, and then it has to do more work than it cost without costing you then more on maintenance. And the hands specifically are really hard on this standpoint. Like, the hands are the most expensive thing to make, and they're also the easiest thing to break. And if those hands can't do whatever the jobs are, then like maybe there are low-hanging fruit jobs that it can do, and it can produce a positive ROI, and as long as you don't produce more bots than you can allocate to those jobs, you can keep the whole project on a positive ROI. Um I'm just Yeah, like I'm heavily discounting the timeline on Optimus robots at least until 2030. Um I do think that Elon is like the production line that they're putting in, they're 100% going to make bots off of that. They're going to be putting those things, they're going to be finding what all the edges are, they're going to break it, but I expect all of that to be a giant money-losing project for a number of years still. And I don't think that the you know, I'm not excited about it as an investment opportunity for Tesla until there's visibility on that project turning to positive ROIC.

they're freaking mind-blown. And they know they're going to get to a million production probably by the end of 2027. And maybe 30,000 of these will go into some Tesla Academy to learn from each other. That's kind of the big vision. But I wouldn't I wouldn't be holding your breath until 2030 till we start seeing big results in these things because they will be deployed in Tesla operations, they'll be probably deployed in SpaceX operations, and then supplier operations.

Well, well, the thing is if you've seen one of the things that also gives me confidence, you look at how Tesla have made vehicles. I've got a 7-plus-year-old Model X outside. It has never had a problem. Zero maintenance. It's like the most beyond perfect. I used to have other cars like BMWs and Mercs and stuff. They'd be in the shop every 12 weeks. >> Yeah. >> Costing me $2,000 a visit. And I've never had that with with any of the cars. There's three of them out there. All absolutely flawless. So, I think it's very bizarre. But this actually this So, the quality of what Tesla make will be second to none. Therefore, I think I know you're a hand bear. I'm going to call you the hand hands the hand bear. >> Yeah. >> But I think they've got that figured out, too. And then

Look at the electric motor within a Tesla. Again, it'll do a million miles. No problems, no service, no nothing. Oh, I I know that they know how, but the question is, you know, it's just has to be so small in that form factor that that means that you're pressing on like the you know, if they had an unlimited weight budget, 100%. But they don't. And so just within side of the engineering envelopes of all the trade-offs that they have to make, you know, to have the dexterity and the reliability. And the power efficiency.

one of the specs they do have for the two hands is to be able to carry 25 kilos or 50 kilos. I can't remember which one. Do you remember? I get confused between pounds and kilos as well, by the way.

Yeah, I would assume that the spec is waterproof out of the box. And they do have the new Optimus 3, little secret here, it does have a waterproof dextrous hand with temperature sensors, weight sensors and stuff on the fingers just like a human hand. They've basically replicated the human hand.