Tesla’s Robotaxi Infrastructure Is Exploding

[00:00:00]Can you believe it? Today marks the one-year anniversary of the launch of Tesla's robo taxi. While ramp up has not been what many thought it would be by now, the progress for setting up the infrastructure and most importantly cybercap production is growing like wildfire. First spy shots continue to show many cyber caps coming out of the factory, sometimes hundreds spotted in the lots. Interestingly, a new batch of cyber caps had the cybercap decals instead of the word robo taxi. The likely reason Elon said the name will be cybercab in cities where the word taxi is prohibited by law. This is a great sign. Plus, cyberc cabs in Austin were spotted testing picking up and dropping off passengers. Second, for the first time, a Tesla semi outfitted with LAR grounding equipment was spotted driving around in public. Looks like Tesla Semi will also become self-driving soon.

[00:00:49]Third, Lemonade's autonomous car insurance just now launched in Colorado.

[00:00:54]This makes four states in total that Lemonade offers a 50% discount for drivers when using Tesla FSE miles. It launched in Arizona in January. Oregon in February, then expanded to Indiana earlier this month and now Colorado.

[00:01:08]Finally, Tesla just trademarked the word Megapod. They described this as modular data center hardware systems for AI computing. In March, Elon announced that Tesla will be deploying millions of dedicated digital Optimus units in the field at supercharger locations. Make no mistakes about it, this is the start of distributed AI data centers, and only Tesla can get this running at scale this quickly. We got Jeff Lus joining us today. He's a multi-deade supply chain executive and chief quality officer.

[00:01:36]He's now CEO of his own high-tech manufacturing consulting firm. Welcome, Jeff.

[00:01:41]>> Oh, hey. Yeah, these are some interesting topics. I think they're really speaking to where the business may be going in a number of aspects. So, look forward to going through it.

[00:01:50]>> Yeah, we often uh hoping to see more signs of robo taxi roll out. It is today the one-year anniversary. I think many people thought that we would be much further ahead in terms of scale, but if you look at the infrastructure, you look at all these developments, it is quietly building up enough that I think they can still catch up in terms of expectations.

[00:02:11]But right now they're definitely behind.

[00:02:14]Uh Tesla themselves celebrated it. It's the one year of robo taxi. So June 22nd I was there. We did a live stream.

[00:02:22]500,000 people watched that live stream of the very one of the very first robo taxi rides. Elon said the Tesla robo taxi launch begins in Austin this afternoon with customers paying a 420 flat fee. Um yeah. So, this is uh the one year anniversary. I'm just gonna play a really quick clip to reminisce about what we did that day.

[00:02:48]>> I'm wondering how the UI is going to be.

[00:02:50]>> I got it. I got it.

[00:02:53]>> Can you see it?

[00:02:54]>> Keep on going.

[00:02:58]>> See? Uh, okay. The corner there. I can see them there.

[00:03:02]>> I'm going to fast forward it.

[00:03:04]>> So, obviously they come in different colors.

[00:03:08]Trying to see if there's any difference in the visualization because we know that this >> we know that this particular um >> keep him out of it.

[00:03:16]>> Yeah. And then >> this particular uh car is using the upgraded version of full self driving, not the kind that we are all using. You can see that on the screen there it does say pull over, stop in lane or support.

[00:03:30]>> Okay, there you go.

[00:03:34]>> So So can >> turn that off.

[00:03:35]>> Can we just >> We don't know. So, >> there is somebody following us, but we don't know.

[00:03:41]>> Well, we'll see when you get out.

[00:03:42]>> Because remember on the way there, there was a guy that was just a fanboy that was following.

[00:03:46]>> Yeah, but that was a cyber.

[00:03:47]>> That was a cyber truck. So, >> this one is a uh a blue Model Y.

[00:03:51]>> Okay, that's enough. But, I mean, it was a lot of fun. That's a short clip of what happened uh that day. Very uh chaotic uh but it was perfect. First experience of a robo taxi. Uh now of course it's been live but very small numbers of cars uh by this point. It is live in Austin. It's now live in Houston and Dallas. Small number of cars and it's still a ride in California. Uh what's your comment about uh the progress, Jeeoff?

[00:04:19]>> You know it's interesting. I I love looking at these one-year snapshots because you can really see the the progress that has occurred. That was a version I think 13 from an FSD perspective and there were some issues but in general you know the the the the core system worked payments worked uh it you know it drove without you know any kind of major incident of course there were issues with pickup and drop off that because that's the area that um FSD and and and robo taxi have been working on over the last year. I think that when you look back on the on the past year, I think on on one hand, you have Tesla and what they they promised in terms of, you know, short-term and and even medium and long-term goals for the service. There's a couple of video clips that talk about the number of cities uh by the end of the year. And then, you know, I think there was a CNBC interview with Elon with, you know, over a 100,000 robo taxis in the following year, which would be this year. So, it would be interesting to understand what the back half of this year is going to look like. It certainly appears that the the the job tasks like the the number of bugs or issues that were in front of the team were probably difficult to predict especially with these with these models but it looks like everything else they've kept everything else moving including of course the improvements in the software what we see in our own I think a year if you look back if we did the same look back a year on FSD software for in our vehicles versus today. It is in a meaningfully better place. It's almost perfect except for these just final details at drop off and sometimes when um when when departing in in parking lots and so forth. But it if just I think the the key takeaway is a year ago to today massive improvements in software and massive growth in the readiness of the infrastructure, the people um getting all of the cleaning, the charging uh the land and the construction of these facilities to deal with servicing these vehicles. We've seen m multiple permits pulled and and uh or service centers expanded. So a lot of a lot of work has happened and I think I think they are positioned actually for a big roll out is a question of is when are they going to pull the trigger and what are those internal metrics that say we're ready to go. Remember the internal metrics are not only about the quality of the software and and the and there's there's two elements of that the safety and the convenience features. I think they're there on the safety end of it, but in terms of perfect pickup and drop off, I think there's still aspects that they're working on, but they're getting closer.

[00:07:22]And then the infrastructure is is significantly grown. I think all these things together. I think it's a lot to plan for. But we are where we are now.

[00:07:32]Let's look at, you know, as we ex we ex exit June and go into the second half of this year. Can we light up more cities and then can we get a meaningful number of of driverless vehicles out? I think I think you'll start seeing more signs of that in the existing cities uh Austin, Dallas, Houston where they have vehicles today. When we see those numbers meaningfully grow, I think it'll be ready for for broader deployment. So >> Mhm.

[00:08:01]>> Anyway, in terms of and if you compare it to one more quick thing, just other robo taxi rollouts, >> it's not significantly off. They're they're hard to measure because we've only had one other major service launch in volume driverless, but it's it's fairly close.

[00:08:21]>> So Jeeoff, I have a comment about this and I want to see how you react to this.

[00:08:24]So first of all, it there's no sugar coating. It's a lot slower than many of us thought we would be by now, one year later. But if you stop and think about it, you realize, okay, well, they had a bunch of Model Y's. I really wanted them to roll out the Model Y's to at least be, let's say, 100 cars per city. That's what I thought. More cities, 100 Model Y, but the reality is that that is not Tesla's robo taxi service. Tesla's robo taxi service is CyberCaps. So, they wait until April to launch Cyber Caps. And then we're watching cyber cap production come out which is seems to on been track. That's the one that should be 900 per city. If you have a,000 cars n 90 90% of it should be cyber capabs 10% should be model wise. So you know what I'm saying Jeeoff like if you're launching a program and your main product is you know in this case subc cabs you don't necessarily want to roll it out with you know too many Model Y's quite yet or you know what I'm saying?

[00:09:22]Like they're different. That's not the product, >> right? No, I think I think I think that makes sense. But I mean Tesla's known the Cyber Cab ramp plan for a long time and I and I think they've known I think they've known that plan. So I think I think I I again I agree with you, but it it I think there's I think they're just encountering a lot of startup issues as they get the as they get the full fleet to scale. But once they get over this hump, I think it'll be it'll be something that scales like unlike anything else that we've ever seen before. But it's it's just it's honestly it's you've got the hard you got to get the your I think I agree.

[00:10:06]We've got to get the cyber capab hardware ramped up and deployed in cities, which is what's happening now.

[00:10:12]Uh and they've got to get through these convenience issues because the convenience issues unfortunately scale as as they scale the service. So that's why you've got to knock them out when you're at lower levels. So, I think that's where they're at. And I think it's I I said a couple weeks ago, I think somewhere in the next 2 to 6 weeks that they're really going to start scaling this. So, I think at some point in July, I think they're I think we're going to see some meaningful movement just based on everything I'm seeing.

[00:10:39]But, who knows? We'll see. Tesla's going to put safety first and they're going to make sure that the system is performing very well just given all the the media attention on anything Tesla, anything Elon.

[00:10:52]>> I've got a number of new stories to share about server cap production um and progress. So, let's start with this one.

[00:10:58]Joe Techmar is back from vacation and uh he said CyberCap day. He took a look at Giga Texas. Clear shots of a bunch of them in the outbound lot sporting the cyber capab logos. I'm going to ask you about that. Instead of Robbo taxi logos on the doors and rear hatch, here's a bunch of uh new shots. Many employees gather around 10 of them on the west side of the end of lot of line and more lined up outside the factory exit.

[00:11:21]Overall, there was provably over 80 plus all over the site, including several doing what looks like test runs and drop offs around the factory, reminiscent of early test late this year.

[00:11:32]So, instead of Robo Taxi, it has the cyber cab uh decal on it. And I have a theory about why that is. Let's talk about it. First of all, they're beautiful. There they are. Uh James Steven asked any theories on why that decal says CyberCat, not Robo Taxi. Uh if you remember, he said the name of the ride hailing app is Robo Taxi, so it should be called Robo Taxi. But if you, this guy explained it really well. I said remember Elon said the name will be cyberc cab where taxi is prohibited by law and inversely robo taxi where cab is prohibited by law and Elon said they might need another name for where both cab and taxi are prohibited. Uh so I thought that that was the right the right answer because uh actually which means that it's actually a good sign to see a a lot of them that are called cyber cap. That means that they plan to roll this out to cities now that you're not allowed to use the word robo taxi.

[00:12:28]So they're going to call it cyber cap.

[00:12:30]That's what I think it means. What what's your interpretation?

[00:12:33]>> Yeah. When you're when you're when you're scaling out a new product uh and you've got different regions, you've got different regulatory, you then you got to build different model types to serve that. It's no different in the smartphone world. As you're as you're rolling over and getting to those new model types, you're invariably just building more volume. So, I agree. This is a sign that, you know, they're going fairly wide with this service.

[00:13:01]And if we were just stuck on the I know someone's going to look at this and say, "Oh, now they're they're obsessing about the the the name on the decal." And but it actually does mean something. It means that they're they're rolling out the other model types, meaning and by the way, it is a different model when they actually have to label it differently on the outside of the vehicle. Even though this seems minor to anybody watching this video, when you're in a factory and you, you know, you've got to deploy tens of thousands of something, you've got to build a particular model and then switch over and then get to the next model, no matter where that configuration occurs. So that's what they're doing. So it is it is synonymous. It is significant. And it it does show that they are they are building an expansive fleet to hit multiple regions and they've got to deal with these regulatory issues of what they can call the service or call the vehicle or label the vehicle. So it is it is a positive sign.

[00:14:01]>> One thing Joe said was that he saw some of the cyber caps testing and simulating pickup and drop off in the Giga Texas.

[00:14:08]There's a person here who spotted it in public in Austin. Tesla's engineering safety right into the cyber cap. That's how he's positioning it. He said a Tesla employee in a production cyber cap simulated a drop off and pickup near downtown Austin. He exited from the passenger door, re-entered from the passenger door, slid over to the driver's side. The cyber cap seats have no middle obstruction, which allows an easy slide across to the other side. So, you're going to enter in from the side, not not through the middle of the street. This makes it safer to exit while avoiding oncoming traffic. The cyber cap can also park very close to the curb and open its butterfly doors upward to avoid scraping or hitting the curb. This can make it safer to get in and out of certain situations. And so these are the thoughtfulness. How close can you drive to the curb? Can the door open wide enough so you can go ingress and outgress uh and safer than you know other cars where you need to park further away from the curb or something like that. Here's the photos. They're testing it out in um on the streets.

[00:15:09]Yeah. And then here's a view where CyberCap is driving on a test track uh there as well.

[00:15:16]Yeah. I I I saw somebody commenting on the uh the wheel covers and saying, "Wow, these things these things these things um I guess somebody was bending and folding the rubber back. Oh, these things look cheap." I actually think they're designed for what you just described that they will be getting close to the curb. They will be, you know, they will be, you know, experiencing a lot of wear. And I actually think they're designed to be easily removable and replaceable and low cost where wheel coverings typically are not. And they and they're one of the things that really show the age of a vehicle. So, I actually think this is this starts at the Fron level. This is a fron thing.

[00:16:00]with the engineering team. The engineering team is like, "Okay, here here are the materials we can use.

[00:16:05]Here's what's durable. Here's how we can design it. So, it's a compression fit."

[00:16:09]And France is like, "Well, this is the look. This is um kind of the the profile we want as you're approaching the vehicle. Here's the colors, materials, finishes. CMF is what term is used in the industry. And they both work together. So, they both link up and they work with manufacturing to figure out how do we build something, how do we take something that's traditionally a high cost thing. If you if your um rims in your vehicle need to be refinished with curb rash, you know, the cost on that can be depending on where you're going and how good of a job they're doing. It could be, you know, hundreds, a couple low hundreds uh to get that work done. And here you have something that may be a a low dollar part to just swap out and replace. And it also looks like it's made of a material where uh it looks like it's made of a material where even if it gets a couple of scratches or dings, the underlying substrate under it's the same color. I haven't seen it up close. I'm not 100% sure of that. But it looks like those are designed for easily easy replacement and serviceability. You don't need even need to be in a service center to do that. And these are I think there's a lot of things you have to think through.

[00:17:25]A lot of companies that build robo taxis take existing vehicles that are for personal use that have been designed the traditional way and then put them on the road. Well, they're heavier and the high wear items probably cost a lot of money to replace. So this is why I think this program has taken a couple of years from when we saw it at Ri robot to getting into it to production because these are all the things that the industrial design team under Fron the engineering team under Lars and the manufacturing teams and the supplier realization teams have to work through to come up with materials a seat that you can easily slide on a seat that can be easily serviced and swapped or or the uh the outer upholstery maybe can be easily serviced and swapped. The wheel covers, these are the things you have to think about on a vehicle that's going to get much more ingress egress than ever before. So, I love the examples about a proximity to the curb, being able to go in and out from the sidewalk side. These are very thoughtful things the Tesla team has really worked through.

[00:18:32]>> That's a very good point, Jeeoff. Yeah, these tiny little things. I mean, who would have even thought they let's make the wheel cover rubber.

[00:18:41]Yeah. And then how close can we park to the curb without right because the issue is always a you know curb rash or something that you don't want to do that and that's why they put the wheel cover and yet you can still park really close and then where does the it's brilliant the little things like this this is the takes yeah months and months to figure this out and then like you're pointing out you'd have to go and make changes so things might take longer depending on how important is so the one the one thing I am very disappointed with is the progress with California. So there it is not a robo taxi, it's a ride hill and it's been like that since I don't know since the day one. I don't remember when I can't remember when it was launched in California but um so some little signs I don't think this is very important but some little signs here was that California DMV uh changed Tesla's permit holder from Tesla to Roboaxi LLC. So now it says Tesla's robo taxi LLC. Here's the list of all the companies that can actually test. Um, you can test uh as long as you have a safety driver, you're authorized to test on any public road and uh has issued autonomous vehicle testing permits. So, those are the companies. In the past, it was just um it was just saying Tesla and now it says some robo taxi. Sure. Okay, that's straightforward. And then another one is there's now a little permit here. Says allows a limousine permit again. ride to be able to pick up drop off at the in San Francisco. Yeah, San Francisco airport. So, that's great. What do you think's happening in uh California, Jeeoff? Why is it not moving forward? I thought they needed to get enough data, but by now with the number of cars they have, the number of months they should have gotten the data.

[00:20:29]>> Now, just to be clear, there is no there's no major um incidents or crashes. That's not what's happening.

[00:20:35]But but why why is it taking this long?

[00:20:38]>> Well, we like I said, we don't know the we know well there's it's been talked about in California like permitting cues for a number of different industries just seem to take a long time. We don't know what's actually going on. We do know what's been reported to Nitsa. We we know that there's there's no serious issues.

[00:20:56]There's nothing like that that would be holding up. So it could be imminent. It could be it could be quick. We don't we don't know the the timeline, the expectations, the back and forth between um the California DMV and and Tesla. So, we don't know what's open and remaining, but it's probably imminent. They have a lot of cars out on the road, several hundred.

[00:21:18]I know um some of the accounts we have have um paused recently, but they have a lot of cars on the road doing a lot of miles. So, these are subtle good signs.

[00:21:29]I think the size of that list is very interesting because we know that there's really only in volume who Whimo and some volume of Zuks uh giving rides. So it it's it just shows you how difficult the bar is to enter and to be a player in this. And then I mean none of them are profitable. So Tesla's is is approaching this not just to enter and to be in the game, but they want to they want to dominate. you dominate by being profitable, high gross margins, and have a product that people want to use uh repeatedly. So, that's the safety side of it and that's the cost side of it with the vehicle. So, we'll see. I'll do I'll do some more research online. If there's anything that comes up, I'll uh I'll post it. But, I I imagine that they're in a queue and it's it's fairly imminent. They they've been testing um for I mean north of nine months, I think.

[00:22:25]>> Yeah. Okay. Uh Tesla Semi that looks like it is uh now testing ground truthing equipment. They often do this to just double check that the FSD full self-driving is actually working accurately in certain locations.

[00:22:39]Anyways, I don't know where this was spotted. Um I the guy the guy didn't mention which city state this is seen, but there we go. There you can see that >> that's the semi on that. Yeah.

[00:22:51]>> Yeah. This is not this is not surprising but it's obviously gamechanging. First with with semi if they can just extend the you know with with a driver in the seat and then you know just like we're experiencing with our cars today with FSD when you you're on a couple hour drive or you're doing errands all day and you're letting the car drive you you feel less fatigued you feel better at your destination and you probably could have driven more and more that day. I find myself when we decide like who's driving, you know, in the family, I find myself saying, "I'll I'll do it because it's it's not a chore. It's not it's it it's great." And and so I I think I think out the gate, I think with FSD and semi, it's it's going to be a completely different experience for drivers.

[00:23:43]They're going to feel much better. They can they can be more alert and just going to be a much better experience.

[00:23:48]And similar and synonymous with robo taxi, I think a lot of the heavy lifting, no pun intended, is going to be on pickup and and uh and drop off of of loads out of the out of the out of the vehicle. And that's where a lot of the work is going to have to happen autonomously. I think for a long time, actually, I think you're we're going to have autonomous trucking, but I still think pickup and drop off are still going to be there's going to be a lot of human intervention involved in that. But I think it's going to get to the point where you won't need to be in the driver's seat. But I think the first phase is you're in the driver's seat, but the but the truck and FSD are doing all the work. So this is great to see the semiis ramping in the second half of this year. And this is a kind of an underdisussed thing in the Tesla >> Yeah.

[00:24:39]>> world and an underdisussed thing in the auto and media world. But Tesla has been in the lab with this semi for we know it's been north of 10 years. It's been uh 9 years since they revealed it and we know they've been working on it beforehand. So this is going to be I think a great product >> even better when it's self-driving regard like you said even if it's not robo taxi robo trucking yet but as long as it's full self-driving game changer already. huge.

[00:25:12]>> All right, this morning, Lemonade, they announced that they are now live in Colorado. This is the fourth state in total that's doing this. They um if you remember, they have a car insurance. If you are in these four states, so this is Arizona, Oregon, Indiana, now Colorado.

[00:25:31]I believe they have 11 states in total that they provide car insurance, of which Washington state, where I live, is one that I'm waiting for myself here.

[00:25:38]But every time you drive in FSD, all them FSD miles, 50% discount for your car insurance for those miles. Um, now it's live in Colorado and uh, yeah, this is their this is their this is their marketing gig, which is actually pretty funny. Which is pretty good.

[00:25:57]Yeah, it is funny. And this is good to see. And I think others are going to be forced to to pick up on this. Tesla is let out with with their insurance product. We haven't seen many announcements, by the way, in terms of expansion of Tesla insurance. I love it.

[00:26:14]I mean, so far it's been working out um great for me. Uh and it is essentially the delta for me for like for like insurance in terms of the premiums, the settings, everything. It saves me about $100 per vehicle per month.

[00:26:33]>> And that's FSD roughly. I mean it it vascalates depending on your driving score which now my driving score is consistently pretty high with FSD doing the driving. So for me this has been a wonderful thing. I couldn't recommend it more. I know others have had mixed experiences but most of the experiences I've heard have been have been very good.

[00:26:56]>> Yeah, this is proof uh that if an insurance company is willing to actual real dollars reduce the charge. I mean you can argue maybe it's marketing but reality is they wouldn't do that if they weren't still making money. They can't do it too far too far too often but um they are expanding. This is the force they >> it's smart for them because model Y is such a high in three are such high selling models that it's a very smart thing for them to do to go and get those buyers because they're great products and they they know that they're great products and they see the data in terms of how they're performing in the field.

[00:27:32]with FSD. So, this is a very smart thing for Lemonade.

[00:27:35]>> That's Lemonade. Okay, here's a very big story. It just starts off with a trademark, but it tells you that they're moving forward. This is they Tesla just trademarked the word Megapod. And if you read down here what it is, and this is just a uh a mockup of what this could be, but a megapod is a distributed AI data center servers in where the superchargers are because they've got the batteries there. And uh it's called modular data center hardware systems for artificial intelligence computing comprised of computer servers, computer hardware for artificial intelligence processing, computer networking, hardware, electrical power distribution units and cooling systems sold as a unit self-contained modular computing hardware system for AI workloads. Da da da da. Wow. So here we go. Is this I think this is related to the dedicated digital Optimus units running on superchargers and uh yeah I think so if you think about it this is um this is what Elon said. So if you remember in March he talked about Mark Microheart and the other name is Digital Optimus.

[00:28:41]It's a joint XAI Tesla project and then Gro's going to ma be the master conductor of the world uh to direct digital Optimus which is processing actioning the past 5 seconds of real-time computer screen video and computer and keyboard mouse actions. And then you can think of digital optimist AI being system one instinctive part of the mind and gro being system two thinking part of the mind. Okay. So one that controls the computer the other one that answers questions and so forth. But down below he says oh and it works on all AI4 equipped cars. So your car can do office work for you when not driving.

[00:29:19]Okay. So our cars are mobile distributed AI data centers. Now he also said we're also deploying a millions of dedicated digital optimist units in the field at superchargers where we have 7 gawatts of available power. So I think this might be inference in the cars but this one could be also used for training or inference. I don't know but anyways um yeah what do you think about this Jeeoff? This is if this is just a mockup but how cool would this be? How how cool is it that you know Elon has these assets available to him and him and his teams can come up with I mean here's what's the landscape right now? The landscape right now is there's dozens of states that don't want data centers that are actively protesting against them.

[00:30:07]Illinois pulled a lot of tax benefits for for data centers. And then you have this far out thing that's could be a couple years out of putting them in space where you don't have a lot of these energy power and regulatory or the human issue the complaints um about data centers. So what does Tesla do?

[00:30:29]Because if you if you look at this there's this other thing going on of NeoClouds and and former Bitcoin miners that have power. They have power. So, they're getting into the data center business and they're hooking up with companies. You think of companies like um these Neoclouds, Nebius, uh uh former Bitcoin miners, iron, uh uh Clean Spark, and then now, so what Tesla is looking at, it's like, hey, look, we have all these superchargers. We have power at these. We have a lot of power at these superchargers. And what plugs into these superchargers are AI4 inference computer vehicles. And of course, we can put compute inside of these megapods as well. So, this is Tesla looking at and leveraging their assets like like no other. I mean, this is you'd almost think like this was thought of or planned of. And I think this what actually wasn't. I think this is a bit of a reaction to what's happening with compute and what's happening with the the the strangle hold on power. I think it's just such a great solution though.

[00:31:43]And we'll see. We'll see how it works out. We'll see how people respond to it.

[00:31:48]Um, I I can't I mean I think these things can't go in in more densely populated areas, but in some of these where some of these superchargers are, there's not a lot around it. And if you were to pl pop in something of that size that was in that rendering, I don't think people would have an issue with it. So this is Tesla looking at where where we have power and we have compute putting the two together. And it's kind of this tweener space between having the inference in the vehicle, having this larger capability next to power and of course all the compute that SpaceX is doing, you know, on terrestrial on Earth and there's another deal was even signed this morning. But the far out thing of going out into space, which is a couple of years out. So this is Elon covering all the different spaces all the way from inference down into the device level all the way through up into space.

[00:32:44]And I didn't think about this, but I know I remember this conversation from the macro hard conversation and leveraging superchargers. This is a really I guess is a really thoughtful approach as a response to all these challenges out there.

[00:33:00]>> I have to think this through. I mean, first of all, they said millions of dedicated digital Optimus units in the field at these supercharger locations.

[00:33:07]Now, we know that they're called megapods.

[00:33:11]What I need to think through is that is that uh they likely won't be coherent.

[00:33:16]If if they had an issue connecting macro harder 2 to Macrohard one because of the distance and they were like literally like nearby, u I don't know how they would do this as as coherent. So, maybe this won't be for training. This will just be for inference, which is what which is what digital optimus is for inference. So, and then and of course they've got power, electricity, they've got the mega chargers, batteries, and then now they've they probably outfitted with Starlink, easy peasy. Now you got connection millions of units.

[00:33:50]>> The other thing I'm just I don't know. I don't know how I mean this could be way out there but you do have this Starlink connection as you described and one of the things that Starlink is missing to be a real cellular operator is the terrestrial presence. So, for example, when you're in buildings, when you're that, that's where Starlink wouldn't work if it was direct to cell. But if these pods are everywhere and they have the ability to kind of act as I I mean, it' be interesting to see if there's in in the uh >> in the write up it's what are they talking about in terms of antennas >> and transmission >> um out of these pods. So anyway, it just it could be something. It could be nothing. But I have to think that there could be something else to this related to communication between Starlink satellites and potentially direct to our devices, but we'll see.

[00:34:48]>> Any connection possibly for cyber cab like robo taxi specifically? So if you had mega pods and each supercharger, we know that the robo taxis might use some of the superchargers for charging. Is there any other benefit to having a megapod there for supercharger for robo taxi itself? Every robo taxi will have connection to Starlink. They will need to have some sort of gro as the orchestrator of the robo taxi. Think millions of robo taxis all over each city. Hundreds of maybe tens of thousands per city. And then you need to orchestrate uh where the cars go and all that. I wonder if these mega pods can serve as many brains, many location.

[00:35:31]>> Yeah, I mean it could definitely be for connectivity for sure. But I I think I think in the beginning I I don't know this. I actually don't know this for a fact cuz we haven't seen a tear down of cyber cab if it's coming out. I think they're coming out with traditional cellular modems inside of them, but I could be wrong. Uh maybe that's a backup and then maybe eventually I agree they will have Starlink uh in them. But yeah, there could be a number of different uses. But right now, to me, this is if we just take it at face value, >> this is modular compute at the source of power. And Tesla has many of these sources of power with this supercharger network that they've built up over the last 15 plus years. It's incredible.

[00:36:16]>> All right. Well, today's June 22nd. It is a one-year anniversary.

[00:36:21]I'm just curious what's going to happen.

[00:36:23]Uh I was hoping that Tesla would actually use this day as some sort of you know cyber cap first cyber cap to uh to and be part of robo taxi. They haven't done that. So we'll see what they end up doing rest of day.

[00:36:36]>> The day is young.

[00:36:38]>> Wonderful. Thank you Jeeoff. Appreciate you. uh lots of um you know disappointing how far Robbo taxi has gone but we continue to report on the uh infrastructure that's rolling out and that and and most importantly cyber cap production seems to be >> safe and so far it's been safe it's been issue free since I think February and even those issues were minor that's probably the most important thing if if they had you know 5,000 vehicles and they had a couple of serious issues I think >> we're in trouble >> in a different place >> you're right this is the it's a go and it's just a matter of when and we're just uh anxious. Thank you, Jeff. Follow him on his exac account at the Jeffoffs.

[00:37:18]Thanks, everybody.