The video correctly identifies that the battle for autonomy is won through economic scalability and AI integration rather than sensor redundancy. It highlights how Tesla’s minimalist hardware strategy creates a cost barrier that competitors struggle to overcome.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
SpaceX AI uncatchable! - Xpeng Copied Tesla’s Vision-Only…But Missed the whole point with Cybercab!Added:
Hi, welcome to Best Tesla and the new show Overflow episode 90. Space XI is rising to become the backbone of the entire AI boom and no one really seems to be able to compete. And Xping has come out with their new robo taxi. But even though they copy everything Tesla does, it seems like they have missed the whole point about the Tesla Cyber Cap and Porsche is pausing production in its plant in Stoodgard because of weak demand. And the boring company shared their new profuck boring machine that can continuously mine and we are seeing just how good a value for money Teslas has become in the US. And Tesla hit some big milestones here in Europe with their full self-driving.
And we had yet another example caught on camera of full self-driving saving lives.
>> WHAT THE HECK? HEY, >> WHAT THE [ __ ] >> All of this and much, much more in today's episode. Let's dive right in.
Let's put our living stereo stylus in this groove.
SpaceX AI is rising to become the backbone of the entire AI boom and no one really seems to be able to compete.
We are learning a lot more about SpaceX business because of the IPO filing that SpaceX has done and one of the thing is production capacity of satellite. As Brian shared, SpaceX says in his IPO filing on Wednesday that is Starling satellite manufacturing facility in Redmon produce an average of approximately 70 satellites per week from December 2024 to April 2026. a yearly rate of about 3,640.
So to put that into a little context, someone like Amazon's facility has a peak capacity of up to five satellites per day or over 1,800 per year. So this is also just such a good showcase of how SpaceX is simply uncatchable right now.
SpaceX already have 10,000 satellites in space. Amazon has 300, but they will simply not be able to catch up as they produce half the amount of satellite as SpaceX does when Amazon gets to their peak capacity and they can't even get that many satellites into space right now with their cadence of rocket launches. And on top of this, as the other Brian said from next big future, this will be far to 10x in 2027 with version 3 internet and cell phone satellite with high volume Starship launches and then 10x in 2028 to 2029 for launching the AI satellites. Yes, SpaceX is just playing on a whole different level than their closest competitors. They are already out in front, already running faster than everyone else and is accelerating faster than anybody else. And on top of all of this, we have heard of the big deal that SpaceX made with Anthropic where they pay SpaceX $2.1 billion a month or about $15 billion a year for access to SpaceX Colossal 1. you know, the coherent computer cluster that everybody said was not possible because Elon Musk wanted to build 100,000 coherent GPUs and at the time it were thought that they could only build 30,000 coherent clusters.
bigger than that would simply be impossible. But not only did they make a 100,000 coherent GPU possible, but build the thing about 20 times faster than anybody else even knew how to build a much smaller coherent cluster. As XAI writes on their website, "We were told it would take 24 months to build. So we took the project into our own hands, questioned everything, removed whatever was unnecessary and accomplished our goal in 4 months. So instead of 2 years, they did it in 4 months. As they also write, built in 122 days, outpacing every estimate, Colossal was the most powerful AI training system yet. Then we doubled it in 92 days to 200,000 coherent GPU clusters. This is just the beginning. Yes, they just broke that record once again with Colossal 2 where they added a 110,000 GPU cluster to the Colossal 2 in 64 days and wants to add another 220 GPUs to Colossal 2, getting an up to 550,000 GPU coherent cluster. So SpaceX is once again accelerating and the others don't even know how to do what they did in the first place. And we knows that SpaceX is aiming for 1 million GPU clusters in Colossal 2. But with the cadence we are seeing right now, they could have that online by the end of this year. But of course, power permits and supply chain issues are starting to bite harder at this scale. So, it could of course take a bit longer, but within the next year or so, SpaceX could have Colossal 2 running with 1 million GPUs. And if the deal with Anthropic is anything to go by, SpaceX could rent out half of that capacity, so 500,000 GPUs, and earn another maybe $15 billion a year. Then SpaceX would earn $30 billion a year on just those two supercomputers. But all of that is just the beginning of what's to come. As XAI also writes on their website, Anthropic plans to use this additional compute to direct improve capacity for Claude Pro and Claude Max subscribers. As part of this agreement, Anthropic also express interest in partnering to develop multiple gigawatts of orbital AI compute capacity. The compute requires to train and operate the next generation of these system is outpacing what terrestrial power, land and cooling can deliver on the timelines that matter. SpaceX is the only organization with the launch caden massto orbit economics and constellation operation experience to make orbital compute a near-term engineering program rather than a research concept. If engineering challenges can be overcome, space-based compute offered near limitless sustainable power with less impact on Earth. Yes, space data centers will make the anthropic deal look like a rounding era. Space AI will be the new frontier and nobody is able to challenge SpaceX at the moment on cost of launch, getting mass to Auburn, build a huge data center faster than anybody else. It has almost been 2 years since Colossal 1 came online in just 122 days. But still to this date, no one, not Meta, not Microsoft, not Google has built such a big coherent cluster in such a short time. And SpaceX has already cut that pace in half. But here we are 2 years later and no one is even close to the original record of Colossal 1. I think there is no doubt that SpaceX is going to build the backbone of the infrastructure for this AI boom we are watching play out right now. And have you been noticing how Claude is the most talked about tool on the internet right now? Claude has been launching new models and features like Claude co-work skills connectors literally every week.
It is indeed the most powerful tool right now which has enabled people to go from idea to app to launch without a single line of code creating full stack presentation run their lives on autopilot and even replace an entire marketing team with it and naturally everyone everywhere is telling you to learn claude but rarely do they teach you how to but luckily I come across outskill today's sponsor and they offer this world's first claude athon a deep dive into claude its use cases and more than 10 other AI tools happening next weekend from 10:00 a.m. to 7:00 p.m.
Eastern time. And they have just 1,000 free seats for a limited time. 10 million people across the world has already attended this and rated it 4.9 out of five on Trustpilot. I would highly recommended you take this course.
In the workshop, you will do a deep research on Claude. Build your own artifact dashboard. Create full presentation using Claude. Set up Claude connectors like Indeed to automate your job search. Master more than 10 AI tool.
Build custom GPTs and agents and generate visuals and videos with AI. And at the end of it being in a situation to automate your entire week and if you sign up now as a bonus, you will get 50 secret code that will make it a godle tool for you. and an AI prompt library and personalized AI toolkit built all for free. And the best part is you get mentored by leaders from Microsoft, Google, Amazon, and Nvidia. So what you will learn in these two days will change the next 2 years. So sign up now before they sell out. Use my link down below or scan this QR code to join the WhatsApp community before the seats is closed.
And a big thanks to Outskill for sponsoring this video. But while we are speaking about Tesla Cyber Cap, the Chinese Tesla copycat Xping has now come out with their robo taxi. And even though they are copying everything Tesla do, well, they seem to have missed the whole point about Tesla's cyber gap. As Chan wrote, Xping CEO said, "Lida remains a useful and necessary component in many other industries, but he doesn't think the sensors are needed in the automotive industry anymore." Yes, exactly. Xping yesterday launched its new six-seater SUV GX. The vehicle has adopted a vision solution. 24,800 confirms orders secured in 12 hours of vehicle launch. Very cool. And yet again, Xping validates Tesla's approach instead of the way of the LAR. And this car will also be sold to private persons, but it will also be a robo taxi. So, personally, I think they missed the whole point about Tesla's cyber cab. Tesla has created that to make sure transportation becomes as affordable as absolutely possible, reducing the cost to build the car and the cost per mile in operation down to as little as possible. Xping has gone the opposite way, choosing a very big and expensive electric car as their robo taxis. And don't forget, Xping is still bleeding money on selling electric vehicles. They have still not turned a profit. So like to pierce road Xping copies Tesla's vision only approach as well as steer by wire. Yet when it comes to the robo taxi they chose a traditional full-size SUV and even plan a six and sevenseater and it will be offered to consumers to own as well with steering wheel. Tesla went with a two-seater for the cyber cab as the vast majority of cab rides are for one or max two people and mass producing a smaller car is cheaper. Xping was able to go to market quicker as they can reuse a lot of existing manufacturing lines in their main Guangshu factory. But why option for a more expensive vehicle? Won't this allow others to underpric them on a large part of the robo taxi business in the future? Yes, this is exactly what I've been talking about in many of my videos about the Cyber Cap that it will be able to undercut everybody on price.
The Cyber Caps 165W hours per mile is equivalent to just 0.029 per mile in energy cost with 60 cents per kilowatt hour. Whimo Ipace is basically triple that. And the new Seeker van is about twice that. And I don't think this huge SUV from Xping is going to be any cheaper in energy cost.
So, the Cyber Cap will no doubt be able to undercut Xping on this very expensive robo taxi, both on cost to produce and cost to operate. But cool to see Xping hitting the street with their robo taxis and is going to be very excited to follow and see how they do because other automotive insiders in China still think Tesla is way out in front in capabilities. As Ray for Tesla reported, Lioto's autonomous drive head Kung Xang said in a recent interview that Tesla's full self-driving remains industry benchmark, praising its safety and capabilities. He added that Lee Auto's autonomous driving team is aiming to reach that level. The interviewer, an industry insider, offered an even more blunt assessment of full self-driving competing systems in China. We've consistently acknowledged Tesla's full self-driving as the benchmark. What I do want to criticize is the constant claim from some people that they already surpass Tesla when in reality the gap is still enormous. The only reason they dare to say that is because Tesla's full self driving hasn't fully entered the Chinese market yet. Once it truly rolls out, they'll be completely exposed. It will be like the imperson new clothes being stripped away in public. So, I still think that having a system that can drive everywhere on the planet, that is going to be the big differentiator in the end. And to be able to have an AI system that can do that, you need a lot of real world data. And Xpang still has a very limited fleet size collecting that data compared to someone like Tesla. But it is going to be very interesting to see when Tesla get approval for wide release in China, which should be very soon. Then we can start the timer and see when someone like Xpane will be able to deliver just as capable a system to customers in China. And another automaker that is trying once more to get into the autonomous game with a couple of failed attempts behind them is Mercedes. As Sawyer reported, Mercedes says it has plans to roll out its urban point-to-point autonomous driving system in Germany from the end of this year, aiming to be among the first car makers to offer this technology in the country, just like they were the first automakers to offer level three autonomy in the country. That ended up being a complete failure and have now been taken off the market. So, the system will be available in selected German cities from the end of 2026 before expanding nationwide from early 2027. Our ambition is clear. We want to continue to be at the forefront of international competition in assistant and autonom.
This includes providing people with the best possible support in complex traffic situations from traffic light to lane changes to dense urban traffic. This is exactly what our urban point-to-point navigation makes possible. Sounds great.
And now that they have taken a new approach with vision only with Nvidia's platform and tools, the chance of success of course is much much higher.
But I have seen people saying this is not Mercedes, this is Nvidia. Yes, the tools and the platforms are Nvidia, but Mercedes still have to do the heavy lifting, collecting all the data and train their AI system to drive. So, if they are able to get that to a nationwide release in Germany in early 2027, so the car can drive everywhere in Germany by itself pointtooint, just like Tesla's full self-driving supervised, I am going to be very impressed. And let's squeeze in the last short news topics into this new show. Yes, it's time for the Tesla shorts. Porsche is Porsching production for a full week in his Schuvenhen plant in Stoodgat due to weak demand for his electric tyan following an initial shutdown day in early May as part of a broader sales decline since the peak of 2023. Q1 2026 data show Porsche's deliveries fell 15% globally to 61,000 units and 21% in China with operating profit dropping around 22% amidst intense local EV competition and the end of certain incentive in key markets. And don't forget that 21% drop in China comes after 2025 where Porsche did drop 26% in deliveries as well. The situation highlights pressure on German premium automakers from slowing EV uptake in Europe, high production cost, and a strategic shift towards more profitable hybrids and combustion models to stabilize finances, which might be a strategic for short-term profits for long-term survival. And Whimo is doing weird stuff again as AI driver shared a picture of Whimo Ipace robo taxi in cyber cap golden collar and wrote, "Wow, gold paint. Such a novel an original idea for robo taxis. This just seems like a weird move from Whimo." Also, because the sensors are still white, making them stand out even more. And here is a nice example of how far you can go for €10 here in my country of Denmark. So, as you can see, no wonder that everyone living in a house will buy an electric car. Like me, I can charge my Tesla from 0 to 100 for around $6 to $7. Actually, yesterday we had so much renewable energy, I got paid to fill up my Tesla.
And actual smart summon is starting to work as intended. As Lawrence wrote, "Best Tesla actual smart summon ever pouring rain in Victoria, Texas. But no problem to logo luggage and self incomplete dryness." Nice. And Xiaomi continues to deliver more competitive versions of their two models. As Chan reported, Xiaomi just released the YU7 Standard Edition that cost the equivalent to $33,000.
It has a 73 kWh LFP battery from CL with 643 km of range and does a 0 to 100 km an hour in 5.9 seconds. Xiaomi is once again highlighting its competition with Tesla's Model Y with this new trim. Yes, this is a very compelling electric vehicle for a very compelling price. And don't forget their performant version of the YU7 just set a new record around the Nurburg Ring in Germany. And Tesla Asia shared a little post about the Tesla speed. Then breaking ground 2 years later shipping mega packs worldwide.
Yes, nobody does it quite as quick as Tesla. And the Boring Company shared a new video and wrote, "See how Prof continuously mines, pushing and building at the same time, the machines from Las Vegas to Dubai are remotely controlled from our Bastrop operation center. These advantages help the Boring Company deliver more miles each year in the battle against soul destroying traffic.
>> The best remotely operated continuously mining tunnel boring machine in the world.
Constructing a tunnel comes down to two tasks. First, pushing where we're actually digging the tunnel. And second, building, where we're constructing the concrete rings that make up the tunnel structure. Continuous mining means these two tasks happen simultaneously, meaning the machine never stops moving forward.
Here's how it works. We're starting at the front of Brew Rock. Just behind this wall, the cutter head rotates with enormous torque powered by these four electric motors. Tools on the cutter head gouge away at the earth and force material back to the buck chamber. A massive augur takes the material in the muck chamber and feeds it out onto the conveyor belt, which takes the muck up to the surface. As the cutter head digs, propulsion cylinders extend, pushing the machine deeper and creating more space for the next ring of tunnel. At the same time, the mining crew loads the liner truck, a giant remote controlled vehicle with concrete segments, each weighing 4,000 lb. When liner truck reaches the front, it is unloaded and a robotic crane picks up each segment and delivers it to the shuffler. The shuffler moves each segment up to the front and the erector picks it up and moves it into place. Most tunnel boring machines alternate between pushing and building.
Push and then build. Push and then build. Our tunnel boring machine pushes and builds at the same time. And this is where we can control everything remotely. Our global operations control center. From here, we can control machines from Las Vegas to Dubai. That's remote continuous mining. Muck flowing out of the screw, holes and cylinders extending, a new ring going up. All at once, all without stopping, all operated from the surface. Instead of mining miles per year, this will allow us to mine miles per week.
Just amazing how far the Boring Company has come and how many innovations they have done. Building miles of tunnels in weeks just make this so much more doable. building huge tunnel network under our big cities and Tesla once again showed off their community involvement at their Berlin factory as Tesla manufacturing shared 20,000 runners 1,000 just from Tesla.
Gigabeline once again brought the biggest team to the Berlin famine. Nice.
And Nick shared just how good a deal a Tesla has become in the US. The average price of a new car in the US in April 2026 was $51,600.
You can get a brand new Tesla Model 3 for as low as $37,000 and the Model Y for as low as $40,000.
Both of these cars will drive you around everywhere with a click of a button. It makes no sense to buy gas cars in 2026.
Yeah, a full self-driving car for $10,000 less than the average car in the US. That is just one hell of a deal. And we are seeing more and more people realizing just how much value for money you get with a Tesla compared to your average ICE car in the US. As Owen shared, it's true. My neighbor is a police officer. He bought a Model 3 after I showed him for self-driving. He didn't buy it because it's electric, but because it can take him home safely after an 18-hour shift, has incredible performance, and is cheaper to operate than a gas car. Yesterday, he told me five other officers in his department have now bought a Tesla in the 10 months since he got his after taking them on rides and talking about how much he loves it. Best car, period. and Musk economy shared X is launching X money soon. SpaceX S1 filing reveals X is preparing to launch a money product offering payments, banking and financial service including enable users to purchase goods, send money and access to other financial tools. SpaceX says it has already obtained payment license in the United States and other jurisdictions in preparation for the launch. Nice. And James reminded us all, two years ago, Elon destroyed the Tesla brand globally. It will never recover.
Today, number one selling EV in California, Tesla Model Y. Number one selling EV in the US, the Tesla Model Y.
Number one selling EV in Europe, Tesla Model Y. Number one selling EV in China, the Tesla Model Y. Yes, I made countless of videos about this topic, so I completely agree with James. But let add one more to the list. number one selling EV in the world, the Tesla Model Y. And to all the haters that has been telling me that Tesla was done in Europe, done globally, huge brand damage and demand problem. Well, I told you so. Just don't quite cut it. And Alexander said, "The Swedish Transport Agency most likely had to approve this, but Tesla's full self-driving with no major indigence after 15,000 km in Europe doesn't get approved." And yes, they have already passed 20 million now without a major incident in both Holland and Lithuonia.
But Sweden doesn't seem to want to approve this lifesaving technology. But only this autonomous bus that crashed on the first day on the job. And yes, I do mean lifesaving as we got another example of Tesla saving the day. Saving these two guys in Holland from being t-boned.
WHAT THE HECK? WHAT THE [ __ ] >> Holy moly. JESUS, MAN.
>> Holy [ __ ] What the hell?
>> Oh my.
Oh my.
Hey, best.
Honda.
>> What the hell, bro?
>> I don't think there is any doubt what both these two guys next car is going to be. That is all we have time for in this Overflow news show. Hope you enjoyed it.
If you did, don't forget to hit that like button. It really does help this video out. a lot so other people can find it on YouTube. And if you did like it, maybe you want to consider hitting that subscribe button and notification bell so you don't miss out on future videos just like this one and help this channel grow so we can reach more people about the EV revolution and the future of sustainable abundance. And if you're already a subscriber to the channel, thank you so much for the support. But if you want to support the channel even more, remember you can for as little as $1. Become a patron of this channel. To get a shout out, add some extra perks like early access, add free versions of my videos, exclusive video, check out Patreons to find out more. And don't forget to follow me on X. I post pretty much all the news as it comes out and a lot more. And check out the merch store to get some merchandise and support the show. It is also possible to support the show without becoming a member or a patron. There is a donation option down below or on my website bestinttesla.com.
You can also become an executive producer and get news articles, exclusive videos, behind the scenes, and access to all my research, charts, and spreadsheets also on besttesla.com.
And thank you for watching. And until next time, take care out there and be nice.
Hey, hey, hey.
Related Videos
OpenHuman VS Hermes AI: Who Wins?
JulianGoldieSEO
285 views•2026-05-29
Long-Running Agents — Build an Agent That Never Forgets with Google ADK
suryakunju
142 views•2026-05-30
5 Mind Blowing Omni Uses Cases
PaulJLipsky
1K views•2026-06-02
This computer is made from real human brain cells. And you can buy it.
Talktmsmedia
3K views•2026-05-28
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
I Made the Same Anime Fight Scene in Every AI Video Generator
NobleGooseAnime
295 views•2026-05-30
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
I Tested NEW Opus 4.8 on Four Projects (Updated LLM Leaderboard)
AICodingDaily
298 views•2026-05-29











