Welcome to Electrified. It's your host Dylan Loomis, quick shout out to my newest patrons, media cafe online and Vance M. Thank you for choosing to support the channel.
As we get closer to the Cybertruck launch event, I'm pretty sure they will start slowly rolling out new wraps and hopefully some Cybertruck accessories. We know they have their own in-house team working on specific accessories for this vehicle, and unless they don't want to complicate early production, which is definitely a possibility, if Tesla was smart, that's what they would do, slowly start showing the public what type of accessories they've been working on. And while many of us were excited to finally see the matte black Cybertruck, which was most likely a color PPF wrap, the search volume on Google still does not compare to when Elon went on the Joe Rogan podcast.
So here we have it, X's new AI platform called Groc or Groc, whatever you prefer, to possibly be integrated natively in Tesla vehicles using local compute in the future and Tesla vehicles to run inference for Groc queries when the robot taxi fleet is not in use driving people around. I think the Tesla community can tend to get over their skis when it comes to quasi announcements like this. So is this a future billion dollar business for Tesla or is it something that may not actually ever happen at all? Rather than cop out with the answer being somewhere in between, which by the way it most likely will be, here are a few things to consider.
Internalized the first line here, provided our vehicle AI computer is able to run the model. For now, there are no guarantees Tesla's FSD hardware will be capable enough to run Groc, which then begs the question, would it only be cars with hardware 4 that may be able to run Groc? What about hardware 2.5? Is it bound here in terms of what percentage of Tesla's fleet might be able to do this in the future?
So sadly I'm going to sound a little bit like Wall Street here, but it's actually very tough if not impossible to put numbers on this right now, it's just far too soon. When Elon says Tesla will probably have the most amount of usable inference compute on Earth, bear in mind using FSD hardware when the cars are sitting idle will indeed use some level of energy. How would a Tesla owner be compensated for letting Tesla remotely run inference compute on Tesla vehicles? This would certainly result in an accelerated type of phantom drain, only this time there would be some type of return to the customer for this energy use.
And there are more questions, how much energy would running inference require? Would the energy drain be worth the presumably small amount of compensation that Tesla or X would end up offering to the customer? The customer being the owner of the Tesla vehicle. To make sure everybody's up to speed, inference compute is basically the process of answering questions that people would ask AI like Grock. It's the process of making predictions or decisions based on data. This is entirely different than training an entire model. So think about simple questions like what's the most popular restaurant in Austin, Texas. Tesla cars may in the future be able to use the FSD computer to spit out an answer to a query like this and of course many others.
SETI-S-E-T-I is the search for extraterrestrial intelligence and Elon did clarify he did not mean Tesla vehicles were going to be looking for aliens when they're sitting idle, at least not yet. I do want to reiterate, all of this is in such an early stage that drawing too many conclusions is a fool's errand. Grock itself is still in very very early development and what it will eventually become is not yet fully known. However, thinking 5 to 10 years down the line, things do become incredibly interesting. You know how Siri and most voice assistants are pretty frustrating at times? Well, think about having a conversation with your car, in your car, being a Tesla, asking it questions and getting real-time news and answers because Grock references the entire X library of content with news in real time. This would also give Elon, Tesla and X even more data and more of a pulse on what customers are inquiring about.
Even just the idea of using Tesla's future fleet of vehicles for any type of distributed computing, whether it be Bitcoin mining or something else, those applications become intriguing opening up a whole new world of possibility. But there's also a chance they're never really used in any meaningful way in that capacity. We do know the rest of the industry is struggling with software and the idea of smart connected autonomous vehicles.
Grock running natively in Tesla's could catapult Tesla even further ahead in this realm, giving customers a potentially very useful tool to use with voice commands in the future, anywhere on the road. To have a built-in voice assistant with real-time data from X natively in your Tesla could bring an entirely new realm of productivity and content consumption. Could Tesla charge a monthly or annual fee for the ability to run these Grock queries? I could speculate here for a while but here's my takeaway. I'm not ready to say this is a new multi-billion dollar business for Tesla but I'm definitely not ready to write it off as Hopium either. I am however very excited to watch this develop in the next few years because this definitely could turn into a major moat for Tesla relative to the competition. So we can absolutely go ahead and add this to the growing list of home run potential ventures for Tesla.
As Grock continues to develop we will pay close attention but the Tesla integration natively in the next few years paired with Robotaxes could become quite the selling point for the public and yes Tesla's distribute compute capability could eventually be monetized in ways we have not yet even really thought about. What a time to be alive.
Clean Technica shared a chart of the top selling EVs in Europe. This was just for one month for September of this year but it's still fun to look at the Model Y leading the way. Outselling places 2 through 5 combined. And we have VW's ID4 and ID3 combined doing less than 10,000 units for the month.
I believe some of you in Europe have already seen these updates now for a few weeks but in case you haven't Europe is supposed to be getting some of these FSD visualizations bringing them over to just autopilot. Those improvements like the dynamic vehicle sizing and open door visualizations should be rolling out over in the EU.
We got an update on the Tesla strike situation over in Sweden. We have this source Dagen's industry saying the negotiations between Tesla and IF Mattel have completely stalled. Now the union is flagging for more strike action saying there will be more sympathy measures. These sympathy strikes are where other companies not directly Tesla related may actually get involved in the strike to limit Tesla's business operations in the region. Basically as a sign of solidarity because around 90% of Sweden is unionized. What I could find online the union believes Tesla is engaging in strike breaking. The word on the street I can't confirm is that Tesla has brought in people partly from overseas, partly from other locations in Sweden to cover for the members that are striking. So basically Tesla doing whatever they can to maintain the business operations. This is why the other unions are now saying there will be more sympathy measures. Right now as far as I can tell there are no more scheduled negotiation times and there are no compromises in sight. A union rep said I don't know what that would look like. We want a collective agreement. They Tesla say they don't want it. When asked if Tesla had planned any countermeasures like leaving Sweden they said no we have not heard of that.
Tesla has actually settled with a UK owner who bought a Model 3 back in 2019 and he ended up paying around 6000 quid for FSD but he sued Tesla saying he wanted his money back because Tesla never delivered on the service. The initial settlement had clauses including a non-advice clause and a confidentiality clause but the customer wasn't having it because he wanted to be able to share his story about what happened. Tesla then agreed to remove the clauses and the case was resolved with the customer receiving the full claim amount of 5800 quid and extra compensation of 2200 quid basically for the interest he could have earned over that time. So let's call this roughly 10,000 US dollars that Tesla had to pay back.
There are certain locations globally where Tesla has been accepting payments for FSD but it hasn't really delivered the service whether it's because of a software issue or something to do with regulations. Naturally the question becomes now that this settlement is public will there be a line up of other people looking to do the same thing, get their money back from Tesla. Let's just pick a number 50,000 people that may hypothetically be in a similar situation where they paid for it and haven't got what they were expecting. That would roughly be potentially about 500 million dollars that Tesla would then have to pay back potentially reducing some of their FSD revenue that they're expecting to recognize in the next few years.
That of course just a quick and dirty thought experiment of what a potential worst case scenario could look like but in the months ahead we'll see if any of these arise. Here we have a quote over the weekend from a veteran auto analyst saying the way Toyota builds cars has been considered the standard but it's extremely shocking to think that what Tesla is proposing is likely to become the standard for producing EVs. The impact on Japan's car manufacturing will be monumental. He was of course referring to the Gigacast method. An exec from a European automaker added the way Tesla is making cars is quickly moving to become an industry standard. FT also shared an infographic on Tesla's older serial production process and then contrasted it with what they're looking to do with the modular or the unboxed approach.
We've heard about this all before but little things like previously they would have to put the doors on, then paint, then take the doors back off to go to the next step and from a first principles approach it was just inefficient. Compare that to what Tesla will be looking to do with the next gen platform and they said the Gigapress castings allow floor assemblies to become part of the structure of the car, making modular units that more people or robots can work on at the same time in parallel sub assemblies. So if you want a gross oversimplification you have all of these outside areas we can call them pods where you can do all the work, do all the painting, assemble these different pods and then you ultimately just put them together like Legos. Toyota's chief production officer said to be honest we are behind in Gigacasting since we've yet to roll out a product. But we have done casting for a long time and ultimately we believe we can produce a product that will be cheaper, lighter and thinner with high performance.
Yet again we have people concerned saying that the big question will boil down to repair ability when these vehicles made with a Gigacast technology are actually in accidents. However if you've been following along daily with Electrified you will remember we talked about Tesla's Gigacast are actually modular in themselves. So if a certain part is actually broken or damaged you can remove individual pieces that are actually replaceable. Yet another Tesla innovation that Legacy Auto is going to have to figure out. In summary Tesla has gone from any relevant golf cart maker to literally revolutionizing the entire global auto industry. Not too bad.
On X Edge shared a few recent images from November of the site where Tesla is building its new diner in California. You'll see a few wall structures are indeed up. So the work is underway. I said it last time we talked about this but I'll say it again. This site just doesn't look that big to have 32 plus stalls and this big restaurant that we were all expecting. Is it just me?
On X Edge最近分享了一些11月份特斯拉在加利福尼亚建造新餐厅的工地的照片。你可以看到一些墙体结构确实已经建起来了,所以工程正在进行中。上次我们谈论这个话题时我说过一次,但我还是要再说一次。这个工地看起来并不像我们都期待的那样大,能够容纳32个以上的摊位和这个大餐厅。难道只有我这样觉得吗?
This is a fun little story we have Tesla sponsoring an engineering team at the University of Manitoba. Every year this team builds electric race cars that are then entered into competitions. This time around Tesla is giving them 2170 sales for them to use in their next design. But in my opinion the coolest part of this story Tesla will also be offering consulting services to their engineers about the design and allow them to get some Tesla expertise. Which one of the students said is invaluable to the team. The team said they were one of only 67 that were chosen worldwide for this Tesla sponsorship. The word on the street is the public will also be able to buy some Tesla merch in the lobby through some vending machines. I believe this was the first day it was open to the public so most likely over the next week or two we get a bit more information.
Tesla Chan on X shared some takeaways from a recent Taolin interview. We've already heard about 1 through 3 but some news she said currently a car can come off the production line in about 30 seconds. The last time we heard this I believe it was somewhere in the neighbourhood of 40 seconds. 10 seconds may not sound like a whole lot but it's actually a 25% operating efficiency improvement. Automation in the welding workshop now close to 100% Gigacasting reduces manufacturing time from 1 to 2 hours down to 3 to 5 minutes which doesn't even make sense and the Megapack factory in Shanghai will indeed export to the global market. We've heard that the cost competition for battery energy storage in the Chinese market is as intense as anywhere on the planet so it may make sense for Tesla to start exporting the Megapack to the European and other regions. And of course as Tesla scales production for its Megapack in time with higher volume they should be able to drive down costs and then maybe become more cost competitive in the Chinese local market.
And when people ask how Tesla is actually making EVs profitably of course there are thousands of reasons not the least of which is relentless continuous improvement. American advisors have said nearly half of the dealers that were queried indicated they have no trust in Ford and Toyota ended up as the most trustworthy brand when it comes to dealer relations. 48% of dealers said they have no trust in Ford consistent with the expectation of a decline in future Ford franchise profitability due to the OEM's EV future retailing strategy. Nissan not far behind though as they followed Ford with 43% of dealers saying they have no trust in the brand. Meanwhile a whopping 72% said they have a high level of trust in Toyota. This of course begs the question though is this a case of the blind leading the blind to which I would say most likely it's a resounding yes. I will say you have to be careful with these numbers because obviously the dealers are not going to want to change their business model they've been making historically high profits nobody wants to go and learn an entirely new vehicle of the EV. They have to spend money on charging and infrastructure and training so naturally any OEM that tries to get the dealers to onboard with EVs there's going to be some level of pushback. Just another messy situation that Tesla does not have to spend one second thinking about.
According to L. Economista Tesla has established a holding company in Barcelona Spain. You may recall that Tesla was in negotiation talks to open a $5 billion Gigafactory in Spain but then some of the details of those plans were leaked and Tesla said nope not doing it anymore. The reporting said the purpose of this holding company would be acquisition holding enjoyment administration direction and management of securities and or shares of Spanish or foreign companies or entities residents in Spain or abroad so definitely too broad to make any real assumptions. Tesla's cooking up something in the Spanish market I'm sure we'll find out soon enough.
Previously GM had about $11 billion earmarked through 2028 in US investments for electric vehicles and powertrains now that number over the weekend has been boosted by about $2 billion to 13. GM is expected to build a more affordable version of the Chevy Bolt in Kansas and a new series of premium EVs for Cadillac and Chevy in Michigan. GM is considering whether to build this new version of the Bolt EV at its Fairfax plant in 2025. They're also thinking about a pure electric performance model with a Corvette name. This is all still evolving as over the weekend GM and the UAW were still working on ratifying these new contracts. But they're talking about five vehicle assembly plants and seven parts plants to build EVs and components in the United States over the next five or so years.
The word is the new Chevy Bolt would fall under the master UAW contract. Lucid has finally given in today announcing that their customers will be able to access 15,000 Tesla superchargers in North America in 2025 so Lucid is adopting the NACS after all. Their current vehicles with CCS ports will be able to access the supercharger network with an adapter but not until 2025.
These remain how the V3 superchargers will work with Lucid and how the voltage step up is going to be handled but I'm sure we'll find out in time. Here's the updated NACS list remaining holdouts, VW, Stellantis and Mazda.
In response to the news, you have to love the humor from Elon saying that must have been a bitter pill to swallow.
针对这则新闻,埃隆的幽默说法相当可爱,他说这一定是一颗难以下咽的苦药。
You can find me on X linked below. Please like the video if you did. Hope you guys have a wonderful day. And a huge thank you to all of my Patreon supporters.