Welcome to Electrified, it's your host, Dylan Loomis, quick shout out to my new or updated patrons, hails and jaden. Thank you for choosing to support the channel. 15 months after Tesla first applied to expand Giga Berlin, they have finally gotten approval for the first stage of that expansion. In the end, when the entirety of this approval is done, Giga Berlin should be able to pump out 1 million vehicles per year up from the 500,000 stated capacity currently. But approval of the first phase will only cover the construction of a large haul, which will house storage areas, a battery cell testing lab, and expanded logistics infrastructure. All of the new construction will take place on land already owned by Tesla.
Don't forget what plant manager Andre Tereg said earlier this year. We will not spend several billion on the expansion of the factory without the signals being clear that the market is asking for this. We firmly believe that the market will pick up again, it's certainly a question of how quickly and when. So from that standpoint, this expansion going ahead seems bullish. However, we have to remember what has changed this year, that being the large tariffs for EVs coming from China. Tesla does get a special rate that's lower than some other automakers, but the number is still not zero.
So honestly, even if demand for Tesla across the EU was stagnant, I think these new tariffs would be incentive enough for Tesla to move forward with this expansion, localizing more of that production for the European market. A local source also confirmed that the expansion will also increase the production of battery cells at Gigabrelin from 50 up to 100 gigawatt hours per year. But keep in mind, right now Tesla is not hitting either of these stated capacities, whether it's for battery cells or vehicles. Right now, most of the battery production at Gigabrelin is just certain materials that are being shipped to the United States for final assembly, which is thanks to the IRA incentives.
This source also added that there will also be changes in optimization of the production facilities already built at Gigabrelin, and that for phase one, no additional clearing of the forest is necessary. Also part of this phase one expansion are some temporary facilities like a hazardous materials warehouse, a distribution center, and a company medical center. Gigabrelin produced less than 200,000 model-wise in 2023, and they're likely to be under 250,000 for 2024.
It'll likely take a few months for things to really get moving at Gigabrelin, but over the next six to 12 months it'll be interesting to watch how the export number from Gigashang High changes and how likely the production number at Gigabrelin increases. Just to make it clear from RBB24, a German source, they said the first partial approval does not provide for any new production facilities. The expansion of production and the construction of additional production facilities are the subject of a second and third approval process. Speaking of Gigabrelin, the Tesla Semi is making a guest appearance at the factory.
In this short video, Dan Priestley said there's a huge amount of potential they're seeing in the market in the surrounding areas of Gigabrelin. Bloomberg talked to some Tesla employees after the Wii robot event, and we learned a few new things. The employees said the Optimus prototypes were able to walk without external control using AI. This should come as no surprise, they said originally Optimus wasn't due to feature in last week's event at all. Elon had requested their inclusion about three weeks prior.
The late notice meant that there wasn't time to get the software up to speed making remote operation necessary. They said Elon had been due to discuss a wider range of autonomous driving related items on stage. The presentation was going to include more details about the plan for autonomous tech for Tesla's Semi, details of the app behind a future ride hailing service, and pricing for software suites. But Elon either skipped over the material or removed it from the event. So why would Elon choose to skip over that material or remove it from the event? Let your speculation run wild, but I think the safest assumption is what we talked about yesterday, he's just keeping his cards close to his chest.
There are some headlines out there about this Anthony Leventowski who made some positive comments about Tesla's FSD and its competitive advantage, but I think we need some important context on who this guy is. Now look, everybody has things in their past, but Anthony was sentenced to prison for 18 months for stealing self-driving secrets from Waymo. He took that info from Waymo, then started his own self-driving trucking company called Auto, which he eventually ended up selling to Uber. There was a public admission he said he came to terms with what he did.
It was actually Anthony selling this auto business to Uber that allowed him to join Uber as a high ranking exec. But then Waymo sued Uber for trade secret theft and eventually the two parties settled. But Leventowski was still on the hook and he eventually ended up entering into a plea agreement. And after all that, Leventowski actually sued Uber for just over $4 billion claiming that they owed him money for the prior deal. So with all of that context, Anthony said that Elon holds an incredibly strong position when it comes to FSD right now. I'd rather be in Tesla shoes than Waymo's. One key reason for this sentiment, the data. He said that's ultimately what's really going to be the differentiator here that you have the richest, most consistent data to continuously improve over time. I would say Tesla has examples of maybe 10,000 or maybe a million times more data than Waymo does in terms of all other scenarios of driving.
Responding to Anthony's statements, Elon chimed in saying true. And despite his background and legal troubles, this guy was a hotshot engineer in the self-driving space for a number of years. He's not really telling us anything we don't already know, but hopefully the public starts to pick up on Tesla's lead as well. Uber's CEO just said that going up against Elon can be tough, so of course he's opening to partnering with him. He said that he would love to offer Tesla's Robotaxes on Uber's ride-hailing platform when they're ready to go. Obviously, competing with Elon is no easy matter and we take nothing for granted, we really want to partner with the autonomous industry. He sees three potential paths for Uber once Tesla launches its Robotaxes, competing with Tesla, working with Elon, to have autonomous vehicles on the Uber network or combining the two. He said, we'd love to have it on the platform, but if not, I don't think this is going to be a winner take all marketplace.
We believe in the spirit of partnership, we'll see what Tesla does. I see a lot of people out there saying that Tesla should never partner with Uber, they don't need them, and I totally understand that argument. The problem is we don't know what the partnership would actually look like. Just for example, let's say Uber said you know what, we'll buy all of the cybercabs from Tesla at cost, and we'll take care of all of the cleaning and maintenance and operation of the fleet. In return, all Uber would ask for is a small fee per mile. In that scenario, there's not really much downside for Tesla and there is a lot of exposure that may take Tesla a bit longer to get on its own given the network that Uber already has. Now of course, personally, I would rather see Tesla do everything in-house to control the entire experience and keep all of the revenue for itself. But to argue that there's no partnership that Uber could come up with that would be beneficial for Tesla, I don't think is true. And just to be clear, I also think it's true that Uber needs Tesla way more than Tesla needs Uber, but that doesn't mean that Tesla can't benefit from working with Uber in some capacity. Tasha Keeney from ARK shared some thoughts on Wii Robot and she said Elon mentioning 30-40 cents per mile at scale, which isn't far from what ARK was predicting, compares quite well to the competition. She said the average cost per ride-hailing mile in Western ride-hailing markets is around $2.
Translation, right now the competition for a 10-mile trip would be around $20, whereas Tesla's solution would be around $4 for the same trip if of course they can hit those numbers. Tasha also shared this chart, it just shows us the number of miles in trillions when it comes to the addressable market at different price points. For example, at $1 per mile, the addressable market for ride-hail is right around 1 trillion miles. And again, ARK is saying the existing addressable market for ride-hail companies in Western markets, they charge between $2 and $4 per mile. It's not the best chart ever because you can barely see certain price points like the purple, which is $4 per mile and the black, which is $3 per mile. They're incredibly slim right over here. The first one you can really see is gray at $1 in 10 cents per mile, the light blue is 60 cents and the yellow is 50 cents.
Let's just say Tesla only gets the price down to 50 cents per mile, that's a $2.75 trillion addressable market, which of course is orders of magnitude greater than the addressable market at $2 per mile, which is where a lot of the competition is now. Just a heads up, going back to Tesla's AI schedule that they did incredibly well with when it comes to September milestones, in October, we're supposedly getting Unpark, Park, and Reverse in FSD and version 13, where they're touting a 6x improvement in miles between necessary interventions. The chatter at WeRobot from everything I've been reading is that the Tesla employees were very bullish on version 13 and as we know, these are some of the last pieces for the full suite of FSD, handling parking lots, the last mile, and reversing to get out of sticky situations.
For whatever it may be worth, the early reviews of 12.5.6 have been quite impressive from what I'm seeing. Is the Tesla gonna give them room? See how this traffic starting to stack up here? Right about now, I'd be blinkering over and getting in this middle lane, but I don't know what to say guys, it is doing what I'm thinking. Are you listening? Maybe you're listening to me, but here I am in chill mode. I didn't change my speed much, but I definitely gave those cars room to merge. Didn't give me any sort of alerts, didn't tell me what it's doing anymore, it's just doing it.
Chris from Dirty Tesla reiterated the excitement around version 13 as he was at the event, and he said the feeling is that banish and reverse come together to really complement the full robotaxi feel for customers current cars. A question we've been asking, it does not seem the cars learn in between updates yet. Tesla is not changing weights, flipping toggles, or doing anything between you and installing updates. Map data can update and change in between software updates, but outside of that, not much is happening without an update. And finally, the team does know about the speed control problem, and they said a fix is coming soon.
On X, James Damo was having a back and forth with Jordan from the limiting factor, and he said some things worthy of sharing. James said Tesla's AI work is aggressive, but also professional and prudent. In contrast to media coverage of FSD, it's not a wild high stake scandal. Quite the opposite, it's a near inevitability given the needed investment, which is why it warrants the investment. We don't know the final form in detail yet, and things like public reaction, price elasticity, and the adoption rate are hard to predict. But if it were to not happen, that would be a shocking development. FSD happening, and happening on hardware 3 should be the unsurprising outcome.
Later, he said the inevitable success of FSD should not be surprising or unexpected. It's the fear-mongering horse race dynamics of media that promotes the illusion that FSD is somehow unlikely. James said my conviction comes from my familiarity with the underlying tech and its development trends in thousands of hours actually using the software. The noise level on publicly available stats is too high to be informative. Holmars also shared a clip of end to end on the highway. 12.5.6, and I have to take an exit. We're now less than one mile away from the exit. It says 0.9 miles until the exit, and watch what the car does. Does it slow down and get behind this car? No, it actually speeds up and gets in front of the car, and then changes lanes.
And do you think it's going to miss the exit now? Because it waited so long, it waited until the last mile to take the exit? Well, think again, it sped up, got ahead of the car, and still had time to make its way across all four lanes and take the exit just fine. I know that's not the most impressive clip you've ever seen, but something I wish FSD would get over to the necessary lane a bit sooner because I do tend to get anxious that it's actually going to miss the exit or the turn. I am curious though, am I the only one? DHL recently put out a press release about its experience testing the Tesla Semi in the real world. The results? Over a two-week trial this summer, DHL took the Tesla Semi over 3000 miles of normal operations in Livermore, California.
The trial included one long haul of 390 miles, fully loaded with a gross combined weight of 75,000 pounds. During the trial, the vehicle averaged 1.72 kWh per mile, operating at speeds exceeding 50 miles per hour on average for over half its time on the road. The result exceeded our expectations and even Tesla's own rating. Our verdict? The Tesla Semi is ready for prime time. Just one more source saying that the Tesla Semi can get up to 500 miles of range in the real world, fully loaded. Following the successful trial phase, DHL is now carefully planning how we could best integrate these electric trucks into our network as Tesla gears up for large scale production of the Semi in 2026.
Tesla is hiring for an unreal software engineer for virtual reality teleoperation for the Tesla bot. Again, I know a lot of people are pushing the narrative that Tesla will have a business line doing teleoperation, anything is possible, but this job specifically is mainly for data collection. And just to be clear, Tesla has had job postings similar to this in the past. Earlier this year in February, they had an AR-VR telemetry software engineer job listing. And Ashish, who works at Tesla, said what people missed about the around 20 human assisted optimi is Tesla's intent and ability to scale up data collection around 20 times more than 2022 AI day language images, audio are blessed with exact data. So even though they were human assisted, they were still gathering valuable real world data.
Cox Automotive put out its estimates from Kelly Blue Book. I think that's an important distinction, but they had EV sales for quarter three in America at 8.9%, the highest level recorded and an increase of 7.8% in quarter three last year. After seeing sales fall year over year in the first half, market leader Tesla returned to growth mode in quarter three with sales up 6.6% thanks to the Cybertruck. In fact, Cybertruck outsold every other available EV except two, the Model Y and the Model 3. Their estimate has Cybertruck sales for quarter three at 16,692. Meanwhile, the F-150 Lightning 7,162. The Mach-E 13,392. The Hummer Truck and SUV 4,305. Chevy Silverado 1,995. And the Foundation Series Cybertruck is still selling more than Chevy's more affordable equinox EV at 9,772.
Cox Automotive发布了来自Kelly Blue Book的估算数据。我认为这点很重要,他们估计美国第三季度的电动车销量达到了8.9%,这是有纪录以来的最高水平,比去年第三季度增长了7.8%。在经历了上半年同比销量下降后,市场领导者特斯拉在第三季度重回增长模式,其销量因Cybertruck上升了6.6%。实际上,除了Model Y和Model 3之外,Cybertruck的销量超过了其他所有可用的电动车。他们估计第三季度Cybertruck的销量为16,692辆。同时,F-150 Lightning售出了7,162辆,Mach-E售出了13,392辆,悍马卡车和SUV售出了4,305辆,Chevy Silverado售出了1,995辆。基础系列的Cybertruck销量甚至超过了更实惠的雪佛兰Equinox电动车,后者售出了9,772辆。
Lucid has partnered with four seasons hotels starting this fall and guests at select properties in the US will have access to complimentary Lucid Air EVs during their stay. The program is set to expand to locations in Canada, Europe, and the Middle East in the coming months. Arizona now has a new, biggest battery storage project up and running. This one is 1.2 gigawatt hours. But I'm sharing because meta confirmed the green energy would be used at a data center in Mesa, Arizona. And the American president of Fluent said we're seeing some of that come into the space where the power consumption of these large data center customers, coupled with their desire to have green power, is driving some of the storage and renewable business quite a lot in the US. Arizona has one of the highest growth rates of electricity in the country due to the surge in data centers and the reshoring of American manufacturing.
Tesla Optimus on X shared this video of the cybercab with a little infographic and inside it looks like the cybercab will be using 4680 battery cells. And they're saying the same batteries, the same cameras, and the same computer will go from the cybercab directly into the Optimus as well. So could the cybercab use LFP4680s made by Tesla? I think there's certainly a chance. Tesla stock closed the day at $219.57 up 0.19 percent while the NASDAQ was down 1.01 percent. It was a lower volume day for Tesla trading about 19 million shares below the average volume the past 30 days. Hope you guys have a wonderful day please like the video if you did you can find me on X linked below and a huge thank you to all of my patreon supporters.