Tesla conference call: Tariff costs increase by $300 million, Musk warns of a "difficult quarter" ahead, but firmly believes Robotaxi and Optimus will create greatness

Tesla delivered its worst quarterly report in recent years, with the company's second-quarter revenue plummeting 12% year-on-year to $22.5 billion, marking the largest decline in at least a decade. On July 23 local time, CEO Elon Musk admitted during the earnings call that due to sudden changes in EV tax credit policies and tariff impacts, the company may face "difficult quarters in Q4, Q1, and possibly Q2." Despite the pressure on performance, Musk reiterated his confidence in the autonomous driving and humanoid robot business, predicting that the RoboTaxi service will expand to cover half of the U.S. population by the end of the year, and the goal for the Optimus humanoid robot is to reach an annual production of 1 million units within five years. The leader firmly believes that "Tesla will eventually become the most valuable company in the world." The company's stock price fell 5.2% during the earnings call and has dropped about 18% year-to-date. Analysts pointed out that investors' views on Tesla are becoming increasingly polarized: those who see it as a traditional automaker are disappointed with its performance, while those optimistic about its AI and robotics prospects remain positive. Here are the key takeaways from Tesla's earnings call: • Due to the cancellation of tax credits and tariffs, Tesla acknowledges it may face "several difficult quarters," with challenges expected in Q4, Q1, and possibly Q2; • CFO Vaibhav Taneja disclosed that tariffs will lead to a $300 million increase in costs; • Tesla successfully launched its autonomous RoboTaxi service in Austin and plans to significantly expand the service area to ten times that of its competitors within weeks; • The company expects to launch its autonomous ride-hailing service in areas covering about half of the U.S. population by the end of the year, with regulatory discussions progressing in regions such as the Bay Area, Nevada, Arizona, and Florida; • The design of the third version of the Optimus humanoid robot has been completed, with plans to launch a prototype by the end of the year and begin mass production next year, targeting an annual production of 1 million units within five years; • The adoption rate of FSD has increased by 25% since the release of version 12, but about half of Tesla owners have never tried the feature; the company's vehicle safety report shows that vehicles using FSD are ten times safer than those not using it; • The company expects its capital expenditures to exceed $9 billion this year, primarily directed towards AI, robotics technology, and manufacturing capacity expansion. RoboTaxi Expansion Accelerates, Covering Half of the U.S. Population by Year-End Tesla successfully launched its first autonomous RoboTaxi service in Austin. According to Ashok Elluswamy, the director of autonomous driving software, the service has driven over 7,000 miles, received positive customer feedback, and "so far there have been no significant safety-critical incidents." Elon Musk stated that the company is accelerating its expansion pace and expects to significantly expand the Austin service area within weeks, with operational areas far exceeding competitors. Tesla is seeking regulatory approvals in places like the California Bay Area, Nevada, Arizona, and Florida, with the goal of covering half of the U.S. population by the end of the year. "The service area and the number of operational vehicles will grow at an exponential rate," Musk said during the conference call. He anticipates that the RoboTaxi fleet will "grow from a tiny scale to a massive scale in a relatively short time" and expects this business to have a significant impact on the company's finances by the end of next year. Policy Headwinds Hit, Tariff Costs Increase by Approximately $300 Million Tesla CFO Vaibhav Taneja warned that the policy changes brought by the Inflation Reduction Act will have a near-term impact on the business. Most importantly, the $7,500 EV (electric vehicle) tax credit will be eliminated at the end of this quarter. "Given this sudden change, our vehicle supply in the U.S. is limited this quarter," said CFO Taneja. The profit support that Tesla has long relied on—regulatory credits—saw a decline of over 26% in the second quarter to $439 million, a significant drop from $595 million in the first quarter and $890 million in the same period last year. With the Trump administration lifting penalties on automakers that do not meet federal fuel economy standards, this revenue is expected to decline sharply. The impact of tariffs is also becoming evident. Taneja revealed, "Quarterly, tariff costs have increased by approximately $300 million, with about two-thirds affecting the automotive business and a smaller portion affecting the energy business. However, considering the lag in invoicing and sales, the full impact will be seen in subsequent quarters, so costs will rise in the short term." Facing these challenges, Musk acknowledged, "Due to the negative impacts of the legislation and tariffs, our business faces short-term challenges... We may have a few tough quarters." FSD Adoption Rate Increases, "Vehicles using FSD are 10 times safer than those not using it" According to Taneja, since the launch of FSD version 12 last year, the adoption rate of FSD in North America has grown by 25%. The vehicle safety report released by the company shows that vehicles using FSD are 10 times safer than those not using it. However, Musk admitted, "The vast majority of people are unaware of its existence," and even half of Tesla owners have never tried the feature. He compared FSD to "a cat that can sing and dance," saying, "Before you see the cat sing and dance, you would think it’s just an ordinary cat." The company plans to engage users through service centers, send videos, and prompt users to try the FSD feature in specific driving scenarios. Musk stated that even without considering other factors, the $99 monthly fee for FSD "is equivalent to getting a personal driver for just $3.33 a day." Optimus Robot Enters Key Phase, Aiming for Annual Production of One Million Units in Five Years Elon Musk revealed that the design of the third version of the Optimus humanoid robot has been basically completed, with no major defects, and a prototype is expected to be launched by the end of the year, with mass production starting early next year. He stated, “We will try to scale up the production of Optimus as quickly as possible, aiming to reach an annual output of 1 million units as soon as possible. Do you think we can achieve this in five years? That’s my guess.” The Tesla Optimus lab currently “looks like a set from Westworld,” with robots undergoing 24-hour continuous walking tests at the Palo Alto office. Musk anticipates, “We will transition from a world with few robots to a world where robots are so common that you won’t even look up to notice them.” Energy Business Grows Against the Trend, Surge in Storage Demand Despite facing tariff and policy challenges, Tesla's energy business achieved “the highest gross profit to date” in the second quarter. Musk explained that the continuous power output of the U.S. grid is about 1 terawatt, but the average usage is less than 0.5 terawatts. “If we increase battery capacity, we can keep power plants running at full load 24 hours a day, thereby more than doubling the annual energy output of the U.S.” As the growth of AI and data centers drives rapid increases in energy demand, customers are beginning to realize that energy storage is “the fastest way to scale energy.” The company expects strong growth in deployment volume in the second half of this year, with the first LFP battery manufacturing plant set to go into production by the end of the year, and a third Gigafactory to be launched near Houston in 2026. The transcript of Tesla's earnings call is as follows (translated by AI tools): Tesla Q2 2025 Earnings Call Company Participants \* Ashok Elluswamy - Director of Autopilot Software \* Elon Musk - CEO \* Lars Moravy - Vice President of Vehicle Engineering \* Travis Axelrod - Head of Investor Relations \* Unknown Speaker \* Vaibhav Taneja - Chief Financial Officer Other Participants \* Adam Wood - Morgan Stanley \* Analyst \* Dan Levy - Barclays \* Emmanuel Rosner - Wolfe Research \* Mark Delaney - Goldman Sachs \* William Stein - Truist Securities Presentation Session Travis Axelrod {BIO 24033559 } Good afternoon everyone, and welcome to Tesla's Q2 2025 Q&A live stream. I am Travis Axelrod, the head of investor relations. Joining me today are Elon Musk, Vaibhav Taneja, and several other executives. Our Q2 results were released around 3 PM Central Time in an updated document, and the release link is the same as this live stream. In this conference call, we will discuss our business outlook and make forward-looking statements. These comments are based on our forecasts and expectations as of today. Due to various risks and uncertainties, including those mentioned in our recent filings with the U.S. Securities and Exchange Commission (SEC), actual events or results may differ significantly from those anticipated. During the Q&A session of this conference call, we ask each questioner to limit themselves to one main question and one follow-up question. Please use the "raise hand" button to join the question queue. Before we move into the Q&A session, Elon has some opening remarks. Elon? Elon Musk {BIO 1954518 } Thank you, Travis. This quarter has been very exciting for us. We successfully launched the Robotaxi service, achieving our first driverless rides with paying customers in Austin. Some of you may have noticed that we have expanded the service area in Austin. It’s bigger and longer now. And it will get even bigger and longer. We expect to significantly expand the service area in Austin, far beyond our competitors. Hopefully, this will happen in about a week or two? (Background sound: Yes, a few weeks.) Elon Musk {BIO 1954518 } About a few weeks. And we are obtaining regulatory approvals to launch services in the Bay Area, Nevada, Arizona, Florida, and many other places. So once we get the approvals and prove our safety, we will roll out the driverless ride-hailing service across much of the U.S. I believe by the end of this year, we may offer driverless ride-hailing services in areas covering about half of the U.S. population. That is at least our goal, subject to regulatory compliance. I think it is technically feasible. So, assuming we get regulatory approval, by the end of the year it could cover half of the U.S. population. But we are very cautious. We do not want to take any risks. Therefore, we will proceed cautiously. But the service area and the number of operational vehicles will grow at an exponential rate. Another noteworthy thing is that the Model Y became the best-selling car in Turkey, the Netherlands, Switzerland, and Austria in June I believe it is still the best-selling car in the world, regardless of type. The autonomous driving capability is one of the important factors. So, even without supervised driving, even if it is just supervised self-driving, it is still a huge selling point. It is worth noting that we have not yet obtained supervised Full Self-Driving (FSD) approval in Europe. Therefore, we believe that once we can provide European customers with the same experience as in the United States, our sales in Europe will significantly increase. I think this is a very important point that needs to be communicated. We have been working with regulators in the Netherlands. I believe we are about to receive approval from the Netherlands. Then it will be submitted for EU approval. It is quite Kafkaesque. In fact, Kafka could not have imagined something like the EU existing. The challenges brought by bureaucracy even surpass Kafkaesque absurdity. But we will get approval. I believe this year, some people in Europe will receive a similar experience to that in the United States in most parts of Europe, hopefully achieving at least part of it this quarter. Additionally, we are also facing some regulatory challenges in China, which we hope to resolve soon, as we currently cannot provide supervised FSD in China either. But we hope to resolve this issue soon. This is also another major factor—it truly is the single biggest driver of demand. In the domestic U.S. market, as we gain more confidence in the safety of different geographic areas, we will relax the requirements for drivers to be fully focused and keep their eyes on the road. Because this is a common complaint. In fact, it does create a strange safety issue: people sometimes disengage autopilot and then do other things, like adjusting the radio or looking at their phones, driving with their knees, and then re-engaging autopilot, which is obviously less safe than keeping autopilot on. So, in any case, this experience will improve in the coming weeks. As we focus on Austin's no one privacy, the production version of the autonomous driving software is actually months behind the Robotaxi version that people experience in Austin. So now we are launching Robotaxi, which will add those elements back, so that people outside of Austin will see a leap in their autonomous driving experience. This really— as you can see, it is largely about the story of autonomous driving. It's like, we need—we need a physical product; without it, autonomous driving cannot be realized, but once you have the physical product, autonomous driving is the factor that amplifies the value to astronomical levels. We opened a Tesla diner, which is very popular. It has actually attracted global attention, which is unusual for a restaurant. Typically, restaurants do not make global headlines, but this is quite a special restaurant, and if you are nearby, it is worth a visit. Frankly, in what was originally a somewhat desolate open landscape, it stands like a shining beacon of hope So, it really is a very, very lovely experience. The team there is doing a great job. Continuing to talk about full self-driving, we are making significant improvements solely through software. We expect to increase the parameter count. At this point, I think we might be able to increase it by almost 10 times—yes, about a 10-fold increase in parameter count. This is actually a very tricky thing. Because when you increase the primary parameter count, you encounter memory bandwidth bottlenecks. But we currently believe we can increase the parameter count by 10 times. Relative to the level people are currently experiencing. So, not just 4 times, but actually a 10-fold increase in primary parameter count. Yes, therefore, there will still be a lot of improvements happening on existing hardware. The energy business is growing very well. Despite facing tariff headwinds, especially supply chain challenges, Megapack is rapidly expanding its capacity. We also have an upgraded version of Megapack that will make it even better. We achieved record strong deployments in the second quarter. So I think batteries will be a huge field. The scale of batteries, the demand scale for batteries, I think not many people really understand how enormous it is. You can understand it this way: the continuous power output of the U.S. grid is about 1 terawatt, but the average usage is less than 0.5 terawatt. If you add batteries, you can keep power plants running at full capacity 24/7, effectively doubling—if not more—the annual energy output of the U.S. just with batteries. But this is a big deal. It really is a big deal. Regarding Optimus, we are iterating on the design of Optimus, truly bringing it to an extraordinary design level. We are currently using Optimus version 2, which is considered version 2.5. I think Optimus version 3 will be a beautiful design, and as I have said many times before, I predict it will become the biggest product of all time. This is a very difficult problem to solve. You have to design every part of it from first principles of physics. There are no off-the-shelf parts that really apply, so you have to design your own motors, gearboxes, power electronics, control electronics, sensors, and mechanical parts. You also have to train Optimus to use its limbs and sensors, which requires neural networks, but we will apply the same technology used in cars (essentially a four-wheeled robot), while Optimus is a robot with arms and legs. The same principles that apply to optimizing AI inference in cars also apply to Optimus, as they are essentially different forms of robots. It is important to note that Tesla may be the best in the world in the field of real-world AI. A clear proof point is that if you compare Tesla with Waymo, Waymo's cars are equipped with God knows how many sensors, yet, isn't Google not good at AI? They have not been good at real-world AI so far. Tesla is actually much better than Google, far surpassing anyone else in real-world AI. Moreover, Tesla has the best inference efficiency. I believe a key metric for AI is how much intelligence can be generated per gigabyte. People talk about parameters, and so on, but I think people should stop talking about parameters and talk about gigabytes, because parameters can be 4-bit, 8-bit, or 16-bit, but the actual constraint in hardware is: how many GB of RAM you have and how many GB of data can be transferred from memory per second. Therefore, it is not a constraint on the number of parameters, but a constraint on the number of bytes. Tesla has the highest intelligence density among all AI so far. I have deep insights into this because xAI (Musk's other company)... xAI's Grok is currently the smartest AI overall, but Grok 4 is a terabyte-level behemoth. So it is important to understand that Tesla has the best intelligence density, and intelligence density will be a very big issue in the future. It already is. So for Optimus 3, this is the truly correct design; it currently has no major flaws, but we will redesign many things. Therefore, the prototype of Optimus 3 may be completed by the end of this year, and mass production will start next year. We will try to scale up the production of Optimus as quickly as possible, aiming to reach an annual output of 1 million units as soon as possible. Do you think we can achieve this in five years? That is my guess. This is a reasonable wish to reach an annual output of 1 million units within five years. This seems to be an achievable goal. Overall, so far, 2025 is a very exciting year, achieving many important milestones. We have made demonstrable progress in autonomous driving, while many "experts" (NASA, should be naysayers) have asserted that we could not achieve it. It is worth noting that we have done what we said we would do. This does not mean we are always on time, but we eventually did it. Our critics are left speechless. The Tesla team has made tremendous, tremendous, tremendous progress. I do believe Tesla continues to excel in vehicle autonomy and humanoid robot autonomy. It will become the most valuable company in the world. There is still a lot of execution work to be done from here to there. It will not happen by itself. In terms of robotics, we are executing very well. I believe Tesla has the opportunity to become the best company in the world. Clearly, I am extremely optimistic about the company's future. The best way to predict the future is to create it, and we are creating the future here at the Tesla team. Therefore, I want to thank all of our supporters, and I believe we are about to usher in an incredibly exciting future Travis Ackerman {BIO 24033559 } That's great. Thank you very much, Elon. (Background sound: Bedrock (possibly referring to another company or project mentioned by Musk) also has her fingerprints.) (Note: This sentence is unclear, possibly a slip of the tongue or background noise) Vaibhav Taneja {BIO 20985733 } Thank you, Travis. As Elon mentioned, the second quarter has been an interesting quarter in several ways. We have started to ramp up production of the new Model Y across all our factories. We launched the robotaxi service in Austin and achieved fully autonomous delivery from the factory to customers' homes. Reaching this milestone is groundbreaking. I mean, it took tremendous effort, and I sincerely thank every member of Tesla for making this possible. It was not easy, but we did it. It took time, but we have begun the next phase of the company. That "one big bill" contains many changes that will impact our business in the short term. The primary change is that the $7,500 electric vehicle tax credit provided by the Inflation Reduction Act (IRA) will be eliminated at the end of this quarter. Given this sudden change, our vehicle supply in the U.S. this quarter is limited, as ordering parts for vehicle production is already in the delivery cycle. We have rolled out all planned incentives, and as we begin sales, we will start to gradually recoup these incentives. If you are in the U.S. and want to buy a car, please place your order now, as we may not be able to guarantee timely delivery for orders placed in late August and beyond. We will also modify certain emission standards by reducing the penalty amount to zero. This, in turn, will affect our sales of new regulatory credits to other original equipment manufacturers (OEMs), leading to reduced emissions. Although we have never planned our business around such sales, it will still have a significant impact on our future revenue. In terms of our automotive product lineup, the entire product line has been updated globally, and we are seeing an increase in test drives. We plan to start production of lower-priced models in the first half of 2025 as scheduled. However, considering our focus on producing and delivering as many vehicles as possible in the U.S. before the EV tax credit expires, and the additional complexity of launching new products, the ramp-up for this model will occur in the next quarter, slower than initially expected. One aspect that the automotive market severely underestimates is that all vehicles in our product line are equipped with autonomous driving capabilities. This is undoubtedly our biggest differentiator from competitors. Our vehicles have already met safety standards, but with FSD equipped, they now and will continue to set new safety standards for regular traffic. We released the vehicle safety report earlier today, and you can see that cars with FSD enabled are ten times safer than those without FSD. In recent months, we have seen an increase in FSD adoption in North America, which is a very positive trend To give you an intuitive feeling, since we transitioned to FSD version 12, we have indeed seen an increase in adoption rates. We see... in terms of automotive revenue, despite a decrease in regulatory credit revenue, total automotive revenue increased by 19% quarter-over-quarter, even though total deliveries only increased by 14%. This is mainly due to the increase in average selling price (ASP) brought by the new Model Y. This also led to an improvement in profit margins quarter-over-quarter, while product mix improvement and higher fixed cost allocation also played a role, despite tariffs bringing increased costs. We have begun to see the impact of tariffs in the profit and loss statement (P&L). On a quarterly basis, tariff costs have increased by approximately $300 million, with about two-thirds affecting the automotive business and a smaller portion affecting the energy business. However, considering the lag in invoicing and sales, the full impact will be reflected in subsequent quarters, so costs will rise in the short term. While we are doing our best to manage these impacts, we are in an unpredictable environment regarding tariffs. The profit margins for our energy generation and storage business have improved quarter-over-quarter, while deployment volumes have decreased, mainly due to the completion of high-margin deployment projects. We achieved the highest gross profit for this business to date. Please note that overall deployment volumes will continue to fluctuate seasonally. I believe Elon has mentioned that industrial energy storage will play an important role in driving growth in artificial intelligence (AI) and data centers. As applications—especially energy-intensive AI applications—become more widespread, energy demand is growing rapidly. The fastest way to scale energy is through energy storage. This is something our customers have already begun to realize. Although this business is most affected by tariffs, we see customers willing to accept some of the tariff impacts. The "Big and Beautiful" bill even has certain adverse effects on the energy business, most notably on the residential energy storage business, as the bill prematurely terminated the "carbon zero catalyst" program at the end of this year. Therefore, the challenges facing the energy storage business come from both the bill and tariffs, and we are doing our utmost to try to cope. But this will—we will see a shift in demand and profitability. Our services and other business profit margins improved quarter-over-quarter, mainly due to higher profits from supercharging and improved profitability from insurance and service centers. Operating expenses (OpEx) also increased quarter-over-quarter as we continue to invest in AI projects, including additional costs related to employees, such as higher stock-based compensation and depreciation of AI computing. Our operating expenses, especially R&D-related expenditures, will continue to grow. We believe that even in the current environment, it is the right strategy to continue investing in these areas to position ourselves for long-term development. Other income increased quarter-over-quarter, mainly from mark-to-market adjustments of Bitcoin holdings, recording a gain of $284 million in the second quarter, compared to a loss of $125 million in the first quarter. Just a reminder, this will continue to bring volatility based on Bitcoin prices Although operating cash flow has increased quarter-over-quarter, our capital expenditures (CapEx) have also increased, resulting in $146 million of free cash flow. We continue to invest in all aspects of manufacturing, such as Cybercab (robotic taxi), Semi truck production lines, other manufacturing expenses, and the expansion of our AI engine. Our latest capital expenditure expectations for this year will exceed $9 billion. In summary, due to the negative impacts of legislation and tariffs, our business faces short-term challenges. However, our investments in AI, robotics, and our leading position in the energy sector lay a bright future for us. I want to thank the entire Tesla team, our customers, investors, and supporters for their continued trust in us. Q&A Session A - Travis Axelrod {BIO 24033559 } Q&A Session Great. Thank you very much, Vaibhav. Now let's start answering the pre-submitted questions. The first question is: Can you give us an update on how Robotaxi has performed so far? And what pace do you expect to expand in terms of vehicle numbers, geofence range, cities, and supervisors? A - Ashok Elluswamy {BIO 22648673 } Yes, Robotaxi has performed exceptionally well in Austin so far. Customers love the experience. It is very smooth, very safe, and the overall experience is great. We have completed the first round of expansion in Austin, and we will continue to expand there, potentially increasing the current operational area by more than ten times. As Elon mentioned, we are also testing in many other cities. The next area to expand will be the San Francisco Bay Area. We are working with the government to obtain approval for this area, and while waiting for regulatory approval, we will launch services with a person in the driver's seat to expand services during the waiting period. We are also testing in many other cities across the U.S., including Florida, Nevada, and others. A - Travis Axelrod {BIO 24033559 } Great. Thank you very much, Ashok. The next question is: What are the key technical and regulatory barriers that still exist for unsupervised FSD to be opened to individual users? Can you provide a timeline? A - Elon Musk {BIO 1954518 } We are definitely getting closer to the goal. I believe we will open unsupervised use to individual users in certain areas by the end of this year. We are being very cautious about this A - Vaibhav Taneja {BIO 20985733 } This is not something we want to rush into. We want to ensure everything is safe before broad deployment. A - Elon Musk {BIO 1954518 } Yes, we are just extremely paranoid (cautious). A - Vaibhav Taneja {BIO 20985733 } Yes. A - Elon Musk {BIO 1954518 } I believe by the end of this year, we will be able to provide it to end users in some cities in the U.S. A - Ashok Elluswamy {BIO 22648673 } Yes, it's also worth mentioning that the hardware in the Austin Robotaxi vehicles is the same as in some customer vehicles. We have indeed achieved autonomous delivery from the factory to the customer, which is a first (squadron, should be seminal first). A - Elon Musk {BIO 1954518 } Yes. A - Ashok Elluswamy {BIO 22648673 } Every Tesla manufactured in the U.S. and Europe will autonomously drive itself from the end of the production line to the loading docks. So it's just a matter of a software date (day-to-day, should be day and date software version synchronization). A - Elon Musk {BIO 1954518 } Yes, I think by the end of this year, by default, we will achieve car deliveries in the Austin area and the Bay Area. A good car will drive itself to your location unless you say you don't want it that way. That would be super cool. A - Travis Ackerrod {BIO 24033559 } Yes. Great. Thank you all. The next question is: What specific factory tasks is Optimus currently performing? What is the timeline for scaling its production to a level capable of external sales? How does Tesla view Optimus's contribution to revenue in the next two to three years? A - Elon Musk {BIO 1954518 } Okay. As I mentioned earlier, the design of Optimus 3, I think, is ultimately the right design. There will be further optimizations, but the design of Optimus 3 does not require fundamental changes. It has all the degrees of freedom you really want So it— we will have its prototype in, I don’t know, three months. Then there will be sub-scale production. We will definitely start mass production early next year. The ramp-up speed is always difficult to predict, especially when something involves the entire… when everything is new, because the production speed will depend on the slowest and least competent links in the entire supply chain and internal processes. Therefore, the more new things included in the product, the slower the ramp-up may be, as unexpected supply chain disruptions or internal errors may occur. Predicting the end of the S-curve is much easier than predicting the early stages of the ramp-up. The production volume in the early stages of the ramp-up is usually not really important for revenue purposes. In the early stages, you are usually… you are always in a negative contribution margin state and have to solve many debugging issues. So that’s why, for example, I feel quite confident, or at least moderately confident, about predicting things five years out, but it’s very difficult to predict what will happen in a year or two. That’s why I think that within five years, if we haven’t reached a level of producing about 100,000 Optimus robots per month after 60 months (i.e., five years), I would be shocked. A - Unknown Speaker Okay. Thank you very much. The next question is: Can you provide the latest information on the development and mass production timeline of Tesla's more affordable models? How will these models balance cost reductions and profitability? In the current economic environment, what impact do you expect them to have on demand? A - Travis Axelrod {BIO 24033559 } Well, I think Vaibhav has already answered this question well in his opening remarks. As we said, we started production in June and have been continuously improving quality production and related matters throughout the quarter. Considering that we started production in North America, our goal is to maximize the production of existing models by the end of the third quarter when the IRA subsidies end. We will continue to push hard on existing model production to avoid complexity (referring to the complexity introduced by new models). Unfortunately, when the subsidies disappear, we will be ready in the fourth quarter to offer more affordable models for everyone. The goal of these products is not to negatively impact revenue or gross margins, but to create a car that everyone loves and is more affordable. A - Unknown Speaker Great, thank you, Lars (should be Travis). The next question is: Can you talk about the benefits of Tesla's investment in xAI? A - Elon Musk {BIO 1954518 } This is not the occasion to discuss this topic. What I mean is, if there are things that need to be discussed, we will discuss them separately. A - Lars Moravy {BIO 22525342 } I think it's clear that we are a publicly traded company, and shareholders can propose any shareholder proposals they wish. I have recently encouraged this, and then let the shareholders vote, and we will act according to their wishes. A - Unknown Speaker Great, thank you very much. The next question is: Can you tell us more about what is happening in the Tesla design studio? (Background sound: Do you want me to answer this? We usually say that what happens in the studio stays in the studio.) (Background sound: These conference calls are not the place to disclose new product content, but we are working hard to ensure that Tesla and the product line have an exciting future.) A - Elon Musk {BIO 1954518 } Yes. There are many exciting things happening in the studio. It is not static. What is really going to happen in the coming years is a fundamental shift for the company from the "pre-autonomous driving era" to the "post-autonomous driving era." I am developing a new overall plan to outline this. As a team, when you transition from the pre-autonomous driving era to the post-autonomous driving era, you will encounter some teething pains, but I believe Tesla's future vision is extremely exciting and will profoundly change the world in a positive way. It's not like, I think if we... let's put it this way, if we effectively execute that plan - you actually have to execute - then Tesla will far exceed being the most valuable company in the world. A - Unknown Speaker Great, thank you. The next question actually duplicates the one about unsupervised customer vehicles. We'll skip that. The next question is: For hardware 3 (HW3) users, is there any news about obtaining hardware 4 (HW4) or future hardware 5 (HW5) upgrades or retrofits? A - Lars Moravy {BIO 22525342 } What I mean is, what we want to do first is to complete unsupervised driving on hardware 4. Once that is done, we will look back at what needs to be done for hardware 3 vehicles. As I said, the focus is first on launching unsupervised driving, and then we will look back to see what more needs to be done. A - Travis Ackerrod {BIO 24033559 } Great. The next question is: Can you update us on the status of Dojo? Will xAI become a customer of Dojo? A - Elon Musk {BIO 1954518 } Dojo 2, we expect that sometime next year Dojo 2 will reach scalable operation, with a scale roughly equivalent to 100,000 H100 (NVIDIA AI chips). Then there is AI5, which is also very impressive. I rarely use the word "impressive." The AI5 chip is expected to enter mass production around the end of next year, but its potential is enormous. I think, considering Dojo 3 and the AI6 inference chip, intuitively we want to try to find a fusion point there, basically the same chip but used in different places, such as using two in cars or Optimus, or using more on the board, like for training or inference. If you want high-bandwidth communication between chips during training and inference, this seems intuitively to be a wise approach. A - Travis Ackerman {BIO 24033559 } Great. The next few questions have actually already been covered, so we will end with this question: What impact will the cancellation of the tax credit for solar projects under the BBV (should be BIB, referring to the "Big and Beautiful" Act) have on the sales pipeline for Megapack? A - Ashok Eluswamy {BIO 22648673 } (To Mark) Yes, Mark, our sales pipeline is very diversified in terms of customers and market segments, so we do not heavily rely on Megapack projects paired with solar energy. As we discussed in the opening remarks, we see energy storage rapidly gaining recognition for its ability to enhance grid efficiency and its rapid deployment to assist the grid. Furthermore, despite the recent legislation being unfavorable to solar energy, we believe solar projects will still be built because energy is essential, and projects are well-developed and ready for execution. Considering the delivery cycles and prices of gas turbines, there are indeed no alternatives in the short term. We also see continued growth in the data center sector and independent energy storage projects providing capacity for the grid in multiple markets across the U.S. Overall, with the increase in deployment, we predict a very strong second half of the year. Finally, we continue to invest heavily in domestic manufacturing in the U.S. to mitigate the impact of policies and tariffs, and we expect our first LFP battery production facility to be operational by the end of the year, with our third megafactory launching near Houston in 2026. A - Travis Ackerman {BIO 24033559 } Great. Thank you, Mike (should be Ashok). We will now turn to analyst questions The first question comes from Emmanuel Rosner of Wolfe Research. Emmanuel, please unmute yourself when you're ready. Analyst Emmanuel Rosner asks a question (specific question text not provided) A - Travis Akserrod {BIO 24033559 } Yes. (The analyst's question may be about Robotaxi operating mileage and safety record) A - Unknown speaker Yes. We have exceeded 7,000 miles of operating mileage in the Austin area. Just because the service has just started, we currently have only a few vehicles, but we are working to expand the service area and the number of vehicles in Austin and elsewhere. So far, there have been no notable safety-critical incidents. Sometimes we have our own limitations, for example, we limit speeds on roads to 40 miles per hour, and if the vehicle wants to go onto a faster road, we will stop, but these are for convenience rather than safety-critical reasons. So far, the response to the service has been very positive, and we will continue to expand it. (The analyst may follow up on the unit economics of Robotaxi) A - Elon Musk {BIO 1954518 } Okay. Well, Cybercab (vehicles optimized for autonomous driving), I think its cost potential in the long run could be below $0.30 per mile, maybe $0.25. Maybe $0.25. It really is—if you design a cost-optimized robot taxi from scratch, like Cybercab, for example, we are not trying to make it corner as well as a Model 3 or Model S, or even a Model X. All of our human-driven cars are super fun, they have amazing acceleration and cornering capabilities. But we are confident that very few people want to experience high-speed cornering in a Cybercab. Therefore, we limit the top-end speed, which means we can use more efficient tires. We don’t need that strong of acceleration capability. We don’t need such a large braking system. We want to have braking distance, but we don’t expect it to be used aggressively very often. It aims for a smooth ride. Essentially, if you design for a smooth ride, then you have a more optimized design point. That’s why I think we might achieve that goal (low cost), especially if Optimus can take care of cleaning the vehicles, performing maintenance, and achieving automatic charging. So I think the actual cost per mile for Cybercab will be very low. The cost per mile for our existing fleet will be higher, but still very competitive. Yes, maybe over $0.50, I’m just guessing. Yes, so this—Tesla's Robotaxi fleet operation scale will grow from tiny to huge in a relatively short time My guess is that by the end of next year, it will have a substantial impact on our financial situation. Great. Thank you very much. The next question comes from Adam at Morgan Stanley. Adam, please unmute yourself when you're ready. (Analyst Adam Wood asks a question, possibly related to corporate governance and Musk's control over AI) A - Elon Musk {BIO 1954518 } Yes. This is a major concern for me, as I mentioned earlier, I hope to address this issue at the upcoming shareholder meeting. But yes, this is a big deal, like, I don't want to find out that my control is so limited that after having this army of humanoid robots, I could easily be ousted by activist shareholders. I think, as I said before, I believe my control over Tesla should be sufficient to ensure it moves in a good direction, but not so much that I can't be removed if I go crazy. (Analyst may follow up on whether Tesla plans to hold an AI Day to showcase progress) A - Elon Musk {BIO 1954518 } Yes, this is a bit tricky because when we hold AI Day, we find that some competitors study our slides frame by frame, everything we say, and then mimic us. So I would say, what’s the trade-off? It does help with recruitment, but competitors will study and mimic us very closely? That said, we might want to— I mean, we could consider using the shareholder meeting as a way... we might delve into some discussions about Optimus, AI, and our chip technology at the annual shareholder meeting. Yes, we are also severely underestimated in AI chip design and AI software. So, even though it has been out for a few years now, there is still no AI chip we would prefer to put in our cars (referring to Tesla's self-developed chips being superior to external options). We believe AI chips will be a profound game changer. In fact, they are so powerful that we have to nerf them to some extent to sell them in markets outside the U.S. Because they far exceed export restrictions. So unless export restrictions change, we will actually have to weaken our AI, which is a bit strange. Hopefully, we can continuously raise the threshold for export restrictions, otherwise it will be a bit ridiculous. We will showcase some Optimus robots at the shareholder meeting. The Optimus lab is cool and worth seeing. It basically looks like a scene from Westworld. The robots are at different stages of readiness. Some are at different preparation stages. I don't know, like a combination of the Tatooine junkyard and Westworld. Very cool. Optimus is walking around the office in Palo Alto So it's 24 hours non-stop, it just walks around normally. I think for Optimus, it's like the Tesla restaurant, they will pop on suddenly. Yes, so we will transition from a world where robots are rare to a world where robots are so common that you won't even look up. A - Travis Axelrod {BIO 24033559 } Great. Thank you very much. The next question comes from Edison at Deutsche Bank. Edison, please unmute yourself. Edison, please continue to unmute. Well, while Edison addresses this issue, we will move to the next question from Dan Levy at Barclays. Dan, please continue to unmute. (Analyst Dan Levy may ask about the timeline for adding private cars to the Robotaxi fleet) A - Elon Musk {BIO 1954518 } We haven't seriously considered this yet. I really need to ensure that it works properly when the vehicles are fully under our control. Yes, it's a bit like taking it step by step. We don't want to rush it. As I said, we are very paranoid (cautious) about safety. So it's like—but I guess that will be next year—I can confidently say it will be next year, I'm not sure when next year, but confidently say that next year people will be able to add or remove their cars from the Tesla fleet. A - Lars Moravy {BIO 22525342 } I mean, keep in mind that we will set some standards. Because even if you put the car on Uber or Lyft, I still need to go through a whole set of checks to ensure everything is okay. A - Elon Musk {BIO 1954518 } It's like Airbnb and the like. A - Unknown Speaker Yes. Yes, so we will do something similar. A - Elon Musk {BIO 1954518 } Some sort of vetting process. A - Unknown Speaker Yes. As Elon said, we want to better understand safety. I mean, in terms of assets, things like tread on the tire can affect safety. That's why we want to do proper validation before allowing other cars to join. A - Travis Axelrod {BIO 24033559 } Dan, do you have a follow-up question? (Analysts may ask about the funding sources for fleet expansion) A - Elon Musk {BIO 1954518 } Well, once any product has a clear cash flow, you can finance it with debt. A - Vaibhav Taneja {BIO 20985733 } I mean, during the transition period, we will use our balance sheet. But once we reach a certain scale in current revenue, as Elon said, we can do simpler deals to try to obtain financing. A - Travis Akserod {BIO 24033559 } Great. Thank you very much. We will now turn to Mark from Goldman Sachs. Mark, please unmute yourself. (Analyst Mark Delaney may ask about FSD adoption strategies) A - Vaibhav Taneja {BIO 20985733 } We certainly have seen a significant increase in FSD adoption since we launched FSD version 12 in North America, as I mentioned in my opening remarks. Additionally, we did something last year where we lowered the price and made subscriptions more affordable. Since then, we have seen a 45% increase in adoption. This is an encouraging trend. But honestly, we are just getting started on working to explain the benefits of FSD. As I mentioned earlier, we released the vehicle safety report. Even if you don't believe anything else, the fact that FSD-enabled cars are 10 times safer should be an incentive. Another point that people don't realize is that even at $99 a month, it equates to about $3.33 a day for having a private driver. This is undoubtedly the biggest game changer, and I know we have been talking about it partly because we are living in it (referring to the focus on FSD), but I feel that— A - Elon Musk {BIO 1954518 } Most people still don't know. The vast majority of people are unaware of its existence, and even half of the Tesla owners who could use it have never tried it once. They actually don't realize this. Obviously, this is where we want to educate them. So we have to—when they come in for service, we will reach out to them and send videos on how to make it work. It's very shocking. They think the car can't do that. So you have to show them and get them used to turning it on and off. It's very simple, but it's like saying you have a cat that can sing and dance, but it just looks like an ordinary cat. Until you see the cat singing, dancing, and talking, you will think it's just a cat. That's what Tesla's FSD is. Our cars are smart A - Las Moravi {BIO 22525342 } So what we are going to do, as Elon said, we have been giving people free trials of FSD, but we will start providing more prompts, like, "For this specific trip, try FSD." Because seeing is believing. Just like Elon said, think of it as a cat. It looks like an ordinary cat, but this cat can sing and dance. The same thing. That is intelligence. A - Travis Axelrod {BIO 24033559 } Great. Yes. That 25% comment refers to the adoption rate increasing by 25% since we released V12 and V13 in North America. Great. Thank you. Mark, do you have any follow-up questions? (The analyst may ask about short-term financial outlook) A - Elon Musk {BIO 1954518 } Well, we are in a strange transition period, and in the U.S. we will lose many incentives. In fact, many other places in the world still have incentives, but in the U.S. we will lose them. And we are still in a relatively early stage in terms of autonomous driving. On the other hand, from a regulatory perspective, autonomous driving is the most advanced and available in the U.S. So, I mean, does this mean we might have a few tough quarters? Yes, we might have a few tough quarters. I'm not saying it will definitely happen, but it could happen. The fourth quarter, the first quarter, maybe the second quarter. But once you achieve large-scale autonomous driving in the second half of next year, of course by the end of next year, if Tesla's economics are not very attractive, I would be surprised. A - Travis Axelrod {BIO 24033559 } Great. The next question will come from Will at Truist. Will, please unmute yourself when you're ready. (Analyst William Stein may ask about the appeal of economic models or AI strategy) A - Elon Musk {BIO 1954518 } Oh, we won't discuss the appearance because then you would go observe... it's like the Model Y. They just pulled the cat (cat, should be car) out of the bag (referring to unveiling). That cat that can sing, dance, and talk. But it really can talk, sing, and dance. That's the cool part. Yes, I mean, fundamentally, the biggest barrier is still that people don't have enough money in their bank accounts—some people don't—while the desire to buy a car is very high. It's just that people don't have enough money in their bank accounts to buy it. Literally. It's not a lack of desire, but a lack of ability. So, we can make cars as affordable as possible I believe that when people can release their cars into a fleet and make money from them, it will be a very big deal. As I said, I’m very bullish—I confidently say this will happen in the U.S. next year, at least in the U.S. In the U.S., we are legally allowed (with appropriate disclaimers). This will greatly increase affordability. It’s like having an Airbnb where you rent out your house when you’re not there, or renting out a guest room or something like that, which significantly increases the affordability of your home. (Analysts may ask about the differences between Tesla AI and xAI) A - Elon Musk {BIO 1954518 } Well, they are doing different things here, so... you know, xAI is working on models at the scale of terabytes, multi-terabyte scale. Tesla's models are 100 times smaller. So, one is real-world AI, and the other is... I guess, artificial superintelligence (ASI) type of stuff. I mean, the real genesis of xAI is that some people simply didn’t want to join Tesla as AI engineers because they wanted to work on open ASI... AI, and they wouldn’t join Tesla. I thought, well, maybe they would join a new company. I think Tesla's vision is extremely important, but not everyone agrees with me. So rather than letting them join OpenAI or Google or whatever, why not... create a company like that, right? That’s xAI. So that’s it. And then let them choose what they want to do? Work on superintelligence in data centers, or work on real-world AI? Both are compelling questions, but some people want to do one, and some people want to do the other. A - Travis Ackerle-Rod {BIO 24033559 } Great, unfortunately, our time today is up. Thank you very much for your questions today, and we’ll see you next quarter. A - Unknown Speaker (Commentator after the meeting) Okay, everyone, give me a second. Pause the music. Tesla has wrapped up, and we just need to go over some key points. Please let me know if the sound is normal. Hope it looks like there shouldn’t be any issues. Make sure everything is set up. Yes. It looks like the stock might have slightly risen or fallen before the conference call. Now it seems to be down about 4.5%. Not sure if there’s any specific reason; it seems to be a continuous decline during the call. I don’t know if a specific comment caused it. If anyone thinks it did, please let me know. Well, back to the notes. But I don’t think this was a bad call. I think there was a lot of exciting content in it. Anytime management acknowledges that there may be some tough quarters ahead, the market tends to get a bit nervous. Typically, company management—this refers to most companies—will paint things as optimistically as possible, not to say that Tesla hasn’t tried to do that, or even inadvertently did so But I also think Tesla may, some people might find this a bit ironic, but I do believe Tesla is more directly affected by some negative factors. Maybe not in all cases, but in this case, Elon specifically mentioned that, oh, there might be some tough quarters in the fourth quarter and the first quarter. I mean, we discussed this earlier today, right? This shouldn't surprise anyone. It's very predictable, especially when you understand that the electric vehicle tax credits are about to end. So, I don't mind admitting that. I think it's completely—again, I emphasize, it's expected. But as we discussed earlier today, the most important thing is to ensure the business is at a level that can continue to support the investments Tesla is making and the future they want. Shareholders hold stocks for this future potential, for those who agree with this direction. I think there were many exciting aspects in this conference call, and those less exciting parts, I think, are well-known or fairly understood, just some challenges Tesla is facing right now between being in a real growth phase and a transformation phase. As Elon mentioned, but in terms of what Tesla is becoming, I think there are many exciting things. I think Elon said at the beginning of the call that we need to go back (referring to reviewing the opening remarks). ...Just going over the notes again. Confirming the audio is fine. It seems so. So yes, expect to greatly expand the service area. I think it was possibly mentioned accidentally, but the service area will far exceed the level of competitors. We briefly discussed Tesla's expansion with Waymo earlier today. I mentioned that Waymo's expansion might not be as significant because they also collaborate with Uber, so they don't need to ensure the entire area is well covered since they have Uber drivers as backups. Tesla is different; that is the Robotaxi service area. If you're in one part, calling for a ride, it has to come pick you up. It doesn't have the option to send an Uber driver. So Tesla expanding the service area is extremely meaningful. Maybe not in this initial version, but they mentioned here that it could reach about ten times the current service area, which I think is more like their ultimate goal. But just the fact that they expect this to happen very quickly is important. I think Elon mentioned here that it might be next week, or certainly in the coming weeks, to see if I can find that record, that they will be expanding that service area in Austin. So I think that's encouraging. I think the discussion about focusing on the FSD in the Austin service area and the Robotaxi launch, Tesla seems very satisfied with this. And I think, as they mentioned before, the factors driving that performance haven't really entered the production software that people can use yet. The current version is 13.12.9. But once these things actually enter these vehicles, there should be another leap in improvement, just like we've seen with all the major releases from Tesla recently. I think the change from version 11 to version 12 was a huge shift I believe there is a huge change from version 12 to version 13. We have been using version 13 for a long time now, and there is ample reason to believe that, even setting aside Robotaxi and the achievements Tesla has made there, considering the time and resources Tesla has invested, there is good reason to believe that with the release of the next version, FSD will see another significant improvement. It has been so outstanding; I mean, Elon said a few months ago that people might laugh at him for this, but now it’s hard to distinguish which version is better because interventions have become very rare. That is the real experience I have using my own vehicle. So when we start talking about even taking a step further from there, it will become almost difficult for customers to notice because it has already been so excellent. But Tesla has a good understanding at the top level of what it will be like when it is deployed on vehicles that drive a billion miles every quarter. Then you will get very good data to showcase this improvement, and perhaps most importantly, reach a point where: okay, these interventions become so rare that we are satisfied with the initial launch of Robotaxi and the subsequent expansion in other regions, ultimately basically rolling it out to the entire fleet wherever regulators allow. We are on that path, right? There are many positive signs indicating that this path will continue. So that is basically what I am most focused on in this conference call because that is Tesla's future. I think a lot of positive factors regarding that future were discussed in the call. Particularly here, this point I noted, the number of parameters increasing tenfold, I think is significant. In version 13 software, Tesla will add (four times the number of parameters) in the upcoming version. They will have four times the number of parameters. So it sounds like they are not only confident about this but also convinced that they can take it further, achieving ten times the number of parameters on the current hardware, and this is just the current hardware, soon. We will move into AI 5 hardware, which has... since I drove version 13, it feels like, yes, this is amazing. I see no reason why another hardware iteration cannot easily accomplish this, given the pace of progress we have seen from Tesla and the level of advancement released by hardware 4. I believe we are still in the early stages of hardware 4, definitely not at the endpoint of hardware 4's progress. So yes, I think what Tesla will be able to do with autonomous driving in the next two years is very exciting. I think the achievements Tesla has made in autonomous driving over the past year are very exciting. I think that is why the stock remains at around a $1 trillion market cap, because we have seen progress, even though it is slow. It has been happening for many years, but it is happening and will continue to happen, and in hindsight, it may seem that it is happening quite quickly here. That is my expectation for the next 12 to 18 months. So from this perspective, to emphasize again, I think this is the viewpoint of Tesla shareholders, perhaps not the viewpoint of traders, which can largely influence daily fluctuations. But for long-term Tesla investors, there is a lot to look forward to, and many real, tangible advancements are being made in those truly important aspects of the business's future However, yes, when the electric vehicle tax credits end, there may be a tough fourth quarter, and I hope the launch of this more affordable model can ease that a bit. It sounds like Elon said it looks like the Model Y, which could be a good thing, just giving a rough hint so people can adjust their expectations accordingly. I don't know, maybe this is part of the market reaction, but I'm a bit surprised by it, but I hope so, and I think the timing of the model's launch will align well with the end of the electric vehicle tax credits. I hope this can fill that gap, but it will put pressure on the average selling price (ASP), especially if it is indeed very similar to the Model Y. This could put quite a bit of pressure on actual Model Y sales. So, let's see. FSD as a huge selling point, these comments, I'm not sure I completely agree or understand. Yes, I do think FSD is a huge selling point. For me, it would be very difficult to buy a non-Tesla car because I would really miss FSD. So for me, yes, it is a huge selling point, but at the same time, the adoption rate is low, so people are clearly buying Teslas for many other reasons. I think there are also many good reasons for that. So they acknowledge it as a huge selling point. They acknowledge that many people do not fully understand it. It's like, guys, let's just combine these things. And then maybe do some advertising. I know Tesla's attitude towards advertising is like... (hydrophobic should be averse) to put it another way, but this—seems like a pretty obvious solution. At least try it, at least try something like putting an ad for FSD during the Super Bowl. The risk is very low. About $5 million or something like that. I don't know. For the scale of Tesla's business, trying more significant advertising is inconsequential. I hope they can get over that hurdle. I mean, Starlink does advertising too. So maybe not the most traditional way. I know Tesla has done a little, but I would like to see them try some more proactive ways and see what happens. I understand this hesitation. It's a bit like stepping onto a slippery slope, but this seems like a perfect situation: hey, low awareness, we have the resources to raise awareness. Maybe we can combine these two things. Well. Anyway, I've talked about this many times. That's it. Okay. I feel like I've gone a bit off-topic in the notes, but let's see. Yes. So the service area far exceeds the level of competitors, and I hope in about a week, I think Ashok said a few weeks. So, at some point this quarter, we should see another expansion from Tesla. I hope we can continue—I hope someone can specifically ask about the safety rider issue. The problem is when people ask five questions at once, Tesla doesn't answer them all; they pick one to answer, either one they want to answer, one they don't mind answering the most, or just the last one in the sequence, that one will be answered So people really need to learn to ask just one very specific question that they want to know the answer to. For example, "Can you tell us when the safety driver will be removed?" Just ask that. Then ask your follow-up questions. I hope that's the case, but I don't think we have that answer yet. Tesla may not have that answer yet either. So it's a bit half-joking, but I believe they want to remove it as soon as possible. But it's more like—does that actual decision really unlock something for Tesla? Not really. It's more like an external signal that indicates Tesla has reached certain milestones. So it will happen when it happens. I don't know if Tesla knows exactly when that will be at this moment. It's great to see they are discussing it. It sounds like discussions are happening with the Bay Area, Nevada, Florida, Arizona, and possibly other places. I think Elon said they expect to launch Robotaxi by the end of the year, and we only have five or six months left now. To launch in possibly five or six different jurisdictions, time is tight. So, I think that's also very exciting. Back to the positive aspects of this earnings call. I absolutely count that in. I think this is actually more optimistic than what we heard earlier this year, which is quite a significant change. Tesla initially indicated mainly Austin and the Bay Area. Then possibly other places, and now it sounds like we have more specific information about those places, and it seems more optimistic about the timeline, both of which are very encouraging because you are closer to that actual timeline becoming a reality. So yes, I think that's exciting. By the end of the year, it could cover half of the U.S. population. It's not like that; I don't think that's the expectation. I don't think anyone really, I mean, the most optimistic people might expect 100% global coverage by the end of the year, but in terms of the range of participants, I don't think there are many people—and I—Tesla has set ambitious goals, so it might happen later than they aim for, but even so, I think people don't fully understand or realize that scale. When Tesla really aims to be able to do this, it's easy to dismiss the current state of Austin and say, yes, there is a safety driver. The public cannot use it. Relative to half of the world's population, the service area is very small. It's easy to downplay it now, but the ability to downplay it will disappear very quickly. If Tesla's expansion is close to what they expect and aim for, the ability to criticize it will disappear very, very quickly. So—we are on the edge of whether that will happen, on the edge of whether the skeptics are right. I think Elon lives a bit in the future. He said the skeptics won't be able to hold on. I don't think they necessarily feel that way right now, but if Tesla does what they are saying here, then it will be almost indisputable that soon we will be in that outcome. I think that's where Elon is, and that goes back to his comments. Tesla is always three to six months ahead, just because they can see what's actually happening, they can see the data, they can see simulations of future software versions, they have that insight. And we are just looking at, oh, it's Austin. What we are seeing might be software from three to six months ago That might be what is currently running in production, even in Austin as well. So, just trying to remind people of that. Again, I emphasize that this is why this conference call feels quite positive to me, because I think Tesla is very optimistic about those developments. Yes. The Model Y situation, FSD in Europe. I really hope they can resolve this issue. It would be great to reduce driver monitoring. It—I feel like they improved it, and then, I don't know if they ran into trouble or what, but it seems to have become strict again. Browsing notes, Tesla restaurant. Okay. The number of parameters increased tenfold. We discussed that. Energy, a lot of bullish comments. It's great to hear them contextualize some of that with the "big and beautiful" bill and tariffs, etc. They expect there to be an impact, but it sounds like it might be a bit smaller than people are worried about. So I think that's good. A lot of discussion about Optimus. So looking forward to seeing that version 3. Obviously, we've seen the hands of version 3, or maybe it's like version 2.5. We'll see if the hands of version 3 are another iteration, but definitely looking forward to that. I like that you want to put this in the context of AI differentiation. AI now, I don't know if it's like a catch-all term, but it's a broad industry. It has many elements. I think it's hard to argue against your point about Tesla leading in intelligence density, which is crucial in real-world AI; you can't walk around with a massive xAI training data center. So that's a very different issue. Obviously, they work together in many ways, but there are definitely differences here. The Optimus production thing. The second quarter was an interesting quarter; these are the CFO's comments, and it sounds like FSD adoption has increased by 25% since version 12. I don't know, that might be a bit lower than I expected, just because it has gotten so much better, but I do think they correctly pointed out that people are not aware of how much better it has become, whether within the Tesla customer base or externally. I think the awareness is quite low, so I think that's why the adoption rate, while seeing an increase is definitely good, and I think they are talking about percentage basis on percentage (percents on percents), so if it goes from 4% to 5%, that's like a 25% increase, I don't know. That's my understanding of how they talked about it. I might need to go back and listen, but I don't think it's from 10% to 35%, that would also be a 25% increase, but just the way you talk about it, frame it. So my understanding is the former, a smaller increase. Automotive revenue, we discussed a bit earlier today, it sounds like it's mainly driven by ASP—or sorry, ASP is mainly driven by the product mix, especially the new Model Y. Tariff costs, I didn't catch completely; sometimes it's hard to accurately hear what they're saying, I tried, but I didn't fully grasp the tariff cost thing, mentioned about $300 million, it sounds like it will affect future quarters, and may also impact this quarter, but I might need to go back and listen We may see tariff information in the 10-Q form, but the 10-Q usually also provides some background information on how the numbers are affected, at least. Elon’s good points about energy, I think people here might understand, it’s just that energy demand is almost impossible to meet, and how batteries can help solve this problem very quickly. I think the "quick" part is important; if you’ve dealt with xAI, you’ll understand the magnitude of the challenges with power and energy, and that batteries can help address these challenges and can be quickly put into production or become part of a project. R&D continues to grow, with capital expenditures of $9 billion this year, all of which I support. I think Tesla has a lot of exciting things worth investing in, and I believe they are investing. Yes, so here’s that 10x comment; we might expand the service area in Austin by more than 10 times, and next will be the Bay Area. It sounds like they will first launch the driverless Robotaxi system in the Bay Area. I think Waymo is doing the same thing, and people will mockingly say, oh, Tesla is just completely copying Waymo. First of all, no, even if there are some similarities, they are not exactly the same. Secondly, even if they followed the exact same process, they are doing it with very different hardware. So even that criticism is wrong. It’s like, oh, you’re doing the exact same thing at a cost that’s five times cheaper. That’s lame. But obviously, that’s extremely impressive. So that criticism always makes me chuckle. The unsupervised timeline for FSD. That is not at all the focus of their work right now. I mean, it is, but clearly the focus has been on the Robotaxi in Austin, and when it runs to a level that Tesla is satisfied with for expansion, they can start to shift the focus to customer vehicles and other aspects. More on Optimus. I think it’s something Elon has talked about before, nothing too crazy new. The shareholder meeting stuff. We will see Optimus at the shareholder meeting. Tesla will talk about progress there. It sounds like Elon is confident there will be some sort of compensation vote at the meeting. Elon may have the overall plan he mentioned, which will obviously include Optimus. I expect that either before the shareholder meeting or at the meeting itself. So I think that will be the fifth overall plan, maybe the fourth. Looking forward to that. The fundamental transformation of the company. Dojo 2 and Dojo 3. Takie (possibly referring to AI chip names). Sorry, I just got an alert. AI 6 is the next chip in the ongoing iteration after AI 5. Elon sees it merging with Dojo 2. I know Tesla has been using hardware 4 in simulation or inference. Maybe in training. I might need to go back and check, but obviously, if they are going to use the AI 4 chip or hardware 4 chip in vehicles at the server level or remote level, they realize the value of these chips goes beyond just being in vehicles. So this is something Elon has talked about a bit before. For example, can you leverage the computing power that will be idle in these vehicles? The more points of integration between these designs, the better you can utilize that. So I don’t know, this is just—this is a fascinating little intersection that Elon is particularly suited to navigate, given his experience at Tesla (which has covered a wide range of AI fields), and obviously, his history at OpenAI Now, of course, there's xAI. Elon is in a very unique position to make decisions on how to best adjust these levers to optimize these areas. So I'm looking forward to hearing more about that. I think Adam Jonas asked about AI Day, and I mean, that's a good question in that context; we may not be at the point where they can talk about those things yet, but it's something we would love to hear more from Tesla about because the developments in these areas are a very exciting time. Well, regarding energy, we know about LFP, we know about the third Gigafactory. $0.3 per mile, guessing Cybercab. Tesla, I think, has talked about this before. Other vehicles might be $0.5 per mile. So, compensation matters, Optimus matters. I think that's mostly it. Yes, I think we've covered most of it. So yes, I want to look at today from a higher level. I think it's, excitingly unexciting. It's like, I don't think there are a lot of negative surprises here. And the positives we've heard reinforce the thesis. I mean, at least the reason I hold Tesla is that future of autonomous robotics, first through electric vehicles, and then obviously recently humanoid robots. Within that framework or thesis, I think Tesla is executing, and I think they’ve given a lot of optimistic reasons. I think their optimism is not just for show or for the sake of being optimistic. I think they are genuinely optimistic, and they have ample reasons to be optimistic. So that's basically what I was looking for before I came here—Is Robotaxi on track? Will it continue to expand? Yes. Yes. It sounds like it will happen in the coming weeks, and the service area could be much larger. I don't think Waymo will surpass that in the week after. The scale competition we are having between Waymo and Tesla, clearly Tesla started later in terms of scale, but the real question is how fast the progress is from there? Elon has said hyper exponential. Obviously, Elon has been wrong in the past about timing predictions on these topics, but if that happens, I think it doesn't take much for people to quickly realize it's a bit like a game-over situation. So, we might not be far from that realization. Maybe we are far. It really depends on how things develop, but that's what I'm focused on right now. I think everything they said either meets my expectations or exceeds my expectations in terms of how things are progressing. So, excitingly unexciting. We'll see. Let's see how things develop. If we don't get another expansion in two weeks, or in four weeks, then it starts to feel disappointing. But if those things happen, then I think the reasons for excitement still exist, and we'll see Then on a more negative note, I don't know, those things are either expected, not to say they are unimportant, but I certainly don't think there are any unexpected negative factors here. We know about the tariffs. We know about that "big and beautiful" bill. We know these things. Obviously, there will be impacts. Tesla will have to deal with these impacts, which may put pressure on the business at some point. But the end of the electric vehicle tax credit has almost nothing to do with Tesla becoming a self-driving taxi service provider. It's like those are not even related at all. So it may not be unimportant, but it's not the most important thing. Alright. Alright, a quick look here, are there any questions? I see there are a few super chats, so I want to thank you, Greg. That’s a huge tip. I really appreciate it. It means a lot that you find this helpful, that you think my comments are worth listening to, even if not the smoothest presentation. But I really appreciate it. It truly means a lot. ESR, thank you. Here’s the question. Let’s take a look. So, having a CEO who knowingly helped elect people involved with Epstein, is that a help or a harm to the stock? I’m not sure, curious about your analysis. So I think that’s half-joking. You probably didn’t expect me to answer that. But hey, I’ll talk about it. I mean, there was indeed a report today saying that the DOJ notified Trump in May that he was part of the Epstein documents. I think that was around the time Elon said that. So the framework of your question actually doesn’t align with that report. Now, depending on your political perspective, people will believe various different narratives about this. That’s fine. I’m not here to comment on that. So everyone is free to form their own opinions. Just like no one has to be a Tesla shareholder. It has performed well historically. So if you don’t want to be a Tesla shareholder, you certainly don’t have to be. Francis, thank you. I appreciate it. Great to see you here. Greg, I know I thanked you already. And then there’s another one. I’m trying to pull out the name here. I’m not sure. 40W, thank you. I appreciate it. Alright. I really don’t want to end on that note. I hope we can find another question here. Another question I see is, why is the audio so bad? If it’s bad, I’m sorry. I’m trying to make sure that doesn’t happen, but I hope it’s okay. It’s not your audio, it’s Tesla’s. Yes. I hope they can provide better audio on the earnings call. It’s like, I thought you guys were a tech company. Hearing bad audio on the earnings call is almost nostalgic. Please go back and share your thoughts on the Hardware 3 FSD comments. I assume this is about the Hardware 3 upgrade path. So yes, that’s a tricky topic, right? Because so far, Hardware 3 may not have met the expectations Tesla set in 2017, at least I think over time, Tesla has adjusted some of those expectations of what it should be So people have varying degrees of expectations for that hardware. So, I've been through this, as I've been a Hardware 3 owner for a long time, right? About four years. I transferred FSD to the new car. Obviously, there was an additional cost, but I never felt it was unfair. My initial expectation was that this would be something that required long-term development, and no one could be sure if it would succeed. So maybe my expectations were a bit lower than what Tesla publicly set. So I recognized this, and I empathize. I do think that over time, Tesla has made a lot of attempts to make Hardware 3 work. I believe they will do their best to ultimately satisfy those customers. But just like a car has a limited lifecycle, right? You can't wait 20 years and then say, oh, we're going to upgrade your car. Oh, you don't have that car anymore. It's like, yes, that solves the problem, but it's clearly not the best customer experience. So I do think Tesla cares about this. I truly believe they will do everything they can to handle it correctly. But I also think they must, they absolutely must ensure they can get unsupervised FSD working, which takes priority over making those customers satisfied or feeling complete or anything else. Part of the reason is even because it's a lifesaving technology, right? It's like you're sitting there making the decision, oh, should we spend time upgrading people, spending resources to upgrade people to this hardware that isn't fully up to standard yet? Or should we ensure this hardware truly meets the goals, crosses the finish line, and starts saving lives? Clearly, the option that can save more lives should be chosen. This doesn't even consider the economic benefits, which provide a clearer picture: yes, you must solve FSD first. So this makes complete sense. It's a reasonable decision. It's the right decision. But recognizing that this decision will frustrate a large portion of the customer base, who happen to be early adopters, is precisely what has brought Tesla to where it is today. So this is a tricky issue, but I think Tesla's answer is correct, even if it is frustrating. So this is a good thing, and I think it absolutely should happen. Allowing FSD transfers, especially in markets where FSD isn't really available yet, should be taken for granted. So I'm glad Tesla is offering this. I'm glad they are expanding this service. And I believe over time, Tesla will do everything they can to satisfy these customers. This will lead to some people having bad experiences, which is unfortunate. I believe this was never Tesla's intention, nor is it the outcome they wanted, but sometimes this happens when you deal with really difficult issues. I'm not sure if there's a way to circumvent this unless it was managed better in the early stages, which is certainly a fair criticism, but would you hate this company forever because of that? I mean, I wouldn't, that's just my personal opinion. So, again, I emphasize that I think Tesla cares about them, cares about those customers, and will do everything they can to provide the best possible solutions. And Hardware 3, just like Hardware 3 has gotten better. It—it's still very good. If Hardware 4 doesn't exist at Tesla, it will be the best driving assistance technology you can buy on any car, like Hardware 3, which is already eight years old, and it’s still the best. So I think that's worth remembering. It's like, Hardware 3 is not bad, nor is it worse than anything else on the road It is much better than anything else on the road, except for Tesla's own other products. So I think this is very important context. Well. Um, I don't think this is my favorite way to end, but it's not bad either. Again, I want to emphasize that I think today was overall quite a bland day. I don't think we got any major breakthrough announcements or anything like that. This is not what I was expecting. I suspect many people were expecting this. Now the most noteworthy thing is what is happening in Austin, what is happening with the Robotaxi launch elsewhere. I think by the time we talk next, at the latest next quarter, we will see some quite dramatic progress. So, if nothing happens sooner, that will be the next checkpoint, and I look forward to continuing the discussion then. Well. Thank you, everyone. The event has ended. ---