Musk: "I cringe at raising prices repeatedly, hoping to lower them in the second half of the year" (minutes of telephone meeting) input: ====== 中国科学家首次成功观测到类星体 Lyα脉冲 ====== output: Chinese scientists have successfully observed Lyman-alpha pulses in quasars for the first time.

Here is the summary of the Q2 2022 Tesla.US conference call. For financial results, please refer to "What can we expect from Tesla if there is no “injection” from Shanghai factory?" on Longbridge.

1. Management report

1. Production: The Berlin factory exceeded production of 1,000 units per week, and the Texas factory hopes to achieve a production capacity of 1,000 units per week in the coming months (equals 50,000 units per year).
2. Battery: We are currently manufacturing cars using 2170 batteries, and we have enough 2170 batteries to meet production demands for the remainder of this year. Therefore, we do not rely on the 4680 battery. 4680 will be important next year, but not this year.
3. Autopilot: We have now provided FSD beta to over 100,000 car owners with urban street driving capabilities. We have collected more than 35 million miles of data, which is more than any other company's autonomous driving system that we know of. The number of miles is growing exponentially. The FSD beta will be released to all North American customers by the end of this year. We will also release FSD if approved by regulatory authorities in Europe and other parts of the world.
4. Manufacturing process: In Austin and Berlin, the number of body welding robots decreased by 70% per unit of production capacity. The body shop floor area is three times smaller and more lightweight and more affordable than normal conditions. In terms of noise, vibration, and roughness, we have better production and testing equipment and processes.
5. Cybertruck: Still hoping to start production in the middle of next year.
6. After the commissioning of the Berlin and other factories, factory investments have begun to be included in the manufacturing costs of cars, which will put some pressure on gross margins.
7. The most of our Bitcoin holdings were sold to lock in profits, but the remaining holdings were devalued, resulting in restructuring and other costs of 106 million yuan. In addition, targeted layoffs also incurred some restructuring costs.

2. Q&A with Analysts:

Q: China EV companies seem to be doing better than foreign competitors in software and design innovation, apart from Tesla. How can Tesla maintain its leading edge in China and other places in the face of such competitors?

Musk: Now, the best vehicle company in China is Tesla. We are the best, but I really take my hat off and salute the Chinese people and manufacturers, especially EV manufacturers. I think that China's EV industry will be a global force that everyone will see in the future. They are smart and diligent, and any company without their competitive edge will experience a decline in market share.

Q: When will Tesla have a unified vector space for static and moving object networks? Will it be in V11 or a higher version? Can you explain in layman's terms why this is a difficult issue?

A: I think only 0.01% of the audience can understand this answer. Essentially, if you combine static and dynamic objects in C++ within the same neural network context, you don't need to debug them in C++ heuristics. I believe achieving fully autonomous driving may not be necessary, but it will slightly improve the efficiency of autonomous driving. Of course, this is what we want to do. I think we will almost certainly achieve this goal by the end of this year.

Q: Musk said that once inflation cools down, the price will be lowered. How big will the price cut be? How do you view the long-term pricing trend of cars?

Master: Deciding whether to raise or lower car prices requires foresight. I think prices will fall by the end of this year. Some bulk prices are indeed declining, but macroeconomic factors are unpredictable.

When we see signs of a decline in the inflation rate, we don't need to raise car prices. But this fundamentally depends on the macro economy, which is beyond our control.

I want to once again urge entrepreneurs to enter the lithium refining industry. Although lithium is everywhere, mining is relatively easy, while the refining process is much more difficult.

The main thing is to refine lithium into battery-grade lithium carbonate and lithium hydroxide, and the purity must be very high. Doing this now is like printing money, which basically has the same gross profit margin as selling software.

Executive: In terms of battery metals, the price of lithium has risen sharply. We used to spend $11 to $80 per kilogram. But not every situation is so bad.

Carbon steel and aluminum have begun to trend downward, and we may see a decline in inflation in the second half of this year or early next year. Moreover, compared to some bulk commodities, the pricing of lithium is insane.

Q: What is the cost optimization path for 4680 compared to the traditional 2170? Is it economies of scale or technical optimization?

A: The 2170 and 4680 architectures are different and will consider better utilization of physical and chemical properties and battery pack structure optimization. The battery pack is used for energy storage on the one hand and for automobiles on the other. We will consider optimizing different 4680 battery cell application scenarios and shapes. In terms of cost, we need to scale up and collect more data to optimize technology. Technology and scaling are both very important for reducing costs.

Q: What is the progress of FSD? Andrej Karpathy is leaving, does this have a significant impact on the timeline or potential progress?

A: Because Andrej wrote all the code himself, things naturally got stuck. Andrej is also someone we greatly respect. He decided to make more contributions to core AI at the academic level, and then return to independent coding work. But our software AI team has 120 talented individuals. I am very confident that we will solve the problem of full autonomy, and it may even be this year.

Q: What is the progress of 4680? Has the Austin factory started producing 4680 batteries?

A: Yes, the Austin factory has started producing 4680 batteries, but the quantity is not high enough to meet production needs. We expect that by the end of this year, the weekly output will be able to supply more than 1,000 vehicles.

Last quarter, in Texas, the battery equipment was completely installed and debugged, and we produced the first batch of test car batteries. Our goal for Texas is to start production this quarter, and the goal is for Texas to exceed the weekly output of the Fremont Kato battery factory by the end of this year. Q: How is the production ramp of Austin and Berlin? What about the supply of components like semiconductors and batteries? Will cost increase affect the profitability of other factories?

A: Tesla purchases around 1600 types of chips from 43 semiconductor companies. Therefore, challenges always exist for such a combination of purchases of this scale. The latest generation of chips is more stable, and we still see supply shortages of older generations of semiconductors, especially in the field of analog and mixed-signal chips. But we have ways to solve the problems faced by Austin and Berlin.

In terms of batteries, our profit is not bad, mainly due to the strong production of our supply partners, and the production capacity and factory output plan are matched within the time frame we can see.

We have greatly increased our battery production over the past 12 months and have signed long-term contracts with all key battery metal partners. Therefore, we have not seen any significant issues with components, of course, unless there are unexpected pandemic-related lockdowns.

Regarding the profitability of new factories, the past two quarters are the most inflationary ones among many quarters in the past, and every factory has this problem because it is a problem of the overall supply chain.

Last year, key raw materials began to rise, rose in the first quarter, and rose faster in the second quarter. From an annual perspective, I don't think the third quarter will be higher than the second quarter.

Of course, we will not be fully exposed to the risk of price increases: For key raw materials, we all have contracts in hand, but there are indeed some that need to buy spot goods, and some contracts are being renegotiated.

Q: How much improvement has been achieved with the 4680 battery?

A: The optimization of the 4680 cell structure is very important. We strive to improve efficiency, reduce complexity, and achieve scalability in the future, but we don't want to be limited by the technology itself. We need all links to work together to fully tap the potential of technology. In order to achieve more production capacity ramp-up, we will consider the introduction of new material technology.

Our focus now is on dozens of small issues that are inhibiting the growth of 4680 production. Some more challenging issues are powering the positive and negative electrode materials because we are using this revolutionary dry electrode process, which has many unknowns that must be solved. We are confident in solving these unknowns, but it is very, very difficult. So, the first thing to do is to do the basic work well, obtain high capacity and high reliability, and then iterate quickly on this basis to improve energy density and reduce battery costs.

We will maintain close communication with suppliers, show them our important technologies, and achieve better coordination with them. Throughout the entire supply chain, our efficiency is subject to the slowest link.

Q: You always mention the annual revenue growth target of 50%. How do you view the difficulties in commodities, raw materials, and price fluctuations?

A: We do see the limitations of materials required for refining lithium-ion batteries. What I want to emphasize is that this is not due to the scarcity of raw materials. Take lithium as an example. Lithium is one of the most common elements on Earth. It is almost everywhere. However, it is quite difficult to extract lithium and refine it into ultra-high purity battery-grade lithium hydroxide and lithium carbonate, which requires a large amount of mechanical equipment and is difficult to scale up. It is also difficult to manufacture anodes and cathodes. But anyway, there is not a shortage of materials.

The company believes that the cathode material has a relatively large development potential, especially the iron element in it, which performs better in energy storage batteries;

The company will continue to adopt integrated casting technology, which can reduce costs and improve automotive energy efficiency.

Q: Is there a trend of slowing consumer orders? Will the company be concerned that economic issues will affect Tesla's orders?

Our problem is mainly production capacity, and it takes a year or more to deliver a car. Production capacity is the core issue. From Tesla's situation, there is no macro impact on orders.

In terms of vehicle prices, we have indeed been raising prices recently, to a somewhat embarrassing level, but we also have supply chains and trucks, and we are facing crazy inflation. Although it is not a commitment, I hope our prices can drop a bit.

Q: Can you talk about the pricing strategy for FSD? With the full promotion of FSD functions, how do you think it will develop from the rest of the time in 2021 to 2023?

A: Yes, we will increase the price of FSD later this year. I think the value of FSD is very high, but it is not well understood by most people. Assuming that FSD is achieved, its current price is ridiculously cheap.

Q: Tesla has undergone huge changes in the past three years, from being on the verge of death to having stable cash flow and industry-leading profit margins. I want to know if you can talk about your personal role in the company and whether you think there will be any changes in your role, commitment and time in the company in the next 3 to 4 years?

A: I've done a lot of unusual things. Basically, if there's only good news, I'm not on call all the time. But if there are difficult situations like the closure of Chinese pneumonia, I think I would be on call — relatively speaking, if there is bad news. But I'm committed to the long term — what I mean is that I will work at Tesla as long as I can effectively promote sustainable development and autonomous driving.

Q: How to deal with the problem of semiconductor shortage? Are there any internal efforts to improve supply chain performance?

A: We have been working with suppliers. We are not currently planning to make chips ourselves. We do not believe there will be a need to manufacture chips, but we have been working closely with some suppliers, and we already have a lot of customized new products in the car, such as microcontrollers, so we are trying to find the real technical advantage. In the future, I think we will focus on where there are suppliers.

A: I think Tesla is both a hardware company and a software company. Therefore, one of the ways we can solve the front-end supply chain problem of chips is to rewrite the software so that it can use different chips, or in some cases, achieve dual use of a single chip, which is even better. In fact, frankly speaking, the chip shortage has forced us to reduce the number of chips in cars. Yes, we originally had more chips than we needed. But this is proof to our software team that we can install a new chip in the car without interrupting production and write a brand-new software patch for that chip. Our goal is to integrate more features into fewer chips as the platform matures and expands, much like laptops and smartphones. It's the same in cars. We're working hard to do this as quickly as possible wherever it makes sense. At a high level, we're not designing and manufacturing our own microcontrollers, but working with key partners who understand the architecture requirements to design something for us according to specifications. And we've done it.