Elon Musk: The lunar program is still advancing, of course. If everything goes smoothly, a self-sufficient Martian city will be built within 30 years.

Key points:

• Optimus will achieve large - scale production in 2030, and the cost of each robot is expected to be about $40,000.
• Tesla is developing the AI5 chip, with a computing power about 8 times that of AI4 and a performance improvement of 40 times.
• SpaceX will release Starlink mobile phones in two years and may acquire traditional mobile operators in the future.
• The xAI team is training the Grok 5 model. Musk believes that by 2030, the overall capabilities of AI may exceed the sum of all human capabilities.
• The declining population and low fertility rate have led to a stagnant or even declining state of human intelligence, and humanity faces the risk of "self - destruction".
• On September 15th, during the recently held 2025 ALL - IN Summit, "the Iron Man of Silicon Valley", Elon Musk, appeared via remote connection. He talked about the latest progress of Optimus, Tesla's self - developed chips, the Starlink mobile phone plan, and the Grok model, and shared his views on artificial intelligence, the population crisis, and the future of humanity.

Musk is depicting a future where artificial intelligence will surpass the collective wisdom of humanity before 2030, while also warning that Western society is falling into a crisis of population decline and "self - destruction".

The following is a selected version of some of the conversation content:

01. Musk Devotes Most of His Energy to Optimus

Question: Optimus is considered the greatest product in human history. What's the current progress? How much energy have you invested in this product? Is it the third or fourth version now?

Musk: We are finalizing the design of the third version of Optimus. This robot will be very excellent. It will have hand dexterity close to that of humans, which means it will have a highly complex robotic hand, and it will also be equipped with an artificial intelligence "brain" that can navigate autonomously and understand the real - world environment, and it will achieve large - scale mass production. These three points are the key elements that other robot companies do not have.

Currently, among all projects, Optimus may take up most of my energy, more than any other single project. This requires solving real - world artificial intelligence problems, mechatronics problems of Optimus, and challenges in the supply chain and production. We have to build everything from scratch because the supply chain for humanoid robots does not exist at present. This requires a large amount of vertical integration, and all the actuators used in Optimus cannot be obtained from the existing supply chain. If successful, Optimus will become the largest - scale product in history. By 2030, after achieving large - scale production, the cost of each robot is expected to be about $40,000.

Question: Then what will be the cost of the first batch of robots? When can we buy one for ranch operations? You previously mentioned that once the annual production reaches one million units, the marginal production cost will be about $20,000.

Musk: This will be affected by the cost of the artificial intelligence chips in the robot. The efficiency of the actuators also needs to be greatly improved. Each arm has 26 actuators, including motors, gearboxes, and power electronics. The cost of artificial intelligence chips is not low and may account for $5,000 to $6,000 or even more of the bill of materials cost.

So, when the annual production of Optimus reaches one million units, the production cost is expected to be between $20,000 and $25,000, and the final selling price will be determined according to market demand.

Question: Over the past few decades, you have accumulated rich experience in factory operations. Which is more difficult, building Optimus, developing the Cybertruck, Model Y, Model X, or building the Gigafactory?

Musk: Building Optimus is much more difficult than these, but a little easier than the Starship. The difficulty lies between that of the Model X and the Starship.

Question: Which is more challenging, hardware or software?

Musk: At present, the main difficulty lies in the final design of the hardware, especially the design of the hand. Of course, other parts of the robot are also important, but the design of the hand, including the forearm, accounts for most of the engineering difficulty of the entire robot.

Question: Assuming we overcome the hardware challenges, with the current development of large - language models, what kind of assistance can we directly get from them? Can consumers directly interact with the robot, give instructions, and can the robot understand and execute them?

Musk: Of course, that's no problem.

Question: I noticed that you often interact with artificial intelligence online. Can you create a robot similar to "Any" (referring to a certain type of virtual intelligent agent)? Why choose to design a humanoid robot?

Musk: If you want a robot to be able to do everything that humans can do, a humanoid form is essential. If it is only for a specific type of task, the design will indeed be much simpler. The human form and abilities have evolved for a reason. For example, having five fingers is very practical, and even the little finger has an important role. Moreover, humans have designed the world according to their own needs, so humanoid robots can be directly adapted to the existing environment and achieve "backward compatibility".

02. Tesla Adopts a "Dual - Chip" Strategy, and the AI5 Chip Will Achieve a 40 - Fold Leap Compared to AI4

Question: In addition to large - language models and hand actuators, robots also need chips to drive them. You previously mentioned AI5 and AI6, and it seems that you are very optimistic about the development direction at the chip level. Can you introduce the relevant situation?

Musk: At Tesla, we have two chip projects: one is Dojo for model training, and the other is the AI series of chips for inference. Currently, all Tesla models are equipped with the AI4 chip, and now we are finalizing the design of the AI5 chip.

Compared with AI4, its performance will have a huge leap. From some indicators, the performance of AI5 will be 40 times that of AI4. The reason for such a large improvement is that we have carried out very precise collaborative design in the research and development of artificial intelligence software and hardware and clearly know where the bottlenecks of the existing technology are.

Question: For consumers, in which aspects will this performance improvement be reflected? For example, will the safety of Tesla's Full Self - Driving (FSD) that Tesla owners experience be improved by an order of magnitude? How will the performance of the robot change?

Musk: To be precise, the 40 - fold improvement is aimed at the biggest shortcoming of the AI4 chip, that is, the softmax operation (a commonly used activation function in neural networks). Currently, the AI4 chip needs to run about 40 steps in the simulation mode to complete the softmax operation, while the AI5 chip can complete it in a few steps through its native architecture.

In addition, AI5 can easily handle mixed - precision models, achieve dynamic adjustment of precision, and there are many optimizations in technical details, making its performance more excellent. In terms of basic computing power, the computing power of AI5 is about 8 times that of AI4, the memory capacity is about 9 times, and the memory bandwidth is about 5 times. Because we have solved the core bottleneck of AI4, we can, on the basis of an 8 - fold increase in computing power, superimpose a 5 - fold underlying optimization effect, ultimately achieving a 40 - fold performance leap.

Moreover, the AI4 chip currently installed in Tesla electric vehicles has already made the safety of Full Self - Driving reach 2 to 3 times that of human driving, and it may even reach 10 times. The software adapted to this performance will be pushed in the next few months. This will be the largest - scale software upgrade for Tesla since the version 12 software. We will increase the number of model parameters by an order of magnitude and apply a large amount of reinforcement learning technology.

03. Starlink Mobile Phone Plan: The First Batch of Spectrum - Compatible Phones Will Be Launched in Two Years

Question: You spent about $17 billion on purchasing spectrum with the aim of enabling the Starlink satellite network to directly connect with mobile phones. What will this technology look like in one or two years? Will we abandon the mobile operator Verizon and switch to Starlink's service? How many people want Starlink mobile phones? Is this feasible from a technical perspective?

Musk: This is a long - term project. Through this project, SpaceX can directly provide high - bandwidth network connections to mobile phones via satellites. However, to achieve this, the mobile phone hardware needs to be modified because current mobile phones do not support these new spectrums, and the chipset must be upgraded. The whole process will take about two years.

So, the first batch of mobile phones supporting these spectrums is expected to be launched in two years. At the same time, we are also developing satellites that can be adapted to these spectrums and cooperating with mobile phone manufacturers to add relevant spectrum support functions to new mobile phones. Eventually, satellites and mobile phones can achieve efficient connection, and users can watch videos smoothly on their mobile phones anywhere.

Question: This sounds amazing. Can these spectrums be used indoors like current mobile phone signals?

Musk: If the building has a thick metal roof, it may be affected, but there will be no problem using it in ordinary residential buildings.

Question: Do you hope to achieve such a vision: users no longer need to rely on regional operators such as AT&T, and they don't have to worry about roaming when traveling to the UK or India. They only need to sign up for Starlink to use the network. Eventually, Starlink will achieve global coverage and replace regional operators to become a global operator?

Musk: This is a possible direction, but we will not let other operators be eliminated. They will still exist because they have a large amount of spectrum resources. However, users can indeed have a Starlink account: they can use it with the Starlink antenna at home to enjoy free Wi - Fi and also use the high - bandwidth network on their mobile phones, forming a comprehensive high - bandwidth solution covering home and mobile scenarios.

Question: Then will you consider acquiring an operator to obtain more spectrum? For example, acquiring Verizon?

Musk: This possibility cannot be ruled out.

Question: Let's talk about the Starship again. The recent launch was very successful. How far is it from achieving stable commercial application?

Musk: I think we will be able to recover the spacecraft next year. Currently, there is only one more mission left for the Starship launch plan of the second - generation Starlink satellites, with only one set of the second - version booster and spacecraft remaining. After that, we will launch the third - generation Starship, which will be equipped with Raptor 3 engines, and almost all parts of the rocket will be upgraded.

Due to the large design changes, the third - generation Starship may encounter some problems in the initial stage, but it can achieve full reusability and can send more than 100 tons of payload into orbit. As long as there are no major accidents, SpaceX will be able to demonstrate the full reusability of the Starship next year, including recovering the booster and the spacecraft, and can send more than 100 tons of payload into a practical orbit. Currently, the best - performing commercial rocket is the Falcon Heavy. When its side boosters are reused, its payload capacity is about 40 tons, while the Starship's payload capacity is 2.5 times that of it, and it can be fully reused.

Question: There was an explosion during a previous Starship launch, and then there was a lot of discussion about environmental impact and the review by the US Federal Aviation Administration (FAA). But you quickly resumed launch preparations. This is not only an achievement in terms of technology and execution. How could you obtain regulatory approval so quickly? Many people thought there would be a lot of reviews and questions at that time.

Musk: There were indeed many review processes, but we passed them all smoothly. This is thanks to the SpaceX team. They made great efforts, quickly completed the testing and assembly of the next set of booster and spacecraft, and transported them to the launch pad for launch.

Building a fully reusable orbital rocket is one of the most difficult engineering problems in history, at least in the top three. This has been the goal of SpaceX since its establishment in 2002. Now, 23 years have passed, and we have the world's best rocket engineer team. I believe we will definitely achieve full reusability of the rocket next year.

Question: From now until achieving full reusability, what is the technical obstacle you are most concerned about? Is there a key problem that you need to focus on solving, or do you need to integrate previous lessons into the next launch?

Musk: The main challenge in achieving full reusability of the spacecraft is the heat shield (a device or structure used to protect spacecraft, rockets, or other high - speed flying objects from extreme high temperatures).

So far, no one has been able to build a fully reusable orbital - class heat shield. For example, the heat shield of the space shuttle required 9 months of maintenance after each flight. This involves both materials science problems and engineering technology problems.

We are starting from the basic physical principles to develop a heat shield that can withstand extreme high temperatures, is lightweight, does not transfer heat to the main structure, can resist rain, and will not dissolve in the rain. Moreover, tens of thousands of heat - resistant tiles on the heat shield must all be reliable and do not need to be inspected or refurbished one by one after each flight like the space shuttle.

04. Grok Development Progress: Synthetic Data Reconstructs Knowledge, and AI Capabilities May Exceed the Sum of All Human Capabilities by 2030

Question: Can you talk about the progress of the Grok model and the xAI company? You previously mentioned that the training of the next - generation model will not rely on conventional web data crawling but will use a large amount of synthetic data. What is the development direction of the Grok model? Why is this technological innovation important?

Musk: We are investing a large amount of inference computing resources to analyze the original data in the human knowledge corpus. In this process, Grok will sort out each piece of information one by one, supplement the missing content, correct the wrong information, and eliminate the false content. Taking Wikipedia as an example, this processing method is not only applicable to Wikipedia but also to all information carriers such as books, PDF documents, and websites. Specifically, Grok will analyze each page through a large amount of inference calculations, judge the authenticity of the content, and then rewrite the page, delete the false content, correct the half - true and half - false information, and supplement the missing context.

Question: Can you make these corrected contents public? For example, create a "Grokipedia" (a coined word combining Grok and Wikipedia, which can be understood as "Gro Encyclopedia")? Especially, there are many problems with the biography pages on Wikipedia, which are full of biases and often have editing disputes.

Musk: I will discuss this idea with the team. Building a platform similar to "Grokipedia" should be very interesting.

Question: Regarding the training of the Grok 5 model, you are expanding the Colossus super - cluster in Memphis. Can you introduce the latest situation? In addition, from the perspective of Scaling Laws, can expanding the scale of the super - cluster create a more powerful artificial intelligence model? Will there be a situation of diminishing marginal returns?

Musk: There is a natural logarithmic relationship between the investment in computing resources and the improvement of the intelligence level of the artificial intelligence model. Roughly estimated, a 10 - fold investment in computing resources can double the intelligence level of the model. For example, increasing from an IQ equivalent to 100 of a human to 200, which is still a very significant improvement.

Moreover, the intelligence level will continue to improve until we can use the energy of the sun or even the entire galaxy to provide computing support, reaching the computing power scale of a Type II or Type III civilization on the Kardashev Scale (Type II can use the energy of its entire star system (such as the solar system); Type III can use the entire galaxy