一项涉及 Tesla 的潜在合并会合理，并有利于 Tesla 股东，但远未确定：批准需要获得全部股份 50% 的赞成票，而不仅是参与投票股份的赞成票，且特别委员会尚未就条款达成一致。

Regarding the merger with Tesla. Therefore, at the request of shareholders, both SpaceX and Tesla have formed special committees to discuss this idea.

I've never seen anything grow so fast. In short, this will result in SpaceX having annual recurring revenue of between $500 billion and $ 1 trillion by the end of 2027.

Revenue is $14.6 billion, up 161% from $5.6 billion. Net income was $4.4 billion, an improvement from a net loss of $1.1 billion to $5.5 billion.

And consolidated EBITDA is $8.8 billion, up 306% from $2.2 billion. Starting with Starship , Flight 14 was a significant milestone. Starship reached orbit for the first time and deployed operational Starlink V3 satellites.

For Flight 15, which we expect to take place next month , we will try to capture both the booster and the spacecraft. If we are successful , this will be the first time an orbital launch vehicle will be fully recovered.

I think there will probably be 20 or more launch towers at some point. We expect Starship to fly on a weekly basis sometime in the first half of next year , and then twice a week in the second half of the year, or maybe even more frequently by the end of next year.

Starlink has also just started. Flight 14 deployed Starlink's 26 terabits per second bandwidth. This is equivalent to 10 Falcon 9 launches. So, a single Starship launch, which was not fully loaded, perhaps half loaded, did the equivalent of 10 Falcon 9 launches.

So, I expect that over time , which is not in the very distant future, probably within the next 3 years, we will have about a hundred times more bandwidth than the current Starlink constellation.

Because we will launch satellites with ten times more bandwidth and put ten times more Starlink satellites in orbit. So, there will likely be about one hundred thousand satellites , each of which will have more than ten times the bandwidth of the current Starlink system.

So, I think Starlink revenue will be at least somewhat proportional to the total bandwidth delivered, and we will also have high- bandwidth direct-to- cell connectivity capabilities within three to five years.

There will likely be a thousand times more bandwidth demand than what is currently available, because AI or ASI agents will use much more bandwidth per agent than a human, and there will be many more agents than humans.

We plan to bring approximately 200,000 GB300 online per month for the rest of this year and will maintain this pace next year as well. Right now , I would say that I am quite confident that we can bring an additional 10 gigawatts of compute online by 2027.

This will be about 3 gigawatts, maybe an additional 2.5 to 3 gigawatts, that we will have online by the end of this year. So, I would say this, I would be surprised if we have less than 13 gigawatts of compute online by the end of next year.

And as I mentioned earlier , our power, cooling, and electrical infrastructure is somewhat of a limitation. So, to ensure that this does not hinder our construction work, we are aiming to build an additional 20 gigawatts of power and cooling infrastructure next year.

As a result, the amount of power and cooling we have will be much more than what is needed to run the GPUs. So, there is a possibility that SpaceX will bring more than 20 gigawatts of superintelligence compute online next year , which is not zero.

The possibility may be low at the moment, but it's not completely zero. The probability is low, but it was zero before. It's no longer empty.

We also plan to bring the most powerful training cluster online by the end of the second quarter (Q2), where 440,000 Vera Rubin computers will be deployed for optimal training.

I think this will be extremely helpful in making Grok the best model possible. And of course that will make it highly competitive. In the case of Grokbot, I have never seen such rapid growth of any product in my life. I've never seen anything grow so fast.

We also see that the usage of Grokbot per user is increasing, not decreasing, over time. So, as we make Grokbot more useful , we're seeing users using it more, not less, each week.

And we will deploy Grokbot on a large scale for enterprise customers, which is a very important thing, so that it can be used effectively by teams and companies , just as if it were a virtual person.

Um, we've already launched it in the early stages with some trusted customers, and we expect to have Grokbot widely available for enterprise customers within a few weeks.

I think Grokbot will probably use the most of our superintelligence compute. In short, this will result in SpaceX's annual recurring revenue of between $500 billion and $ 1 trillion by the end of '27.

Yes. Sorry. I thought I would say that strange thing. Yeah, I do n't want to give Brad a heart attack, you know. Thank you, Elon. So, regarding the merger with Tesla, I understand that you might have limited say , but if we step back a little , what do you think about the pros and cons of this potential merger?

And as part of that framework , are there any obstacles that you need to address before you can start it ?

Um, I mean, this is definitely a question where I feel like I'm dancing through a minefield, um, but, so I have to be thoughtful about what I'm saying here.

Um, um, I think, um, so, so, so, both SpaceX and Tesla have formed special committees to discuss this idea at the request of shareholders, um, and, um, these special committees are, um, made up of independent, um, directors, and since I'm a conflict of interest party, I can't actually be in the discussions of the special committees.

So, it's like a weird Kabuki dance where two groups of people are talking and I can't stay in the conversation. I mean, I do n't think we should necessarily assume that. I think some people on the outside assume that the merger of SpaceX and Tesla is a foregone conclusion , and I think that's an exaggeration.

Well, then do I think this would be a good plan? Yes, but at the end of the day, it has to have shareholder support, and especially on Tesla's part , we need 50% of all the votes, not just those who vote, but all shareholders .

And that's a pretty big obstacle to achieving it.

So, all I can say is that the special committees have not yet reached an agreement on the terms of any transaction, and there is no guarantee that they will.

But I think there are many benefits to combining the two companies, because we are now at the stage of the superintelligence singularity and there is a lot that can be achieved by collaborating on things like superintelligence software, chip design, Terra Fab, robotics, sending mass into orbit through the Starlink constellation.

Connection. Connection. So, I think it would be a good idea, but it's actually up to the special committees of both companies. They have to come up with something and then present it to the shareholders for a vote.

Now that I'm done with the shock , let's get to some of the important parts of the actual income report. If you want to move forward with timestamps, there's a lot more to discuss , especially with that groundbreaking bombshell news.

We now have confirmed news. Committees have been formed at both Tesla and SpaceX, which are discussing the possibility of a merger between the two companies.

Revenue this quarter was $14.6 billion. Incidentally , this is 161% more than last year and almost double the second quarter. Who will repeat this? SpaceX nearly doubled its revenue in a single quarter.

Superintelligence alone generated $7.56 billion in revenue. More than half of SpaceX's total revenue comes from their superintelligence business alone. And look at these skyrocketing numbers , bro.

Superintelligence revenue grew 195% , nearly tripling, compared to the previous quarter. An increase of 830% compared to last year. SpaceX is also hugely profitable. This quarter alone, they made a profit of $4.4 billion , compared to a loss of just over $1 billion in the same quarter last year.

Operating income has also been positive and not insignificant. It lost $143 million in the second quarter. Profit in the third quarter, i.e. operating income. I feel sorry for the heads of the finance department. Actually, shut up. $2.79 billion.

Super Intelligence's operating income alone is over 420 million, and that's after the 1.2 billion spent in stock compensation after the acquisition. EBITDA, which is basically overall profit, increases dramatically.

By the way, these are Grok's words, not mine. Again, the Grock bot created this presentation seconds after the report was published. It said, " Brother, look at the report." It's getting a little meta , but it's really great.

The Grok bot was monitoring the report. As soon as the report was published , he was briefed in advance, given a run-through , the main points were highlighted, and a presentation was made.

Well, that's all. I'm not exaggerating, just visually creating it in our own style. EBITDA increased dramatically. No lies were caught. Adjusted EBITDA increased 306% year-on-year , while margins were 60% , up from around 45% in the second quarter.

Along with this, capital expenditure ( CapEx) also increases vertically. Capital expenditure of $19.6 billion. The bulk of which, $16.4 billion, was spent on superintelligence.

However, most of this will be primarily AI compute. And to give you a little context , this is against 13.1 billion and 9- month operating cash flow.

As we know, the successful Starship Test Flight 14 launched 26 V3 Starlink satellites for the first time. That is, Flight 14, which is still technically a test flight, is actually generating revenue.

These are the real Starlink satellites. A massive increase in total network bandwidth, which means more users can now use the Starlink network. The more frequently Starships fly, even during initial testing, the more they will recover their costs and make more profit if they deploy Starlink satellites.

Starlink has 13.6 million users, with 1.6 million customers added in the third quarter. Revenue from connectivity alone is 5.25 billion, up 69% from last year. And the average revenue per user is also $69.

In the second quarter (Q2) it was $66. A year ago it was $83.

Obviously, this trend is generally expected to continue to decline as Starlink costs decline over time. They will be able to pass those savings on to customers, which means the product will be more competitive, and more customers will choose Starlink.

But interestingly, the average revenue per user technically increased by $3 quarter-over-quarter. I thought it would go the opposite way.

The compute ramp has now reached 1.7 gigawatts of nameplate compute. In the second quarter (Q2) it was 1.4 GW. A year ago it was 0.6 gigawatts. Not far off from an almost threefold increase per year, from 0.6 to 1.7.

And this is just the beginning for SpaceX's AI compute business, which, I remind you , was a major contributing factor in their highly profitable turnaround from the second quarter to the third quarter.

The Carsdeal has been completed , which will probably be considered one of the best acquisitions of all time. Without the Kars acquisition, Grok Bot would not exist, which is literally the best product of all time.

This resulted in an additional revenue of $608 million. In addition, more than $23 billion in equity awards remain to be vested.

SpaceX also has a huge fund of around $90 billion. This includes cash and marketable securities, that is, anything that they can convert into cash if needed. That's a huge amount. I mean, brother, about 100 billion dollars.

35 launches this quarter. And here's a cool fact: the first launch from Starbase in Louisiana is targeted for 2029.

And the deal that shook up the US telecom market , the 800 MHz spectrum deal for Starlink Mobile, is expected to be completed in the first half of 2027. Very soon.

P.S. In case you forgot, Starlink's subscriber base, especially in the cell business, is growing exponentially and, in my opinion, will soon surpass Starlink's total broadband subscriber base.

These were the main points of the earnings report, but we need to watch this clip again and again. That means, in short, SpaceX's annual recurring revenue (ARR) will be between $500 billion and $ 1 trillion by the end of '27.

These numbers are not made up by Mask. It's important to understand that the vast majority of SpaceX's revenue in the next few years will likely come from AI compute. And it's very easy to calculate how much this income can be, just multiply the AI compute spot rate by the amount of AI compute you have running.

If the price per gigawatt is $50 billion, then you have 10 gigawatts in operation. The calculation is quite simple. This means half a trillion dollars in ARR.

Normally, I would have had a lot more to say about this earnings call. We also need to make a second part of this video. There were n't many topics discussed there, but there was a lot of coverage of the core topics of AI compute business, AI models, and what is now being called superintelligence.

And the biggest surprise was the official announcement that a special committee had been formed to discuss the possibility of a merger between Tesla and SpaceX. Musk cannot be involved in this, as he mentioned.

And the important thing is, as he emphasized and I have been saying for months, that while this idea is very logical in theory, the actual hurdles or standards for making it a reality are very high. 50% of the total shares must vote yes.

Those who do not vote or participate will automatically lose their vote. Many investors around the world literally cannot vote on their own shares. Their shares are held with a lousy brokerage firm that does not allow them to exercise their voting rights as shareholders.

Many people do n't bother about these things, and do n't even want to go through the trouble of voting. It is clear that this deal, i.e. the merger, is not yet certain, even if most investors are theoretically in favor of it.

They may not all be able to or will not vote on this issue.

The merger undoubtedly makes sense and is in the best interest of both SpaceX and Tesla shareholders. I modeled this in great detail in my 20-year Tesla and SpaceX valuation. We have several separate sheets on the merger scenario, its combined benefits, savings, and advantages.

Check it out on Patreon at the link below. It is available to investor-level supporters and above.

But Musk's argument, while theoretically sound and logical , in reality the hurdle to overcome is enormous. The merger is far from a sure thing. The good news is, as I've been predicting for the past few months , actually not even predicting , I'm telling you exactly what has already happened because I have a time machine, it's actually already happened.

I'm telling you before you experience it. Every quarter SpaceX keeps its promises, launches a huge amount of AI compute, monetizes most of it , makes a lot of money, maybe a little less than Musk's incredible estimate of IRR, but close enough to it.

Many Tesla investors, like me, who previously didn't spend time modeling SpaceX, are now doing so. Thanks for the link to my valuation model below,

many Tesla investors, who previously naturally opposed SpaceX because they didn't see its true value , will also be nervous and strongly support the merger in principle.