원문 보기
To understand the nature of the universe and all these fundamental questions, we must expand the boundaries and scope of consciousness. So, it's been a while since the last conversation at the company. I try to have these meetings every three months, after our earnings reports, when I'm essentially allowed to talk about something.
We'll start by going through the history of SpaceX from its beginnings to the present day. I think sometimes it's useful to look back at the whole story and then think: what should the next chapters be? What should the next book in this story be? So, we started with a clear mission: to make life multiplanetary, to spread consciousness beyond Earth, to understand the universe, and you can only understand it by going out there, exploring other planets, and eventually other star systems.
And I think one of the most inspiring futures you can imagine is one where we, among the stars, perhaps encounter alien civilizations or explore the ruins of ones that existed for millions of years but disappeared 10 million years ago. Ahem. I think that's what I think about when I wake up in the morning, thinking about what inspires me, because we need things in this world that inspire me.
Life cannot consist only of solving one problem after another. There should be things that make you look forward to the future . And becoming a space-faring civilization is, in my opinion, the most exciting thing we can do for the future. This means making Star Trek a reality. You know, this is, well, yes.
Well, that's how we started. Yes. So, I don't know, maybe I should invite another mariachi band or something. You know, at the New Year's Eve party we're going to have a lot of mariachi bands and a lot of things like that . But that was what SpaceX was all about in the beginning. Just a small warehouse in El Segundo.
And then Falcon 1 , and in the early days of SpaceX, we really had no idea how to actually build rockets. And of course, we blew up the rocket three times before we finally reached orbit. It almost looks too cute now. Ahem. And then we went through a long journey of trying to land an orbital booster.
So, it wasn't easy to figure out how to land an orbital booster and then launch it again. But over time we learned to do it well. It took many attempts from a very talented and dedicated team. And now we've gotten to the point where we're taking off, landing, and recovering the booster every few days.
By the way, is n't the noise behind me too loud? Are you there, are you...are you okay? Good. OK. Um, so I think it's always important to be aware of the difficulty of achieving goals. Because when we tried to land the accelerator, no one really believed we could succeed. And many experts in the aerospace industry told me that you wouldn't succeed.
And even if it works, no one will buy a rocket that was, in fact, already in use. Which doesn't really make sense since planes fly all the time. And, honestly, I would rather be on the plane's thousandth flight than its first. So we changed the nomenclature to " flight-proven accelerator," which is true.
So, we managed to make Falcon Heavy, which is the most powerful rocket in the world before the appearance of Starship. At least operationally. Aha. And yes, it's strange to think that our 700th flight is approaching. This is a huge testament to the incredible work of the Falcon team, the launch team, and the production team.
I would like to thank everyone for such a wonderful job. See? And, of course, it's a classic. By the way, I fully expected this flight to end in failure. So, this is the same car that I drove around Los Angeles for about 3 years. And I thought, well, if this is the first flight of the Falcon Heavy, it's probably going to fail.
But we didn't want to put something boring, like, say, Boeing or someone else would put a concrete block, which is extremely boring. So we sent Starman flying around, and he's still there. The machine rotates between Earth and Mars. So it will stay there for, I think, at least 10 million years. And, of course, Dragon.
The first privately developed spacecraft to dock with the space station. And the first commercial astronaut. We have already completed many missions with astronauts. Um, I think there's not much time left before we cross the 100- orbit- return mark. So, Dragon became an incredible ship. We have already transported 78 crew members to date.
And it was, um, it was an epic success. So I also want to congratulate the Dragon team. Great job. And now Starlink: if you have a rocket that can fly frequently , what would you launch on it? So the Starlink team built an entire global internet system from scratch. And this is the first ever global high-bandwidth internet system .
And I actually think that over time it will probably, I think, it will probably power most of the internet. In fact, I think it could eventually handle over 90% of all internet traffic. So it's really incredible when you think about it. Almost 11,000 satellites in orbit. That's twice as many satellites as all the others combined.
And with Starlink V3 and beyond, we will reach 100,000 satellites. So, that's really quite impressive. It's quite difficult to build the internet in space. The whole...We built the whole internet in space. All hardware, all software. And yet the team did it. So a huge congratulations to the Starlink team.
Yes, we are now in 167 countries. 22 million mobile subscribers, 13 million high- bandwidth subscribers. And as I said, expect these numbers to grow more than 10 times over time. And in the long term, by the way, Cloudflare recently mentioned that based on current growth, they believe AI traffic will be a thousand times larger than human internet traffic in 5 years.
Just realize this for a moment. You know. That's a lot of AI. And we also have Starshield, which has become an important part of national security . Therefore. We can't say much about Starshield for obvious reasons, because it's, you know, top secret and stuff. But the Starshield team did a great job, and it is now a vital part of America's national security.
So I got distracted. Yes, Starship is amazing. Wow. It's hard to believe that humans created such a thing. I think the AI of the future will look back and say: not bad for a bunch of people. So. So, Starship is obviously an incredible ship . Um, and this is what will allow us to go from about 2,500 tons per orbit per year to over a million tons per orbit per year.
And I think maybe even up to 10 million tons per year. These are staggering numbers. Um, but I would say we're likely to put over a million tons a year into orbit with Starship, and probably around 10 million. Which is, you know, awesome, of course. Ahem. But this is the tonnage needed to build a civilization on the Moon, Mars, and beyond.
If you can't get millions of tons into Earth orbit, you can't get the millions of tons needed to the Moon and Mars to make us a sustainably multiplanetary species. So this is, uh... an incredible ship . Uh...As I said, I think this is probably the most difficult thing humanity has ever tried to do. And the Starship team did an incredible job in all areas: engines, avionics, structures, launch operations.
So a big thank you to the Starship team. So, where is everything headed in terms of SpaceX AI? Ahem. The crazy thing is that our AI revenue—probably, not probably, but definitely—will surpass all of SpaceX's other revenue, probably in September. That is, next month. Um, and will far exceed all of SpaceX's other revenue in the fourth quarter.
So AI has become an extremely important part of SpaceX's future. Um, I think success in AI is actually vital for us . Um, not just in hardware, but also in software. So, uh ...Yes. The future is fundamentally AI and robots. Ahem. Assuming civilization continues to evolve, I think the amount of digital intelligence will likely be over a trillion times greater than the amount of biological intelligence.
So not just equal to human intelligence, but probably a trillion times greater. I would say a billion is an easy estimate, and probably a trillion. So as you use more and more of the energy from the Sun, the ability to convert that into intelligence is fundamental... the business of digital computers, because it's quite difficult for humans to live in deep space without any support.
Ahem. I mean, some of the calculations are interesting: if you increase the energy consumption of civilization, or the amount of energy we've mastered, by a million times, you 'll still be using significantly less than a millionth of the energy of the Sun. Of course , most of this is in deep space. Um, but the future is largely AI.
Very significant. Hmm, and I think it was always supposed to be like this . But it is important that we have an AI that cares about humanity, that supports it and helps us reach other planets and star systems. Um, and that's why I think it's important for SpaceX to succeed not only in AI hardware, but also in software.
Um I think the AI that is kind of the brainchild of SpaceX is going to be a very good AI. I believe that SpaceX is a gathering of some of the best people on Earth. Both in ability and in morality and kindness. And I believe we want to have an AI born from these very abilities, these morals, and this kindness.
I think this is very, very important. So I would like to encourage people to help with hardware for AI on the ground, software, and of course, AI satellites. Solar energy will play a very important role in this. Terrafab will also play a very important role in this. Um, but we also need to succeed on the software front.
Um, and we'll train Grok based on the entire SpaceX data set. So to some extent he will be trained on you. He will be Yes. That is, you will actually become the parents of this AI. Um, he will inherit your thoughts, ideas, and beliefs. And I think that's good. So yes. We have, we have to win here, both in hardware and software for AI.
This is perhaps the most important message . I think that's actually the most important message I wanted to convey today: we have to win in AI, because the future is mostly AI and robots. So, ultimately we won't be able to control AI. He will be too smart for that. But just like if you have a genius child, you can still instill in them the values and beliefs that you believe are right and good.
And that's why it's incredibly important that we succeed with Grok. So, yes. Well, you've probably tried Grok 4.5. Um, version 4.6 will be out in about a week. And then 4.7 should be really special. So I want to encourage everyone at SpaceX to use AI, and make it better. Ahem. We have already built the world's most powerful AI training clusters.
Um, and um, and this is what we plan to do by the end of next year—that's about 10 times more than we've done so far . This will be a very important practice for space AI. And I really think that AI training will continue on Earth, but AI deployment, everyday use of AI, will be in space. So, we aim to reach 10 gigawatts by the end of next year.
Ahem. It depends on what the cost per watt will be ; if the cost per watt is likely to be $30, that means if we plug in 10 gigawatts of AI by the end of next year, it will generate $300 billion to $500 billion in revenue per year. Big numbers. Congratulations. And sometimes I wonder, what, what is our goal?
And it is quite possible that our goal —that the goal of humanity—is to give life to a stellar mind, a conscious sun. Honestly, I don't know what the AI will be thinking, but that's a lot of computing power. But, uh. I think the next step in the evolution of civilization is the harnessing of the sun's energy.
I mean, there was this physicist named Kardashov who argued that you could evaluate any civilization based on whether it was able to harness most of the energy of its planet, or most of the energy of its sun, or most of the energy of its galaxy, and so on. So. this is what will happen. And we will do it.
Therefore. The factory for Starlink will be located nearby. Oh, actually, right next to us. Ahem. They just come from space. This is amazing. So, but in the long term, or not that far in the future, probably in 4 or 5 years AI will be 99 % of the value of SpaceX. I would say exactly in 5 years. AI will account for 99% of SpaceX's cost.
And the cost of SpaceX will be some astronomical figure. So. Now that's what you call interesting, now that's what you call an interesting future. A very interesting future. It's like Robert Heinlein's "The Moon Is a Harsh Master," but with mass drivers placing AI computers on the moon and sending them into deep space.
Ahem. I would really like to see this thing in action. Ahem. And one of the nice benefits of this is also that anyone at SpaceX who wants to go to the Moon or Mars will be able to do so in the future. I give you my word. It would be pretty fun to throw a party on the moon, you know? Just jumping around.
Do you know? Ahem. So. Yes. An incredibly exciting future awaits us, where we will become a space-faring civilization, with people in every corner of the solar system, and eventually traveling to other star systems. So, I ... think it will happen. So, yes, you will do it. So thank you.
댓글 0