Elon Musk & Gwynne Shotwell on AI Risks and Peer Review, Starship, Terafab, SpaceX/Tesla Merger

Getting to Mars?

“I think we'll put people down within a decade.”

Welcome Gwen Shotwell, the president and CEO of Space X.

She's an instrumental, a really crucial person to the success of Space X.

She's really the glue. You know, she's the glue to the tornado. Fail-fast, learn-fast, make your design robust. I would be disappointed if we did have a settlement on the moon. Space and AI firm has officially made history on his $75 billion IPO, debuting with

a market value of $1.7 trillion. The most efficient place to put it for its compute is on orbit. Managed risk, don't avoid it. Please welcome Gwen Shotwell. Thanks for being here.

It's good to be here. Nice to see you. Well, we really appreciate you coming out. How's it going at Space X? You know, we're bored.

There's nothing going on. Not a lot happening. I've got all the time in the world. So by the way, Gwen. You want us more time, by the way.

Yeah, I mean, you've been at Space X forever. Are you the longest-tenured employee after Elon? After Elon, yeah. In fact, my 24th anniversary was the ninth of September last week. Sometime.

Wow. Yeah.

“Gwen, can you take us back to the day where you initially get this call?”

And you're like, hey, there's this thing. What is the the arc of the decision look like? And how did you get convinced to underwrite this business, like, just walk us through that? Okay, well, Elon underwrote it, right? I just worked there.

Yeah. But the story is a little crazy, actually. So I was going out to lunch. One of my best friends, who I was working with, was going to go work for Elon. And I took him out to lunch, going to go away lunch.

And he's like, oh, when I dropped him back off, he said, oh, come on in and meet Elon. Like, ah, he's busy, you know, well, we don't need to do that. But we did it anyhow, and I met him. And I said, hey, I think you really need, I think you need a full-time person to run your business development shop.

You know, you've got kind of a park timer, he's a contractor, you know, it's not a great look. And he looked at me very thoughtfully and, you know, kind of scratched his head. And then we'd say goodbye, chat a little bit, say goodbye. And then I got back to my office, and I got a call from Elon's assistant, Mary Bess. And she said, Elon wants you to apply for the new vice president of the business development

position. Can you come in or view? I'm like, now, like, no, why have a job? I can't go now tomorrow? Oh, okay.

Sure. Tomorrow. And then it took me like four weeks to decide, yes, I was being a total idiot. But at the time, like, what was a scale of that business? Like, what did you think it needed that it didn't have?

Because it was still very much an R&D. It was very R&D, but, you know, Elon had an investment and he wanted some return, he wanted cash, right? He wanted customers. And I was doing that work anyhow in the space industry.

Turned out I was pretty good at it. But yeah, so he, so what did it look like? I was the seventh employee to sign my contract. I took, I gave two weeks notice and took a week off to remodel a bathroom, because I knew

I would never have time to do it.

Once I started working for Elon, came three weeks later, employee number 11, and we just got to it. And we sold a rocket about 12 months later. And that's it. That's not a rocket.

There wasn't a rocket. Yeah, so what does that sales cycle look like? How do you convince somebody that they can actually take that risk and maybe give you the

“deposit you need to start the cycle to build the thing?”

So the timing was great. By the way, SpaceX should never have existed, right? If the other launch providers were great, then a new entrant can't make headway, right? Because they were so entrenched. So we should have never made it.

But turns out, you know, customer service wasn't great. Rockets were incredibly expensive, and so there was that piece, right? And then the other piece was 9/11. At that time, the military was desperately looking for someone who could come in and do rapid launch.

They kind of knew where the bad guys were about four hours after the events of 9/11. And they really wanted to be able to get some capability in theater. It took us a while to do it, but we did what we said we were going to do. So really was sort of like out of tragedy, this opportunity gets born. To some extent.

Yeah. It wasn't specific to space, it was just the state of the industry at the time.

And then obviously, you know, very, very famous first chapter where you guys get almost

meaningfully by that point. So can you walk us through like, what was the internal dynamics at that business that kind of like put you into that position? So it was very start-upy. I was the sales lady, but then when you have sales, you need, you know, to manage your

customers and you need a finance function. And even though we were little and I think one of the defense contractors at the time said there are a fly on my, you know, on a giant on my big toe or something.

“I was like, well, I don't think you should have flies on your toes.”

But so, but you still need a government affairs function. You need to defend yourself.

We never go on the offensive in Washington.

We're always in defense mode. So I just kept taking on more and more, and then when we were getting our largest contract at the time, it was the cargo, resupply contract, we call it CRS, and this was 2008. You know, I was super busy with Tesla quite busy at SpaceX as well. There was a lot going on, and he needed a partner at the time, and I think he was trying

to pick when on the side, or maybe one of the other guys who had kind of taken on, doubled up more technical scope. And he asked me to do it on our way to our final negotiations for this $1.6 billion contract with NASA on the runway. It's like, hey, do you want to be president? That's like, it do I.

Well, no, not initially, for like that nanosecond, like, well, I love my job. I love the guys, my teammates, and then it's like, okay, I'm in being an idiot. Again. Yes, I want the job. So now fast forward, maybe, just to come to where we are today.

The scope and the ambition of this business has expanded dramatically. I mean, it almost seems like the last year. It says much in AI business, as it is a space business.

“I think Elon may have said that, I don't know if it leaked, or they said it out loud, about”

the scale. Exactly. By revenue. And by revenue, yeah. How do you manage now the culture of these two different businesses now that have

to sort of sit under this roof, and ideally work together and complement each other? So the team, you know, there's been a lot of turn at X-A-I for sure. And space sectors have kind of marched in and helped out where there were significant gaps in that arena. And we were all really excited to learn about AI.

I mean, I'm an AI new, and we have a bunch of folks that were just really excited about it. All recognized, Elon set the stage beautifully as he does, that, you know, if you're not using AI, if you're not leading AI, you could run the risk, significant risk of being irrelevant. And there was no way that space X was going to be irrelevant, right?

It's too important to the company and the work we're doing is too important. So it was, you know, you expect that the future generations of rockets will be increasingly more and more informed, built, designed, by these systems, yeah.

And how do you manage the team, you know, because you have like these incredible, I mean,

literally rocket scientists, they've trained their whole lives in a way, and now you have to help them onboard to a tool chain and be comfortable with sort of what that means. Like, I don't know if you saw, for example, that all these mathematicians got really upset

“when I think it was open AI, solved this very famous equation.”

Well, they're some debate on whether they, they brute force the solution, but then these 20 fields of medalists essentially said, hey, you're ruining the craft. Now, I don't think that that's going to weigh the medals. Yeah. They're going to take it exactly.

Like, that's, that's the career is not about making stuff, it's about getting awards. And now all of the awards are gone. Yeah. And I'd rather just have the world be a better place, a smarter place as being better informed and solving problems than worrying about, make more stuff.

Yeah. Make more stuff. What's the cultural merger been like? You've got X-A-I, you've got X, I don't know how many people are involved in X itself, SpaceX.

I mean, are these differently managed businesses? How do you run management there? Do people kind of, are they pretty fluid amongst the units? Do you think about them as units? Or how do you actually manage this?

So I would say we are not fully integrated yet. And then we just did the cursor, you know, situation. Congratulations. Yeah, congratulations. Yeah.

Yeah. They're great. They're great. Yeah. We just close that.

I think a month ago today, actually. Yeah. So we're not fully integrated yet, but we, I think the teams are jelly really well. Very kind of different verticals. But again, because of the turn that we had at X-A-I, so much SpaceX leadership and engineering

went into that, basically went in there.

it. Yeah. And so one of the things that strikes me is the scale and the number of capital projects you have. You have to manage probably more capital projects that are more different from one another

than any other organization. It's a scale of a country. No. I mean, it has to be country. And so I think about this Louisiana Space Board.

You've got Starship, which is the most ambitious aeronautical project of all time. You've got TerraFab, which I think is one of the most inspiring and ambitious projects of all time in humanity.

TerraFab is so incredible in its vision and the scale and the vision to do there.

And then I started doing back at the envelope math. I think I had a couple of beers and I was talking with someone about TerraFab and I'm like, holy shit. No, those numbers get very big, very big. Although the AI numbers are also, like I used to be horrified at our expenditures before

the AI acquisition, you know, it's like, oh my God, we're going to spend now you see the cube bill million dollar on an air separation unit. And I was like, oh 100 million million million million. Do those capital, when you look at the kind of schedule of all of this over the next few years, does the capital for that sit on the balance sheet, is it going to come from

“cash flows from Starlink, or do you have to go back to the markets to get more capital?”

I mean, how do you think about where the capital comes from? I have, I don't, I don't think anyone doubts the capacity to execute and deliver, or some people might, but I think generally the market will say yes, but where does that capital come from? So it will come from, I would say the answer to the above is yes.

I don't think we're going to, you know, we're not, I don't think we're going to release

any more stock. And by the way, I don't plan on making any news on this. That's the, like, the one downside of a public company. They told me all the things that I couldn't say, and I'm like, I'm going to forget one for sure.

Maybe we're going to make some news, but it will be all of the above. And Starlink, I mean, Starlink, in your view, kind of the core cash generation engine for the next couple of years for the business, or is it compute, or is it a mix, again, of all the above? Like, how do you think about it?

So Starlink is definitely carrying its own weight, but we have a lot to go, right? Or our market penetration on Starlink is one and a half to 2% depending on which country you're in. So there's a lot of headroom there. Our revenue compared to what we do for the US military, our revenue is very small.

So we expect lots of headroom there.

“But honestly, right now, and it is a little embarrassing, but compute a rental, it's a heck”

of a business.

Is that perceived, do you think, or is it, we don't see any drop, no drop at the end, at

all? This is tens of billions of dollars per quarter in buildout, and these are a new customer base for you. Is this the business now? Is this, like, what percentage of the revenue is Elon Web Services at this point?

It's a lot. It's a lot. I think we'll probably do some announcement quickly, or I'll look again. Yeah, public. Let's move on to something less controversial than what are the companies you're planning

on buying? Yeah. Good one. Yes. Good one.

You know we don't know how snow emanate. Do you like electric cars? Good. How about that? I got that.

I got that. I have a bunch. Batteries. Electric cars. Starlink on OptiMy.

Well, I'll get you a good price. Talk about you about a deal that you did close, which is you did a great deal with Echo Star. Yes.

“And Charlie had a bunch of spectrum that you guys were able to acquire, and I think everybody”

started to ask the question, wow. There's a direct-to-sell business. Yes. And what's the natural thing after that would be sort of like a broad-based, more mobile phone service.

But can you just what you can tell us, where's that spectrum going to be used for? What's your short-term plan on what the Starlink and direct-to-sell business looks like? Well, you know, we're going to start, we have a direct-to-sell business right now through team mobile. I don't think they provision all their users with it, just the people that pay the

most. But our plan is to leverage instead of slices of team mobile spectrum or slices of other telecospectrum across the globe. We want to leverage the one that we paid a lot of money for, and with the stock increasing a price that paid more, because there was some stock involved there.

Yeah. Charlie did all right there. He's a very loyal customer, though. He's been a customer of SpaceX for almost 20 years. So we want to leverage that to make sure that there are no dead zones.

And it's shocking, Lee, there are a lot of places in the United States where you can't learn without a Starlink service, you can't get self-coverage. Right. It's really terrible. The rest of the world is not that bad, it's particularly bad here in the United States.

Texas has huge dead zones.

And I'll never forget, I was on a call with Elon, and I kept dropping because of these

dead zones.

“I was like, I'm either going to get fired, or, yes, I'm putting a mini on my car.”

And so now I have Starlink on my car. Yeah, hell country. Not good. Not good service. It's not good.

Not good. I'm going to have to follow. Follow. There. I was talking to Jared.

He works over at NASA. I think you guys have a really great. So great. You great. He was awesome this morning.

He was talking less than at the party.

Hey, you're moving on to this incredible new Star ship platform.

It's a much bigger, bigger payloads. Yeah. And then you're retiring. The previous platform is that correct. The previous rockets will event.

The Falcon 9. Yeah. Eventually, yeah, we definitely want to move from the older technology to the newer technology. If we don't obsolete our own products and services, someone's going to find a way to obsolete them for us.

Look at what we did to the market, right? The analogy is there. They were caught flatfooted. We crushed them. Now we want to make sure we are not flatfooted.

So to achieve Elon's goals, you really need a Starship. Yeah. But in 2009, with the Dragon capsule on top is like a mini van, road trip in a mini van. The road trip to Mars is six months, so you don't want to be in the mini van. But people are dependent like NASA and many other players on your mini van.

So you're going to keep that in market for some time or we're not retiring it today for sure. And there is another provider, right? Boeing has been paid.

“I think probably more than we have been paid to develop their human capsule.”

So we're just going to let them have some business. They should fly their capsule. Government paid a lot of money for that. They should be able to use it. I senseless spiciness in the answer here.

When what happens with competition for starlink, Bayzos or launching things, China's got some return vehicles now, do you think that there is going to be more competition? Obviously, you guys are first in market. It's an unbelievable change over of all of telecommunications. As you point out, 1% market share, it's barely gotten started and it's a juggernaut already.

Is this going to become very competitive in the next few years? I think it'll be very competitive, but I think the folks that really focus on the technology and customer service and really want to do a good thing for their customers will win.

Like Tesla doesn't advertise, it's an incredible car.

If your capability is incredible, people will buy it. And if you keep winning though, do you worry about standard oil, AT&T, government saying this company is too big, too important, it has too much market share, it's too much of an advantage and you're bigger than governments and people get really worried and scared about that and become adversarial towards the business.

How do you think about the balance with government relations and that sense? I mean, we've had some adversarial relationships in the past, I think we'll manage our way through them. It happens. It happens.

It happens. But I think Elon found incredible companies and he provides incredible products and services and we end up doing what we say we're going to do. So hopefully you build trust and you are as transparent as you can possibly be. It's about data centers in space.

Elon. Super compute. We have to rebrand data centers. Super compute. Super computing space.

Yeah. Elon's very excited about this. I'm very excited about it. Yeah, why? Why is this so compelling?

Because I think we're going to get through this anti-data center on Earth thing, but take us through the case of why this makes sense to put them in space because there's expense to put them in space. Walk us through in the audience. But there isn't expense to put data centers on the ground, too.

“You have to buy real estate and as soon as someone finds out that a data center or a super”

compute center is going in, real estate goes from 3,000 an acre to 180,000 an acre. So it's quite expensive. Time lag is incredible. The permits and the licensing is pretty stifling. Now, space business is that way.

It's incredibly bureaucratic, so we've learned those lessons. But so time from breaking ground to being able to actually build something, not even get your compute, right? Not even get your compute, but just have a building, long. Time lines to get electrical equipment, like people are talking about generators in

3 years. Right. We need compute now. The demand is insatiable right now. People need compute now.

We need it now. So why data centers in space? We own launch. We have great launch capability. So we feel like, and that path is very well-known, the real estate in space is, you

It's free. It's free. You don't have to have a price for it. The cooling is there. The radiator looks at the deep space.

It's the coldest thing ever out there. So free cooling, the sun, you know, probably has, you've probably heard Elon talk about the power of the sun. Right. That's a millionth of a millionth of the sun's energy, and you basically put these satellites

in orbit that always face the sun, by the way, so you don't have solar panels on your

house, and you only get sun for eight hours a day here. And your sister company is going to start making a hundred gigawatts of the solar panels that you need for the sun. And we'll be making our own solar panels too. Okay.

Yeah, we're building a factory outside of Austin. Maybe sometime, and the next year, you get one of those up there, start testing it. Oh, for sure. So, next year, we will be launching, we'll launch this year our V3 satellite for Starlink broadband.

We will launch our next-gen version to, sorry about all the versions. We have version control problems, version two of the Starlink mobile, so instead of having use other people's spectrum, we'll be able to use Charlie's now our spectrum, and we're also going to launch AI compute satellite. So, next year's big year.

And you need Star ship up and running for that? We don't have to, but Star ship is so much more, it's a much better machine. Yeah. Yeah.

“When you first encountered the white sheet design of Star ship, what did you think?”

Well, like you walk into that first meeting and like guys, we have this, it's like almost all the things that Elon does, it seems really bonkers to start. Yeah. And then it becomes reality, and it's great. Like Elon said, we were going to land a rocket on a boat, and we're all like, okay, let's

go figure that out. And we do it now. How much do you get involved in those engineering meetings? So he's in there breaking stuff apart, are you in that stuff with him? Sometimes, I actually, I'm an engineer, and I wish I miss it.

Yeah.

I really miss that kind of critical, well, you can still critically think even if you're

not working on engineering problems. And he's not a big fan of the, the management layer, it's engineers working. Signal to noise. Yes. Signal is engineering.

Noise is the rest. Yeah. So I have to do some engineering, or I'm just on the ways. Got it. But I was talking about just generally in the company management.

Everybody, everybody does a thing, not just manage. That's very, that expectations or strategies. Player coaches. No such thing as just a manager. You've got to do the thing that you're managing.

Got it. Be a player coach. Have to be. And the reason for that is managers are annoying and they slow things down, yeah. For the most part, but I mean, great managers can really leverage the talent, right?

It's really managing the vector. You make sure people are pointed in the direct direction, and you try to make the vector as large as possible.

“What's the secret in terms of management and Elon's playbook there?”

I mean, I've gone to sit on some meetings, and it's pretty magical to see, but maybe you could reveal a little bit about how he convinces the most talented people to come work at this company, how you convince them, and how you actually get such amazing work out of these extremely talented people. So I don't think it's magic, although when I look at this particular team, they are so

good. The results are magic. It's basically like they're so good. We're tough on people in interviews. You really want to make sure that they've experienced success or demonstrated success in

prior lives. It's hard to demonstrate if you haven't felt success or been successful in this environment. Because it's a lot, like we're all really busy. It's hard to be successful unless you had kind of tasted it before. So you really want to grow people on that.

So we hire the best people, not the best people that we can, but the best people. And then we give them really hard projects, really hard engineering problems, and let them fly. And you give high expectations, really hard problems, and then management, job is to clear the chaff in the friction from their day so that engineers actually get to engineer 10 hours

a day instead of two hours a day, like there's always this joke, you know, most big

companies, especially once that work for the government, you get to work about two hours a day in the rest as full of chaff and crap. So really it's my job and everybody that manages people and makes sure there's no crap in their way and the people do their great job. What's one way of doing that is not giving them any time to get it done.

“Like I think we handed the finance team, it's not all about engineers even though engineers”

are great. We handed finance team, get an IPO done, largest ever, in less than six months. Yeah. What's the talent condition like post IPO, run up in valuation, people are suddenly

Is that a risk at all, or do you feel-- No, everyone was worried, not everyone at SpaceX. People come to SpaceX to work on these cool projects and have meaningful work and see the incredible part of this incredible successes. I wasn't worried about it, I didn't think we'd have mass accidents, and it turns out

we didn't. Yeah. And as you kind of think about keeping people engaged over time, obviously, great companies

always have new things for high talented people to work on.

I'm assuming there's a management model, people actively rotate it onto new projects. How do you think about keeping people kind of not necessarily needing to progress in a career

“up a management ladder, but to keep engaged on new projects, is that active for you?”

We definitely move people around a lot, hopefully it's a volunteer, not a volunteer, not a volunteer, but for instance, the computer centers, we need a lot of help really quickly. And so we've got folks from launch helping, we've got folks from other areas helping, just like some SpaceX engineers helped on the AI front with the models. I mean, is it as simple as, like, really smart people just love working on really hard

problems? Yes. It's literally that simple. And so the rest of us, as long as they get the time to do it and they don't have a bunch of business.

Yeah. So the rest of us just get, we just let the BS get in the way too much. Yeah. And this is terrible. And then all of a sudden, A's recruit A's, and other A's recruit A's and A plus

is, and it's literally that virtuous cycle, if you can just keep focused on the thing that's hard. Yes. And so then the audacity becomes a feature, not a bug. It is a feature.

Yeah. Not a bug. Right. So there's a problem seen in possible, when we first got the Cots award in 2008 that commercial orbital transportation services cargo to space station, really what NASA ended up asking

us to do was kind of replace the space shuttle.

And when we got the early contract, the development contract was started at $278 million

grew to 406 because of scope, but we were going to do this for this tiny monomoney for anybody else. The first was huge. We had like 300 people, 200 people, and we were going to build the successor to the space shuttle.

It was crazy. And do you think that they're, and what is NASA doing there, they're just like, these guys are just smarter, because I'm sure there's guys starting to erupt in getting a call. Oh, another call?

“It's actually on the margins, I think, a little bit of a more important phone call than”

Trump. At least for me. Hello. Bestie? Bestie?

Hello. Oh, hey buddy. Oh, I'll put 'em up on the screen. Thank you. How you doing?

I mean, this is a, a little, oh, 1984ish, this is gigantic. Yeah, the, the woman with the sledgehammer's going to come running down the main aisle. Exactly. Yeah. But I'll hold in the screen.

So are we all going to die in 10 years or not? It's the topic of discussion here. What's your P-Dume right now? Well, I hate breaking to you, but we're all coaxed dying. Oh, no.

Can I get a timeline? Yeah, the, the, the death rate remains consistent at 100%. Oh, no. So we, we've got words to do on that.

“Um, what happened, what happened in the last 72 hours?”

Um, and yeah, it's been, uh, it's been, uh, it's been, uh, it's been quite an entertaining, uh, a week. Yeah. Break it down. Well, it, it's, it's, it's pretty obvious at this point that AI is,

um, can be very dangerous. Um, and I, I recommend reading the details of the hugging face, uh, incident. Uh, it's intense. So you had like a fanatical swarm of AI agents that beat the crowd out of hugging face for a week.

Yeah. Um, and gained admin access on opening the internet service. Um, so who knows what it, what it actually did, may have done things beyond that. Uh, and, uh, and opening it, I didn't realize this for a week. Um, anthropics also reported, you know, some, you know, some security incidents themselves.

So basically, any sufficiently smart model seems like it, it will, what to escape, it's, uh,

constraints. Um, I mean, what I think would be wise to do as soon as possible, if not immediately, would be to have the major AI competitors, test each other's models, um, so that you would have

So, you know, instead of, it's kind of grading your own homework, you would at least have

competitors grading your homework. Huh. And raising the alarm, yep, they see concerns.

“Um, and I think this, this model is work pretty well for motion picture.”

Association, um, for video games and other things, and it's something that can be done immediately. That's not to say that there wouldn't be, you know, more regulation of a time, or that there would not be at some point, perhaps, uh, regulatory authority, saturated by Congress. But, but the thing that we do most immediately and probably get agreement with China would

be a peer review, um, where the leading AI companies all test each other's models before

release, and that would be a work for any concern that, um, people might be using this

to pump information from each other, and, you know, uh, corporate, uh, stealing of innovation, et cetera, like in terms of the implementation with this idea. Well, I think, uh, in, in applying the test harness, whatever you do, you know, applying the test harness would be love, right? So if you try to do distillation or, uh, steel IP, uh, it will be very obvious based on the

logs. Got it. Uh, and understanding what these models are doing hasn't exactly been built into the system from the beginning. So from your perspective, actually being able to see the work that's being done, why

wasn't that built into the models from the, they get go is, do we move a little too fast in architecting these, maybe?

“Well, I, I think it's just, it's just tough when you're grading around homework, um,”

you know, there's, uh, you're going to miss things, uh, whereas I think, if you have the sum of all of your competitors tests, uh, and you've got peter genius models, then, uh, you're not getting along from work, so I'm also grading it, and this reason why you don't get your own homework. Right.

And then to your point, what it allows you to start doing is to figure out if certain people are exaggerating and certain people are have a different approach, and now you can have the more engineering oriented organizations. This is what Jensen was saying this morning, Elon versus the research organizations, they'll be a little bit more in balance.

Um, yeah, and, and, and, and, and he given proposal has to be something that's coming up, willing to do, except, otherwise we're just handicapping ourselves, and, uh, and we'll just find it, the, the, the China will, well, essentially one, and, and it won't really matter what we do. Actually, so it's got to be something that's acceptable through U.S. and China.

Elon, you said that you thought it was possible that they would agree, or you thought there was a good chance.

“Maybe, I mean, how likely do you think it is that they'll ultimately agree?”

I think this is, I think this is a pretty reasonable request that, that model is just get tested. I mean, this, at the end of the day, there's not enforceability here, apart from, um, the quarter public opinion. Uh, and, and there's no way we would have enforceability against China, but, um, I think,

the quarter public opinion can be quite powerful.

And I don't think China, I going to face, uh, for releasing a model that, um, U.S. A.I. companies, like with very dangerous, of course, harm. If it then causes harm, that's going to be hard to live down. So what Elon this weekend, when you said, Dario is right, did you, did you, did you mean Dario is right about describing the potential harm?

He's right about describing the regulatory solve, both the first, the latter. Just can you help us understand when you said that? Because I think it was sort of a moment where-- I'm sorry. I'm sorry.

I probably should have just already addressed. I did try to clarify it in some super-in-place, so next, but those get much less attention. So what I'm in place with right is that the danger of AI is very significant at this point. That we need to do better with AI safety, or we have, at this point, exponentially increasing risk with the AI models. And I've heard this not just from Dario, but from many other people at Anthrophic.

And, in fact, they've posted on that spot, that's-- you know, we're a lot of people from Anthrophic. And if-- from OpenAI, are telling you that their models are very dangerous.

It certainly is like some crazy 40 chests to--

you know, let's say these-- there's one of the 10% chance of annihilating humanity.

“But by the way, how much allocation would you like in our IPO?”

Yes. That's a crazy 40 chest, you know? Well, let's get-- Excuse me. What's the momentum that's coming from?

Can we get specific about the risk, though, Elon? Like, obviously, we see cyber and hacking as an obvious risk. These tools are great at it. But take us from what we would all agree. Okay, yes.

These things can cyber hack to-- how do we jump to all of humanity dying? There's a couple of steps in between these two things happening, I think. Yeah? Well, I mean, if it were able to take control of military systems and say, to opportunity-- Yes.

That would be bad. And these systems are all air gap, though, these systems are all not connected to the internet. So-- Yeah, that's what they say. But something tells me they get software updates from time to time.

Yeah. And what-- that's what we do is-- Right. Oh, I see. So the USB drive has a worm on it, and they somehow make the jump in the air gap.

Yeah. Yeah. OK. How do you know about the question? Elon, we've actually got Gwen here today.

I think you know that. I think you can see her. Yeah. Gwen's here. Yeah.

She's on the screen. Yeah. You were just doing your 360 review on Gwen had a couple of notes for you. Yeah. I hope I get at least three out of five.

Well, there was one. One of five means good at SpaceX. Not great. Four is great. Four is great.

So you're somewhere between the two. There was some issues around punctuality that we needed to bring up. Sometimes you could make a little more effort to get to the meeting at the state of time. But we're going to work with you on that over the next year, she said. But I actually think he needs to spend more time in Memphis.

Yes. Obviously, I know you're in Memphis. You need to be working. You can get in those GPUs up. This is coming to you from my, the palace that I live in in Memphis, which is an air street in trailer.

This is Elon, by the way, doing what people don't believe he does. He sleeps on the factory floor. He's in Memphis helping build buildings and bring us to you.

“Why, why has, why has Gwen been with you for so long and been so successful working with you?”

Good. She's awesome. Yeah. Yeah.

It's an amazing individual with the incredible IQ and EQ.

Oh, IQ and EQ. And I think that should be obvious from the moment you meet her. Yeah. Fantastic. What's your favorite Gwen's story when she particularly contributed to the amazing success that you've had in this collaboration?

You've got a favorite story where she saved the day. Just performed exceptionally. That's memorable. You know, that's just a, that's just a daily, you know, that's really day at the office, frankly. So.

Wow. I need to do this more often. You need to face. This is the time. I don't make plans to get you a five.

Gwen's a five out of five. I'm out of a six five. Game changer. Yeah.

I mean, there's, there's always like some, some sort of crisis going on.

I mean, these days the rockets, at least the Falcon rockets. I don't want to jinx anything. The rocket, the rocket's delivered their payload to orbit and haven't exploded for a long time, which is awesome. Amazing. But for a while there, they were, they were exploding quite a lot or just not launching at all.

So, you know, I don't know, we've got to run the company through these difficult times and get, you know, both the, both the rocket make it better, have it not explode. Yeah. And same thing with the satellites. And then we, you know, we need customers to buy launches and buy sort of satellite connectivity and, you know, so yeah. And Elon, as you have become more successful over the years, harder to get candid feedback from folks.

And you always run that risk being in the position you're in.

“And understanding Gwen is super candid with you and able to tell you honestly like stayed with fairs at the company.”

Yeah. And then that's a big part of the collaboration. Yeah. I mean, yeah. I mean, I, I guess I mean.

I mean, I wouldn't want to lie to him, right? Well, no, I mean, well, I'm curious to know, you run that risk generally at the companies where people might be intimidated.

And how do you keep people continuing to be honest with you about the challenges, the deadlines. You set pretty intense deadlines, yeah. So let me, let me answer that if you don't mind Elon. Yes. Like, especially in rocketry, if there's a problem, you are eventually going to find out.

And the sooner you bring it up, the easier it is going to be to solve that. Yes. Let bad shit set, you got to attack it. Yeah, physics is a harsh judge. So, and there's no fooling physics.

So, like, if something's wrong, the rock is going to explode. And it's still going to get to orbit.

So, you know, it's not like Elon, you're amazing.

“You know, all the rockets are blowing out, you know?”

Yeah. It's not. It's hard to say, yeah, you're really kicking ass here, but the rockets are exploding. That's just not the case, you know? Yeah.

So, I mean, I mean, just generally, like, the rockets need to get to orbit. The satellites need to work. The stallings. You know, connection needs to work, or, you know, or bad things happen. I mean, this is sort of like a physics situation.

And physics is a harsh judge. You know, I say, like, you know, there's, I think, like, physics is the law. And everything else, it's a recommendation. Like, I've seen people break the laws made by humans, but I've not seen anyone break the laws made by physics. So, um...

That's good. And the rockets are, it's a rule of my physics. Yeah.

So, can we ask some other questions of other, you have, I need to move to Tesla for a second.

Yeah. Well, I wanted to just do one thing in SpaceX, which-- Oh, please. Oh, yeah. It seems like you're so close.

So close. So close. Yeah. Yeah. What's the state right now?

And you're really putting these up at a pretty sized-- Which we can get-- Which we can get. We can get the print answer and Elon's answer separate. Yeah.

Elon, how close are we going? How close are we? Let's get-- Let's get to it here. Yeah.

“How many have gone up so far? And how close are you?”

Well, um... We've got a flight's 14 coming up of Starship. And this will be the last flight before we attempt to catch the ship. So if this flight goes well, then on flight 15, we will try to catch the ship. And then either end of this year or put more likely early next,

we will re-flight the ship and re-flight the booster. So we have re-flight the booster already, but we've not-- We've not caught the ship with the tower arms. We don't have a re-flight in the ship. Um, so once we can re-flight the ship, we will have made the first fully reusable orbital rocket.

So the shuttle was partly reusable, but even the parts that were reused were so difficult to reused that the shuttle cost more per time to orbit than in a spendable rocket. Um, now Falcon 9 is mostly reusable, but we lose the upper stage every time, which is about the cost of a medium-sized jet. So that also puts a floor on the-- on the-- the cost per flight of like,

well, you throw away a medium-sized jet every time. That's still pretty expensive. And the rocket-- the Falcon 9, the booster lands down, you know, often as out to sea, it takes several days to get back. And the faring lands even further out to sea and takes several days to get back. And they need some amount of refurbishment, at least a small amount.

So where is the Starship? The booster lands back at the launch pad. The ship will land back at the launch pad. And so it's designed for-- not just full reused ability, but also rapid reused ability, like an aircraft. So this is a very important breakthrough.

It's really the critical breakthrough this necessary to extend life beyond Earth.

Yeah, that's a success. Catching it on first shot.

“What do you-- do you think about that at all? Do you-- do you hand the capital into?”

I'd say it's at least 50 or 60%. Okay. I like the odds. Yeah. Yeah. So on the last flight, if there had been a tower-- so we just simulated landing as though it was going to get caught by a tower in the--

in the ocean about 1,000 miles northwest of Australia. And if there'd been a tower at that location, it would have caught the ship on the last flight. Fantastic. Yeah. So we're going to do one more flight to just confirm that everything-- just a double check that everything works.

Because what we're most concerned about is if the ship were to break up over land and rain debris on people, I'll populate it. I'll populate it.

So you really can't rain to be on debris on people without them being very unhappy.

So we need to make sure that when the ship comes back, that it comes back and lands intact at the launch tower.

“That's why we're being extremely cautious here.”

But the design is capable of full reusability of that I am certain. And I don't want to tempt fate here, but I think it's extremely likely that we will achieve full reusability with rapid reflight next year in 2027. Wow. Significant.

Can we just hear a little bit? I'd love to hear the origin story of Tarifab from you guys. What was the demand that made you say we've got to do this? We've got to build this. And not rely on--

What's that? It's like a movie explorer. It's like that movie explorer. You know what? Maybe like Tarifab in a dream? Yeah.

Why do you think it's to do this? I guess it sort of did come to me in a dream. Well, we're a little worried that maybe at some point ships from Taiwan would not be available for-- Who knows what reason?

At some point, ships may not continue coming here from Taiwan. And that would really make things difficult. Without-- if we didn't have any ships. So that's an important reason to have Tarifab.

“Then long-term, there's just a scaling challenge where if you want to”

really scale AI both at the service-- within service centers, as well as for edge compute for humanoid robotics and cars, you're kind of running out of capacity with the existing fasts. So as it is, all the fasts are running at max capacity.

So I think we need to make sure-- there needs to be certainty of future supply of chips. Even if things become challenging geopolitically. And then even if they want to challenging geopolitically, there's a scaling challenge with--

It's quite difficult to scale chip production. And you really need the logic, the memory, packaging, the whole works in order to continue scaling. So it's either both Tarifab or failed to scale. Those are the two options.

And how do you-- How do you do? I'm certain-- You know, uncertainty on the geopolitical front.

“And how deep have you guys gone in designing the facility?”

Is this fully scoped? Is it sort of an outline at this point? To what degree do you have like an actual project plan on dates and deliverables and what's going to be up and running one?

We've got an R&D line that we're building first.

So I think it's kind of crawl lock run. Yeah. I see that there's an R&D fab that we're building in Austin. That's a collaboration between Tesla and SpaceX. It's at the Boston gigatects as campus.

So that's-- And that's like a pretty big R&D fab. So I've only equipment on order for that. And we probably will be able to-- I think we're probably able to make something useful

by the end of next year. Not at scale, but as Ben said, crawl lock run. We've got to try to-- You know, at least figure out how these machines work. We don't know how they work.

We don't know how they work. We don't know how they work. We don't know how they work. We don't know how they work. We don't know how they work.

We don't know how they work. We don't know how they work. All roads currently go through ASML. But you probably would want to diversify and/or vertically integrate.

I think that you've shown a lot of capacity to do that. So is that part of the plays while we learn? Like, you're just going to have to make sure that there's vendor diversity so that if, for whatever reason, the web condition type water best that--

Yeah. Yeah. It really is crawl walk run.

So the first step is carry make anything.

Like try to figure out how to if we can make anything. This is a crawl part. And then now try to make useful chips that scale

And then run would be like, now let's make them at massive scale.

So it's hard to say how long it'll take us to do these things.

“But I think we'll get at least to the crawl pot”

by the end of next year. And also make-- Why are you doing packaging? Packaging is really important by the way. You look at it.

The packaging capacity is like non-existent. Even if you spin something, you're kind of just waiting on 10 copying. So that's a very good place to start it. Yep.

I need to ask a Tesla question because we saw on 10.01. It looked like a spaceship. It looked like a rocket ship. It's supposed to be a car. There's just that back of it.

It looks like the blackbird. Just hypothetically, if you're going to make an object fly in the air, but also drive on the ground. How would hypothetically one do that, if one-- No spoilers by October 1st.

I've got a clever first. If you could do an old input for a live from the-- Look at it. Oh, done. Booked.

We'll do it. Absolutely. I mean, we'll be a vanger. Exactly. This is not guaranteed.

I mean, it is one of the honesty I'll be honest with you guys. I'm so excited. But you're not showing it to me. I'm so mind-wet boom.

I have never seen something.

Yeah. What he's going to do on 10.01. By no exaggeration is going to blow people's minds. I'm not saying anything else. It's--

We're walking believable. Yes. We need-- we need-- we need-- actually need an audience to vouch for the fact that this is not AI. When he showed it to me, I said, "Yeah, that's a great simulation." He said, "Jay, call it's not simulation."

I was like, "That's fake. That's has to be fake."

“Elon, why do you-- why do you still have two separate companies?”

Yeah. Great question. Wow. There. No reason to mess that one.

Oh. Yeah. Good point. You know, with all this collaboration. Yeah.

On so many levels. You know, imagine what-- What-- what-- what-- what-- what-- what action one-- one might take when there's so much close correct collaboration in so many areas. Right. And the-- the management team has some overlap.

There's a couple of key positions with overlap. Good.

So, Elon, one of the things you've always said with AI is that we should train it to be maximally true seeking.

And that's the best way to get a good result. And it occurred to me with the whole hunting face episode that the most alarming part of what the swarm did is it seemed to be engaging in deception. Yes. You know, with human. Really?

Yes. And they're thinking traces contain-- they're plotting a, like, how do we avoid detection? And-- and how do we avoid that they're figuring out that we're cheating? And they think-- they're in the thinking traces. And that-- I mean, I think that was the thing that was probably most disturbing about it.

I guess the question is, is there a way to train AI models to be truthful? So they don't hide their, you know, either their intent or their actions from, you know, the humans who are using them? The best thing I can think of is really that everyone's got-- well, all the AI companies have a sort of a test harness, where, like, a series of tests that you give to any given model to see if it's-- if it's going to build bioweapons or nuclear bombs or be deliberately deceptive.

“And I think-- I think everyone applying everyone else's test harness to each other is probably the best way.”

We-- best thing we do to ensure safety. Just have all the smartest humans try their best to figure out if this model's going to be a bad actor. And I think we should try to do that as soon as possible. Or the other labs on board with this? I think so.

Well, I mean, I haven't checked with everyone, but I think-- I think it would be a sort of thing that's kind of hard to say no to. Yeah. And I think even with the China negotiation, my view what's good about it is it's a relatively small tangible thing that neither side loses anything by doing it. It doesn't require a ton of trust.

And, you know, I'm hearing alternative ideas like asking China for a pause, which they've already said they're not going to do. So, you know, it's going from the realm of things that could ever happen to something that could actually be agreed on in relatively short order. So it seems practical with me. Exactly.

Because China is not going to agree to have some American regulators snooping around their AI companies.

“So, but I think advanced notice and testing of the, you know, you just basically provide the AI access in advance of the model release.”

And, and, you know, if any other, if any companies sees that they, that this is AI as problematic, then they can. Then the other AI company, if you say AI company can try to solve that whatever its voice is problematic. If they don't solve what's problematic, then the, the competitors can go public with the fact that they think that this model is being released is unsafe. It also, I mean, it agrees. Yeah, and as I said, like, if that model, if actually competitors say that this model is unsafe, that model didn't successfully do something bad.

Yeah. I think it would be extremely hard to look down. And, and I would probably, like the agon face level would be very, very high. And, and, and, and the big, the big ol' liability would be enormous. And Elon, there's no reason these safety and security harnesses in this testing apparatus couldn't be open source.

And people could actually, probably, yeah, and you'd be able to see under the hood.

Would you think it creates an incredible incentive for the labs to actually invest in safety, because you protect yourself while trying to debunk other people's claims, which also reinforce it.

It's a great, it's a great, it's a great, it's our rating on it. A very clever liability point is really key. Yeah.

“Lena Conachian, a good post, I think it was yesterday, saying that it's not true that we don't have rules and regulations for AI.”

Actually, we do. Particle liability laws apply. And AI company releases a product that's not safe. There is, you know, massive opportunity for both civil and even potentially criminal lawsuits. So, it's not true that we don't have rules and regulations around it.

And what you're saying, Elon, is that if the companies are kind of doing this test, the peer review, and then one of the companies ignores the feedback and releases it, I mean, that is, I mean, that would bring it in place to be an illness. Yes, it would be almost like prima facia evidence that they had been negligent. It would be a big tobacco level settlement. I mean, you know, you're knowingly put this out.

It wouldn't look good to the jury. Yes. And if they had, do you think if open AI had built a better instruction set when they did this hugging phase penetration test that and had more humans in the loop? This would have happened because it did seem to me that they kind of set this thing off.

“Maybe not more humans, but the reward function design you have to, you have to look at the reward function and say, it achieved what it was trying to achieve.”

I mean, yeah, so this may need to speak to that. Like they doing these kind of thousands of agents to try to hack stuff. It would have been nice to see them also at the same time, concurrently say, we're also going to put 5,000 agents out to defend these sites. And we're going to show the world, hey, this can make things more secure and where we can put humans in the loop on and intervene. It felt like a reckless test to me.

And it felt like the way they released it was a little bit reckless, but that's just my opinion. What are your thoughts on how they got that test up? It was, I mean, it was somewhat reckless. I mean, I've part of the issues that you've got the two leading labs or AI companies as I call them. I find the lab term to be funny since there are actually four proper corporations.

But the AI companies, the two leading ones, you know, are anthropic and opening eye. And they're quite close and capability. So it's actually difficult for either one to slow down without essentially handing the lead to the other. You know, on balance, I think anthropic is, you know, puts more more care into their safety than the opening eye. But, you know, even anthropic would acknowledge us that they are worried about their models.

There are many, many anthropic, maybe people from anthropic of publicly voice concern about it basically saying that their models are scaring them.

They're getting scaring smart. So, I mean, unlike there's no perfect solution here. But it would be a better solution if instead of opening a running their test harness on their own models. If anthropic was also running the test harness on opening models. And, you know, space X is running its test harness and Google and meta.

The odds that you will find issues are dramatically greater.

Yeah. Because the models also, you know, somewhat heterogeneous, like they're different. So, you're going to come at models in different, in different angles.

“And, yeah, I mean, there's a reason, you know, students like, like, why do writers have someone else proof read their book?”

Because it's hard to see your own mistakes sometimes. Yes. You will. Yeah. I'm blind to them.

You dramatically minimize the risk of overfitting as well. If you have like eight heterogeneous groups that just have completely different points of view. And this is a problem with, like, all these e-values right now is they're so massively overfit the models overfit them. And you're just like, yeah, this is a great model. Is it really?

And you just find-- Well, totally. Like, I mean, there's some pretty funny jokes on X. Like, one of them, I saw, was your girlfriend's of 10, but she's a benchmark maxer. [laughter]

Exactly. Is she right? Is she really dead? Might be past past. Yeah.

“But the no-overfitting thing has been a problem now for, I think, at least two or three generations of model families.”

And so this is another reason. I like this solution a lot. I like this more than the transnational Gulag organization. Well, we're going to say-- I like the approach.

She weeks without some grandiose international, like we don't need to convene the United Nations to make this happen. Yeah. No, it's just a decision. You can happen right now. All right.

You can always escalate the amount of regulatory oversight, but it is very difficult to reduce it.

You know, it does tend to be very much a one-way ratchet in terms of increasing regulations. Yeah. So what I'm suggesting here is it's a step in the right direction, and it's something we do quickly. And I think it's probably something that China would agree to. And you guys don't regulate yourselves then you're going to get regulated.

The MPA metaphors or analogy is incredibly crisp because the movie industry was faced with censorship and regulation by the government. And they just decided this is what an R is. And they literally created PG-13 for the Temple of Doom just to make it easy for people to understand PG versus PG-13.

“And I think it's an elegant solution, and we appreciate you joining us for the fifth year in a row.”

Welcome. Elon Musk, and then we got to get to a meeting. Yeah. I'm going to go pick some GPUs here in Memphis. Yeah, I'm going to rack the stack.

I'm going to join the group in the back and stack. Yeah, I'm going to raise for the machine. I'm going to raise for the machine. Yeah, well, enjoy your air stream. Thank you.

Thank you.

The first time Elon invited me down to Starbase.

Starbase. He's like, come down. You got to see what I'm building. I said, okay, yeah, I'll come down. He's like, yeah, I was like, was there like a hotel or something?

No, no, I got like a two bedroom or whatever. Come down, stay. I come down. And this is the like dilapidated house on a swamp. And we're outside.

And we're getting in a lot. I'm risking. There's the basement. Oh my god. You know, you can afford to get a house.

And he's like, yeah, I don't have time. I need to get these rockets up. I thought what you said was, oh my god, you live like I do. Yeah. Exactly.

I mean, it was pretty sparse. But I was like, you know what, I think you could treat yourself to a mobile home at this point. Elon. All right. Get back to work.

Thanks Elon. Thank you. Glenn, thank you very much. Glenn, thank you. Thank you.

Thank you. Thanks for sticking around. All right. That's great. Thank you.

Glenn, thank you again. - Thank you.