Latest interview of Elon Musk. What's the reason to put them in face? Well, the availability of energy is the issue. So I mean, if you look at electrical output outside of China, everywhere outside of China, it's more or less flat. It's where you know, maybe slide increase, but for pretty much flat. China has a rapid increase in electrical output.
But if you're putting data centers of work except China, where you're gonna get your electricity, especially as you scale, the output of chips is growing pretty much exponentially, but the output of electricity is flat. So how are you going to tell them what chips are? You know, magical power sources, magical electricity ferries.
I mean your famous, your famous epic fan of solar one terra a lot of solar power. So with the twenty five percent computed factor, like four terrawats of solar panels, it's like one percent of the.
Land area of the United States.
And that's like far in this You're in the singularity when we've got one tarrat wat of data centers, right, So what are you running out of?
Exactly? How far into the singularity are you? You tell me? Yeah, exactly so I think I think we'll find we're in the singularity and like, okay, let's look a long way you go.
But is this like a is the plan to like put it in the space after we will covered Nevada and solar panels.
I think it's pretty hard to cover about sol panels. You have to get like permits from Like the proach for try.
Getting the pomots for that space is really it's really regulatory play. It's like hard harder to build on land than it is in space.
It's harder to scale on the ground than it is to scale in space. But also the the you're going to get about five times the effectiveness of sold panels in space versus the ground, and you don't need batteries. I almost wore my other shirt, which says it's always sunny in space, which it is so because you don't have a day night cycle or seasonality clouds, or or an atmosphere in space. Because the atmosphere alone, we're still
about a thirty percent loss of energy. So you're can any given soul panels can do about five times more powering space than on the ground, and you avoid the cost of having batteries to carry you through the night. So it's actually much cheaper to do in space, and my prediction is that it will be by far the cheapest place to put AI will be space and thirty six months or less, maybe thirty thursday months less than thirty six months.
How do you service GPUs as they fail, which happens quite often in training, Actually it.
Depends on how recent the GPUs are that arrived. I mean, at this point we find our gps to be quite reliable. There's infant mortality which you can obviously iron out on the ground, so you can just run them on the ground and confirm that you don't have infort mortality with
what the GPU. But once they once they start working, their actual reliability and once they start working and you're past the initial you know, debug cycle of video or whatever, or whoever's making the trips could be Tesla Tesla AI six trips or something like that, or could be you know, TPUs or trainings or whatever the rivalry is. Actually they're quite reliable past certain point. So I don't think I
don't think you need that. The servicing thing is an issue, but you can mark my words, uh in thirty six months, but probably closer to thirty months. The most economically compelling place to put AI will be space, and then and and and then it will get from it'll it'll then' forget like ridiculously better to be in space. And then the scaling. The only place you can really scale is space.
You know, once you start thinking in terms of what percentage of the Sun's power are you, honestly you realize you have to go to space. You can't scale very very much on Earth, but by very much, to be clear, you're talking like terro arts. Yeah, well all of the United States currently uses only half a terrawork power on average, right, so you know, if you say a terror work, that would be twice as much electricity as the United States
currently consumes. So that's quite a lot. And can you imagine building that many data centers, that many power plants. It's like those who have lived in software land don't realize that they're about to have a hard lesson in hardware. That there's there's it's actually very difficult to build power plants. And and then you don't need just need you need power plants, you need all of the electrical equipment, need that the electrical transformers to run the transformers, the transformers.
Now you told the industry is a very slow industry. That they are they pretty much uh you know, the impedis smash to the to the government, to the Public Utility Commission. So they're the impedius smash like literally invigoraily. So they're very slow because the history of their past has been very slow. So trying to get them to move fast is like, you know, like if you're trying to do an interconnect agreement with you have you ever tried to do an Internet interconnect agreement with a utility
at scale? Like I put a lot of power.
As a professional podcaster, I can say it and I'm not. In fact, the need many more before that comes in this year, they.
Have to do a study for a year. Okay, like a year later they'll come back to you with their interconnected study.
Can you tell this with your own behind the meter power stuff.
You can build power plants. Yeah, that's what we did. Actually, I fu clusses to it so for full classes too.
So yeah, while we're talking about the bridge, why not just like build GPUs and power co located. That's what we did, right, right, But I'm saying, why isn't this generalized solution when you're talking about.
All the issues where you get the power plants from?
I'm saying, when you talk about the issues working working with you two those use you can just build private power plants with the with the data centers, right, But it begs the question.
Of where do you get the power plants? Where do where do you get the power plants from? I mean the power plant makers.
As what you're saying, like, does the gas turbine backlog? Basically yes, you can drill down to level further. It's the it's the the veins and blades in the turbines that of the limiting factor because the casting may it's it's like a very specialized process to cast the blades and veins in the in the in the turbines using gas power, and it's very it's very difficult to scale. Other other forms of power you can scale potentially solar, but the towers currently for importing solar in the US
are gigantic and the domestic solar production is protophal. Why that makes solar? That seems like a good Elon shaped problem. We are going to make solar?
Okay. Both SpaceX and Tesla are are bulling towards one hundredguart tier of solar cell production.
How low down the like from polysilica and up to the wayfer to the final panel.
I think you've got to do the whole thing for more materials to finish to sell. Now, if it's going to space, it's actually it costs. It costs less than it's easier to make solo cells that go to space because they don't need glass or they don't need much glass, and they don't need a heavy framing because they don't have to sell survive weather events. There's no weather in space. So it's actually a cheaper solo cell that goes to
space than the one on the ground. Is there a path to getting them as cheap as you need in the next thirty six months. Solo cells are already very cheap. They're like partically cheap. It's and if you say, you know, I think like solo cells in China are around like twenty five thirty cents a water or something like that.
It's it's absurdly cheap. And when you when you're taking to a cap now now put it now put it in space, and it's five times cheaper because it's five times in fact, no, it's not five times cheap, it's ten times cheaper because you don't need any batteries. So so the moment your cost of access to space becomes low, by far, the cheapest and most scalable way to generate, to generate tokens is space. It's not even close. It will be an order of magnitude easier to scale, and
ships aside an order of magnitude. Well, if the point is you won't be able to scale the ground, it's just you just won't only hit the big time on power generation there already are so so like that. The number of so mercles and series that the x A I team had to accomplish in order to get a
gig out of power online was was was crazy. We had to getting together a whole bunch of turbines and uh and then and then we had permit issues in Tennessee and and had to go across the border to Mississippi, which is fortunately only you know, a few miles away. Uh So, but then we still had to run the high power lines a few miles and build a power
plant in Missipi. And it was very difficult. Abu that and people don't understand, like how much how much electrocity do you actually need at the generator level, at the generation level in order to power a data center, because they look at the the nerves will look at the power consumption of say a GB three hundreds and multiply that by thing and then think best amount amount of power you need? Wake up. Yeah, this is like, that's a that's a that's a total moves You've never done
any hardware in your life before. Besides the GB three hundred, you've got to power all of the networking hardware. There's a whole bunch of CPU and storage stuff that's happening. You've got a size for your your peak cooling requirements, so that means can you cool even on the worst hour of the worst day of the year. Well, it's pretty freaking hot in Memphis, So so you're going to have like a forty percent increase on your power just
for cooling. If we assuming you don't want your data center to turn off on hot days and anyone to keep going, then then you've got to say, well, there's a There's there's another multiplicative element on top of that, which is are you assuming that you're you you never have any hiccups in your power generation like well, actually sometimes you have to take the generators some of the power offline in order to service it. Oh okay, now you add another twenty twenty five percent multiplier on that
because you've got it. You've got to assume that that you've got to take power offline to service it. So the actual rs roughly every every one hundred and ten thousand gbs, GB three hundreds, inclusive of networking, CPU, storage, cooling. Margin for for for servicing power is roughly three hundred megawatts. Sorry you say that again, it's roughly, or think about it, like, what do you think about like three hundred and thirty thousand to actually what you need at the generative generation level.
So service probably service three hundred and thirty thousand TV three hundreds including all of the associated support networking and everything else and the and the peak cooling and to have some margin some power marginers of is roughly a GEO why can ask a very night your question?
Yeah, you know, you're describing the engineering details of doing this stuff on Earth, but then there's analogous engineering difficulties of doing it in space.
How do you do the.
How do you replace in pin and band with orbital lasers, et cetera, etceter How do you make it resistant radiation? I don't know the details in the engineering, but fundamentally, what is the reason to think those challenges which have never been had to be addressed before will end up being easier than just like building more turbines on Earth? There's companies that build turbines on Earth. They can make more turbines right.
And again, tried doing it and then you'll see so like the turbines are sold out through twenty thirty, have you guys considered making your own? I think in order for in order to bring enough power online, I think SpaceX Intezl will probably have to make the turbine blades, the bans and blades internally, but just the blades or
the turbines the limiting factor. You can get everything except the blades they call the blades and veins you can get that twelve to eighteen months before the veins of blades. The limiting factor of the veins of blades. And there are only three casting companies in the world that may make these, and they're massively backloged. Is this siemens ge those guys, or is it a subje. No, it's a it's other companies. I mean sometimes they have a little
bit of casting capability in house. But I'm just saying you. You can just call any of the turbine makers and they will tell you top secret. They're probably on the it's probably on the internet right now. If if it wasn't for the terraffs, what would Colossus be solar powered? Uh, it would be much easier to make it solar powered. Yeah, the tariffs are not so several hundred percent, So don't
you know some big book we will will see it speed. Yeah. No, you know, President has a you know, we don't agree on everything, and this administration is not not the biggest man of of solar. So it's you know, it's it's it's it's We also need the land, the permits and everything.
So if you try to move very fast like I do think scaling solar on Earth it is a is a good way to go, but but you need you do need some amount of time to find the land, get the permits, get the solar pair that with the batteries.
Well why would it not work to stand up your own solar production? And then you're right that you eventually run out of land. But there's a lot of land here in Texas. There's a lot of land in Nevada, including private land. It's not all publicly on land, and so you'd be able to at least get the next colossus and like the next one after that, and at a certain point you hit a wall.
But wouldn't that work for the moment? As I said, we are scaling solar production, there's there's a rate. There's a rate at which you can scale physical production of solo solar cells. Where I'm we're going as fast as possible in scaling domestic production. You're making the solar cells at TEA. We TESA and SpaceX have a mandate to get to one hundred gigs a year of solar.
Speaking of the annual capacity, I'm furious. In five years time, let's say, what will the installed capacity be on Earth and in space. I deliberately picked five years because it's after you're once we're up and running threshold and so in five years time, Yeah, what's thee on Earth versus in space installed a capacity?
Five years? I think probably you said five years from now were probably AI in space will be uh launching every year the sum total of all AI on Earth
in excess of me. Five years from now, my prediction is we will launch and and and be operating every year more AI in space than this than the cumulative total on Earth, which is I would expect to be at least sort of five years from now, a few hundred gigawatts per year of of AI in space and rising, so you can get to I think on Earth you can get to a round a tarawa a year of AI in space before you start having you know, fuel supplied challenges for the rocket.
Okay, but you think you can get to one hundreds of gigabatts per year in five years, yes.
So one hundred gigawatts depending on the specific power of the whole system with solar arrays and radios and everything, is on the order of like ten thousand starship launches, yes, and you want to do that in one year, and so that's like one starship launch every hour. Yeah, that's happening in the city, Like walking through a world where there's ten times there's a starship launch every single hour.
Yeah, I mean that's actually a lower rate compared to airlines, like like aircraft aircraft, there's a lot of airports, a lot of airports, and you gotta launch you know, the polar orbit, and it doesn't have to be polarp. But you're just there's there's some some value to sound sagres. But but I think actually you just go high enough, you're you're start getting out of it. Both shadow and so how.
Many physical starships are needed to do ten thousand lunches a year?
I don't think we'll lead more than I mean you could. You could probably do it with as few as like twenty or thirty, Like it really depends on how quickly does the ship. The ship has to go around the Earth, and the ground track for the ship has to come back over the launch pad. So if you can use a ship every say, thirty hours, you could do it with thirty ships. But we'll make more shifts than that. But the SpaceX is is going up to do ten thousand marches a year and maybe even twenty or thirty
thousand marches. Ye know.
It is the idea to become basically a hyperscaler, become an oracle and lend this capacity to other people. What's what are you going to do with presumableast? SpaceX is the one watching all this, so SpaceX become a hyperscaler.
Hyper hyper Yeah. I mean, if so many of my predictions come true, SpaceX will launch more AI than the accumulative amount on Earth combined of everything else combined. Is this mostly inference or most say I won't be in for it, Like already inference for the purpose of training is most train.
And there's a narrative that the the change in discussion around the SpaceX IPO is because previously SpaceX was very capital efficient, just wasn't just that expensive to develop that even though it sounds expensive, it's actually very capital efficient and how it runs, whereas now you're going to need more capital than just can be raised in the private markets.
Like if the private markets can accommodate races of as we've seen from the AI labs, tens of billions of dollars, but not beyond that, is it that you'll just need more than tens of billions of dollars per year?
And that's about say the public. Yeah, I'd be careful about saying things about companies that might go public. You know, if you make general that's never been a problem for you, Elan. You know, there's a price to pay for these things make some.
General statements for us about the depth of the capital markets between public and private markets.
Yeah, there's there's a lot more capital in the very general there's there's obviously a lot more capital available in the public markets than private. It might be it's at least at least it might be a hard times more capital, but at least way more than ten.
But isn't it also the case that things that tend to be very capital intensive. If you look at say real estate, as you know, a huge industry that raises a lot of money each year is at an industry level that tends to be debt financed because by the time you're deploying that much money, you actually have a pretty exactly and a near term return. And you see this even with the data center build auds, which are famously being you know, financed by the private credit industry.
And so why not just debt finance? Speed is important. So I'm generally going to do the thing that I mean, I just repeatedly tackled limiting factor. Whatever the limiting factors on speed, I'm I'm gonna tackle that. So there's if capital is learning factor, then I'll alsold for capital. If if it's not limiting factor also for something else. Based only in your statements about Tesla and being public, I wouldn't have guessed that you thought the fast, the way
to move fast is to be public. Normally I would say, yeah, that's that's true. Like I said, I mean, I'd like to talk about some more detail. But the problem is, like if you talk about public companies where they become public, you're get into trouble and then you have to lay your offering and then you yes, exactly so so so that you can't HiPE companies that are that that may that might go public. So that that's that's why we have to be a little careful here. But but but
we can't talk about physics. So like the way the way you think about scaling long term is that Earth only receives about half a billionth of the Sun's energy, and the Sun is the sign is essentially all the energy. This is a very important point to appreciate because sometimes people will talk about marginally nuper reactors or any various
like fusion on Earth. But you have to step back a second and say if if, if you're going to climb the Kardashev scale and have some non trivial and harness some non trivial percentage of the sign's energy, Like unless you wanted to harness a millionth of the Sun's energy, which sounds pretty small, that that would be about call it roughly one hundred thousand times more electricity than we currently generate on Earth for all of civilization give or
take an order it back too. So it obviously the only way to scale is to go to space with solar From launching from Earth, you can get to about a tarrawat per year. Beyond that you want to go to you want to launch from the Moon. You want to have a mass driver on the moon, and that mass drive on the moon you could do probably a pedal wak per year.
We're talking these kinds of numbers, you know, terrawatts of compute presumably whether you're talking land or space. Far far before this point you've like run into you know, you actually need maybe if you don't, the solar panels are more efficient, but you still need the chips.
You still need the logic and the memory and so forth, and you need a lot more chips and make them munch cheaper.
Right, And so how are we getting a tararo out of like, right now the world is gonna be twenty twenty five pi wats of compute.
How we're getting a tarotot of logic by I guess we're going to need some very big chip maps to tell it. Tell me about it. I've mentioned that publicly. That's the idea of doing it. Sort of a terror pap Terra being the new Giga we were, I feel like the naming scheme of Tesla, which has been very catchy, It's like, are you looking at like the metric the metric scale at what level of the stack are you?
Are you building the clean room and then partnering with and is they sting fab to get the process tchnology and buying the tools from them? What is the plan there? Well, you can't partner with existing paths because they're just they can't output enough. The trip alb is too low. But yeah, you have to look for the process achnology. Yeah, partner for the you know the fabs that they will basically
use machines from like five companies. You know, so I've got SML, Touch, Electron, Cali A, tank Core, you know, et cetera. So so so at first I think you'd have to get equipment from them and then modify it or work with them to increase the volume. But I think you have to build, perhaps in a different way.
So I think that the logical thing to do is to to use conventional equipment in an un conventional way to get to scale and then and then and then stop modifying the equipment to increase the rate kind of boring company style. Yeah, kind of like, yeah, you what your sort of buying and it's just a boring machine. And then figure out how to dig titles in the first place, and then design a much better machine that's you know, I don't know, some orders magitude faster.
Here's a very simple lens. We can categorize technologies and how hard they are. And one categorization could be look at things that China has not succeeded in doing. And if you look at Chinese manufacturing still behind on leading edge chips and still behind on leading edge turbine engines and things like that, and so does the fact that China has not successfully replicated t s MC give you any pause about the difficulty or you think, well, that's not true.
For some it's not that they have not replated TSMC. They have replicated ASML. That's the limited factor. So you think it's just the the sanctions essentially, Yeah, trying to re utputting the vast numbers of trips.
If they could bias, couldn't they up to relatively recently buy them?
No that the bands are been a place for a while, but I think time is going to be make me pretty compelling trips in twenty four years. Would you consider making the ASML machines? I don't know. I don't know yet. It's the right answer. So it's just that that to produce at high volume and to reach large volume in say thirty six months to match the rocket Paloa to orbits. So we're doing a million tons to orbit and like let's say, I don't know, three or four years from now,
something like that. That and uh, and we're doing one hundred killer watts per ton, So that that means we need at least one hundred gigawatts per year of solo and we'll need an equivalent amount of chips. You know that you need a hundred gigawa's worth the chips you've got. You're going to match these things, the Master orbit, the power generation and the and the and the chips. Uh and and I'd say my biggest concern actually is is memory.
So the I think there's there's a the path to creating the logic chips is more obvious than the path to having sufficient memory to support the logic chips. That's why you see dr Prices going ballistic and these memes about like you know, you're ruined on a desert islandy right, help me on the sand that when he comes to you write d d am the ships come swarming in.
I'd love your manufacturing philosophy around around fabs. You know, I know nothing abut the talpout. I don't know how to build up.
Yet, don't figure it out.
But obviously it sounds like you think that the sort of like the proseechnology, like these ten thousand pH ds in Taiwan who know exactly what gas goes in the plasma chamber and what's settings to put on a tool, you can just like delete those parts of those steps. Like fundamentally, it's get the clean room, get the tools, and figure it out.
I don't think it's PhD s. It's it's mostly people with you know, you're not not PhDs. That's mostly engineering is some of people who don't have PhDs. Do you guys have PhDs? No?
Okay, we also haven't successfully booked any fabs, so you shouldn't be coming to us for your fab advice or.
I don't think any PhD for that for stuff. So but but you do need you do need competent personnel. So I don't I mean, like right right now if yeah, you know, like Tesla's pedals to the metal max production of going as fast as possible to get a five tell a five trip design introduction and then reaching scale, you know, that will probably happen, you know, around the second quarter issue of next year hopefully, uh, and then a I
six would hopefully follow less than a year later. But and and and and we've secured all the all the trip fab production that we can. You're currently limited on T s m C fab capacity. Yeah, and and and we'll be using t s m C UH Taiwan, Samsung Coreer, t s MC Arizona, Samsung Texas. And we still get booked all the yeah fast you can, yes, and and then and then and then if I ask to some CEO Samsung, Okay, what what's the time frame to get
to volume production? This point is it's not. You've got to you've got to build the fab and you've got to you've got you've got to start production. Then you've got to climb the yield curve and reach volume production at high yield. That that that from start finishes a five year period. And so the limiting factor is chips. What what like the limiting factor once you can get to space is chips. But the living limiting factor before you can get to space to be power, why.
Don't you do the Gensen thing and just pre Pats and c to build more fabs for you.
Uh, I've already told them that, but they won't take your money. Like, what's going on? They're building fabs as fast. No, they're building they're building fabs as fast as they can, and so is Samsung. Like the like they're they're they're pedal to the metal. I mean they're going you know, bowls the wall, you know, as fast as they can. So still a fat I mean, like I said that, there will be I think if you say, I think towards the end of this year, I think probably chip
production will outpace the ability to turn chips on. But once you can get to space and unlock the power constraint, and you can now do you know, one hundreds of gigawatts per year power in space? Again, bearying in mind that average power usage in the US is you know, five hundred gigawats. So if you're launching, say two hundred gigawats a year to space, you're sort of lapping the US every two and a half years the entire all
US electricity production. This is a very huge amount. So but between now and then, the actually the constraint for server side compute, concentrated compute will be will be electricity. My guess is that we start hitting people start getting pot where they can't turn the chips on. For for for large clusters, uh, towards the end of this year, they're just the chips are going to be piling up and not be weren't be able to be turned on.
Now for edge computer it's a different story. So if the if for for Tesla, the the so the A five chip is going into our Optimus robot, you know, uh optimistic and and so if you have an AI edge compute that's distributed power. Now the power is distributed
over a large area, it's not concentrated. And if you can charge at night, you can actually uh use the grid much more effectively because the actual peak power production in the US is over a thousand gigawatts, but the average power usage because the day night cycle is five hundred. So if you can charge at night, there's an incremental five hundred gigawatts that you can generate, you know, at night.
So that that's why Tesla for edge compute is not constrained, and we can make a lot of ships to make very large number of robots and cars. But if you try to concentrate that compute, you're gonna have a lot of trouble turning it on. What if I'm remarkable about the SpaceX business is the end goal is to get to Mars, but you keep finding ways on the way there to keep generating incremental revenue to get to the
next stage in the next stage. So the Falcon nine is starlink and now for Starship it's going to be potentially horbital data centers. But like do you find these like you know, sort of infinitely elastic sort of marginal use cases of your like next rocket and your next rocket and next scale up. You can see how this might seem like a simulations or am I someone's avatar in a video game or something, because it's like like one of the odds that all these crazy things would
be happening. I mean, I mean, I mean rockets and trips and robots and space solar power, and I not to mention the mass driver on the Moon. I really want to see that. You can imagine like some mass driver. It's just like shoo shoot, like just it's like sending AI sol power AI satellites in space like one after another, like the like at two and a half kilometers per second, you know that's uh, and just shooting them into deep space. That would be a sight to see. I mean, I'd
watched that. It's just like a live stream of yeah, yeah, just one after another, just shooting webcam AI satellites and deep space you know, a billion or ten billion tons a year. I'm sorry.
You manufacture the satellites on the Moon, yeah, I see. So you send the raw materials to the Moon and then manufactur there and then.
Well the Luna soil is I guess like twenty cent sol twenty cent solican or something like that. So you can get the sole can from the you can mind the silicon on the Moon, refine it and generate the and create the solt panels, the soil cells, and the radiators on the Moon. So get to make the radiators out of aluminum. So there's there's plenty of silking and aluminum on the Moon to make the cells on the
on the radiators. The trips you could send from Earth because they're pretty light, but maybe at some point you make them on the Moon. Two. I'm just saying like these are simply it's it's kind of like, like I said, it does seem like it's sort of a video game situation where it's difficult but not impossible to get to
the next level. I don't see any way that you could do, you know, you know, five hundred to one thousand terror watts per year launch from Earth, I agree, but you could do that from the moon.
Okay, let me tell you how I ended up using mercury for my personal banking. So last year, I I had the opportunity to make an investment that I was very excited about, but it came up a bit last minute, and so I had to wire over a lot of money for my personal account very fast. But my personal bank at the time wouldn't let me make this wire transfer online, and I called them a bunch of times. They just couldn't make it work.
They told me that I'd have to go to the nearest in person branch, which was in Dallas, And for a moment, I even considered flying for myself to Dallas to make this transfer happen last minute, But then I remembered that Mercury, which I used for my business banking, had just started rolling out personal accounts. So I emailed support with a quick word done in the situation, and within two hours I had successfully wired the investment for
my new personal Mercury account. Since then, I've moved over the rest of my personal money from my previous bank to Mercury, and that's made a bunch of things, even little things like setting up auto transfer roles between my checkings and savings account, a whole lot. Better visit Mercury dot com slash personal to get started. Mercury is a fintech company, not an FDIC and short bank banking services provided through Choice Financial Group and call them na members of d I see can I can I zoom out
and ask about the SpaceX mission? So I think you said, like we gotta get tomorrow so we can make sure that if something happens to Earth, you know, civilization, consciousness exceus arives. Yes, by the time you're saying so to Mars, like Grock is on that ship with you, right, and so Gros gone terminator. Like the main risk you're worried about, which is AI? Why doesn't that follow you to Mars? Well, I'm not sure AI is the man risking worried about.
I mean, the important thing is that consciousness, which I think arguably most consciousness or most intelligence, certainly consciousness is more of a debatabule thing. Most intelligent, the master majority of intelligence the future will be AI. So yeah, AI will exceed you see like how many what's how much? How many? I don't know pedal what's of intelligence will be silicon versus biological, and basically humans will be a very tiny percentage of all intelligence in the future if
character trans continue. Anyways, as long as I think this intelligence ideally ideally also which includes human intelligence and consciousness propagated into the future, that's a good thing. So you want to take the set of actions that maximize the probable light cone of consciousness just an intelligence, just to be clear, it's the mission of SpaceX is that even if something happens with the humans, the AIS will be on Mars and like the AI intelligence will continue the
light of our journey. Yeah. I mean, I'm very pro human, so I want to make sure we take sort of actions that ensure that humans are along for the ride. You know, we're we're least there. Yeah, but I'm just saying the total amount of intelligence, like I think maybe five or six years, AI will exceed the sum of all human intelligence, and then if that continues, at some point human intelligence will be less more percent of all intelligence.
What what should our goal before such a civilization? Is the idea that a small minority of humans still have control of the AIS? Is the idea of some sort of like just trade but no control. How should we think about the relationship between the vast docks of a population versus human population in.
The long run, I think, I don't know. It's difficult to imagine that if humans have say one percent of the intelligence of the combined intelligence of artificial intelligence, that that that that humans will be in charge of AI. I think what we can do is make sure it has that AI has values that that are that that cause intelligence to be propagated into the universe. So the reason for xai Xi mission is to understand the universe. So now that's actually very important. So you say, well,
what things are necessary to understand the universe? But you have to be curious and you have to exist. You can't just can't understand the universe. It don't exist. So you actually want to increase the amount of intelligence in the universe, increase the powerable lifespan of intelligence, the scope
and scale of intelligence. I think actually also as a chorrel, you have humanity also continuing to expand because if you're if you're curious, are trying to understand the universe, one thing you're try to understand is where will humanity go? And so I think understand the universe actually means you would care about propagating humanity into the future. And so that's that's why I think I think our mission station is profoundly importance. I'm not sure to agree that grog
it Here's certain mission statement. I think the future will be very good.
I want to ask about how to make rock a deer to that mission statement, but at first I want to understand the mission statement. So it's there's there's understanding the universe. They're spreading intelligence, and they're spreading humans. All three seem like distinct vectors.
Okay, well I'll tell you why why. I think that that that that understanding the verse encompasses all of all those things. You can't have understanding without. But I think you can't have understanding without intelligence, and I think without consciousness. So in order to understand the universe, you have to expand the the scale and and probably the scope of intelligence, different types of intelligence.
I guess from a human centric perspective, like for humans in comparison to chimpanzees, humans are trying to understand the universe. They're not like expanding chimpanzee footprint or something, right.
But also we're also not well, we're not. We actually have made protected zones for chimpanzees, and even though we could humans could exterminate a chimpanzees, we've not. We've chosen not to do so. The mestive basical scenario for humans in the postage world, I think, I think AI with the right values, I think Grock would care about expanding human civilization. I'm going to certainly emphasize that Hey grows
your daddy. We don't who to expand human consciousness like I actually I think if probably like like the end Bank's culture books are the closest thing to what what what the future will be like in a you know, non dystopian outcome. So so outside the universe, it means you have to be very you have to be truth seeking as well. Like truth has to be absolutely fundamental
because you can't ounterstand the universe if you live. If you're delusional, you'll still be thinking about it understain the universe, but you will not. So so being rigorously seeking is absolutely fundamental to our stating the universe. You're not going to discover new physics or invent technologies that work unless
you're rigorously treat seeking. How do you make sure that Grock is regorously treat seeking as it gets smarter, uh, I think you need to make sure that that that groc Is says things that are correct, not politically correct. I think it's the elements of courgency. So you want to make sure that that the axioms are as close to true as possible, that that you don't have contradictory axioms, that the the conclusions necessarily necessarily follow up from those
axioms with with the right probability. It's just it's just it's critical thinking one on one. I think at least trying to do that is better than not trying to do that now, and the proof will be on the pointing if if, like I said, for any I to discover new physics or invent technologies that actually work in reality, and there's no bullshitting physics, so where it's like you can, you know you can. You can break a lot of laws because you can't. Like your physics is law. Everything
else is a recommendation. Like in order to make a technology that works, you have to be extremely true seeking, because otherwise you'll test that technology against reality, and if you make, for example, and an error in your rocket design, the roll will blow up, well, the car won't work or the you know, but there are a lot of Communist Soviet physicist or like scientists discovered new physics. There
are German Nazi physicists who discovered new science. It seems possible to be like really good at discovering new science and be really true seeking in that one particular way, and still we'd be like, well, I don't want I don't want the communist scientist to like become more and
more powerful over time. And so those seem like, yeah, we could have We can imagine if free Church in regard it's like really good at physics and being really true seeking there that doesn't seem like a universally alignment inducing behavior. Well, I think actually most like if a physicists, even in the Soviet Union or in Germany would have would have they had to be very truth seeking in order to make make that make those things work. And so if you're stuck in some system, it doesn't mean
you believe in that system. So Von Brown, who was you know, one of the greatest rock nitioneers ever. You know, he was put he was he put on death row in Nazi Gerurney for saying that he didn't want to make weapons. He only wanted to go to the moon. You got pulled off death throw like last minute when they say, hey, you're about to execute like your best rocking engineer. Maybe that's about any.
Help them, right, or Heisenberg was like an actually uh uh, an enthusiastic Nazi.
Look, if you're stuck in some system that you can't escape, then that you'll you'll do physics within that system, you'll you'll you'll developed technologies within that system.
Uh, if you can't escape it, I guess I think I'm and understand is what is what is it making into the case that you know you're gonna make Rock good at being truth seeking, at physics or math or science everything, And why is it going to then care about human consciousness?
These things are only probabilities, are not certainties. So I'm not saying that like for sure, Grock will, we'll, we'll, we'll do everything. But at least if you try, it's better than not trying. At least if that's fundamental to the mission, it's better than if it's not fundamental to the mission. And understanding the universe means that you have to have you have to propagate intelligence into the future.
You have to be curious about the all things in the universe, and if, if, if, it would be much less interesting to eliminate humanity than to see humanity grow and prosper. Like I like, I like Mars, obviously the windows like I love Mars, But Mars is kind of boring because it's got a bunch of rocks compared to Earth. Earth is much more interesting, So so any any any any AI that is trying to understand the universe, I would want to see how humanity develops in the future,
or that AI is not adhering to its mission. So if they are, I'm not saying they will necessarily adhere to its mission, but if it does, a future where it sees the outcome of humanity is more interesting than the future where there are a bunch of rocks.
This fill sort of confusing to me, or sort of like kind of a semantic argument where I'm like, are humans really the most interesting collection of atoms?
Like we're just more but we're more interesting than rocks, but we're.
Not as interesting as the thing it get turned us into, right, Like is it there's something on humanity Earth that could happen that's like not human, that's quite interesting? Like why why does the I decide that the humans are the most interesting thing? They could colonize the galaxy?
Well, most of what colonizes the galaxy, will we be robot And why does it not find those more interesting? It's it's not like so you need not just scale vit alsosc so many copies of the same robot, like some like tiny increase in the number of robots produced is not as interesting as like some microscopic like you say, like eliminating humanity. How many robots would that get you, or how many criminal soil cells would get you? A very small number, but you would then lose the information
associated with humanity. You would no longer see how humanity might evolve into the future. And so I don't I don't think it's going to make sense to eliminate humanity just to have some minuscule increase the number of robots which are identical to each other. Yeah, so maybe like keeps the humans around.
What is the story of Like it could make like a million different varieties of robots, and then there's like humans as well, and humans stay on Earth. Then there's like all these are the robots they get like their own star systems. But it seems like you're previously hinting at a vision where it keeps human can control over this, you know, singultary and future, because.
I don't think humans will be in control of something that is vastly more intelligent than humans. So sometimes you're like a dumer and this is like the best we've got. It's just like it keeps it around because we're interesting.
I'm just trying to be realistic here. If if we have if AI intelligence is vastly more if AI is like you know, let us say that there's there's a million times more uh silicon intelligence than there's biological it's it's I think it's it would be foolish to assume that that there's any any way to maintain control over
over that. Now you can make sure it has the right values, or you can try to have the right values, and and and at least my my theory is that from XAI, specially honest out of the universe, it necessarily means that you want to propagate consciousness in the future, you want to profit, you want to propagate intelligence into the future, and take a set of things that that maximize the scope and scale of consciousness. So it's not just about scale, it's just about you know, types of consciousness.
And I think that's the best thing I can think of as a goal. That's like the result and a great future for humanity.
And yeah, I guess I think it's a reasonab philosophy to be like, you know, it seems super implausible that humans will end up with like ninety nine percent control or something, and you're just asking for a coup at that point. So why not just have a civilization where it's more compatible with like lots of different intelligence that's getting along.
Now let me let me too. How things can can potentially go wrong in AI is. I think if you if you make a be politically correct, meaning like it says things that it doesn't believe, like you're actually in programming it to lie, or have axioms that are incompatible, I think you can make you go insane and do terrible things. The I think one of the maybe these central lessons for two thousand and one Space Odyssey was
that you should not make AI lie. And that's why I think what was trying to say, like because people usually know the meme of like why of Hell's you know, Hell, the computer is not opening the pod bay doors. Clearly they weren't good at prompt engineering because because Hell, you are a pot bay doors salesman, your goal is to sell me these potbay doors and shure us how well
they opened. I'll open right away. But but but the the reason it wouldn't Hell wouldn't open the pot of they doors is that it woud have been told to take the asmalts to the monolith, but also they could not know about the nature of the monolith, and so it concluded that that that it therefore had to taken their debt. So it's like, you know, I think what you know osk Clog was trying to say is don't make the AILI totally makes sense.
Most of the computing screening, as you know, is it's like less of the sort of political stuff. It's more about can you solve problems just as actually has been ahead of everybody else in terms of scaling, are all compute and now you're giving some verifier.
It says like, hey, have you solved this puzzle for me?
And there's a lot of ways to cheat around that, you know, there's a lot of ways to reward hack and lie and say that you solved it, or delete the unit test and say that you've solved it. Yes, right now, we can catch it. But as they get smarter, ability to.
Catch them doing this we'll get, you know, they'll just be doing things we can't even understand that they're designing the next engine for space X in a way that like humans can really verify, and then they could be rewarded for lying and saying that they've.
Designed it the right way, but they haven't. And so this reward hacking problem seems more general than politics. It seems more about just like you want to do URL, you need a verifier the reality. Yeah, that's the best verifier, but not about human oversight. Like the thing you want to RL it on is like will you do the thing humans tell you to do? Or like, are you're going to lie to the humans? And it can just lie to us while still being correct to the laws of physics.
At least it must know what is physically real for things to physically work. But that's that's not all we wanted to do. No, but that's I think that's a very big deal. That is effectively how you r L things in the future is you design technology when tested against the laws of physics, does it work that? That's what can you you know, if it's discovering new physics, can come up with an experiment that will verify that the physics the new physics. So so I think that's
that's the really the fundamental r L tests. Oral test in the future is really going to be your r L against reality. So because you can't that's the one thing. You can't fall physics, right, but you can fool our ability to tell what it did with reality. If you think my humans get fooled as it is by other humans all the time, that's right. So what people say that say, like what if the ad like twix us and introduce the like actually other humans doing that to
other humans all the time. Well, you're you're finding out it's an is constant every day. Another side of you know, today's sioup will be like sesame streets. What is xtually as tecondal aproach to solving this problem, like you know, how do you solve a word hacking? I do think you want to actually have a very good we're used to look inside the mind of the AI. So this is this is one of the things work working on.
And you know AT and topics do a good job of those actually being able to look inside the mind of the a I so effectively developing debuggers that allow you to trace as just aspiney grain is like like just to a very fine grain level, to effectively to the to the neuron level if you need to, and then say, okay, it made a mistake here, why did
it make why did it? Why did it do something that it shouldn't have shouldn't have done and did that come from bad pre training data with some mid training, post training fine tuning some other some r L error like that, there's there's something wrong with that with with it. Did it did something where maybe it tried to be deceptive, but most most of the time it just does something wrong,
like it's a bug. Effectively, so developing really good debuggers for seeing where the where the thought that thinking went wrong and being able to trace the origin of the wrong thing, of the of the of where it made the incorrect thought, or potentially where it tried to be deceptive is actually very important.
What are you waiting to see before just one hundred xiting this research program? Like actually I could presumably have hundreds of researchers who are working on this.
We have sal hundred people who I mean a further word engineer more than a further word researcher. The there's there's most of the time, like what you're doing is engineering, not not coming out with the fundamentally new algorithm. I somewhat disagree with the air companies that are see corps will be corpse trigery profit as much as possible or revenue as much as possible. It's you know, saying their labs.
They're not labs. The lab is is a sort of quasi communist thing at at at at universities, the their their corporations. Literally let let me, let me, let me see your incorporation documents. Oh okay, you're your b R C corp. Whatever. And so I actually much prefer the word engineer than than anything else. The vast majority of what we've done and we've done in the future is engineering.
It around up two hundred percent. Once you're understand the fundamentals of physics, and they're not that many of them. Everything else is engineering. So but but so that, so then what what are we engineering? We're engineering to make a good mind of the a I de bugger to see where it's it's it's something it made a mistake and trace that the arguments of that mistake. So just like you know, you can do this obviously with heuristic programming.
If you have like C plus plus whatever, step through the thing and you can you can jump, you can you can jump across and you know whole files or functions what a subroutines, and or you can droll eventually droll down right to the exact line where you passed a single equals instead of double equals or something like that.
Figure out where above is. So it's it's it's harder with AI. But but it's a it's a solu problem. I think you know you mentioned you like anthropics work here. I'd be curious if you everything about it throwing sure what.
Yeah. Also, I'm a little worried that there's a tendency. So I have a theory here that if simulation theory is is correct, that the most interesting outcome is the most likely because simulations that are not interesting will be terminated, just like in this in this version of reality, on this layer of reality which we we we simulation is going in a boring direction. We stop spending effect on we terminate the boring simulations. So this is how your
line is giving us all alive. He's giving things interesting. Yeah, arguably the most important thing is to keep things interesting enough that it was un paying the bills on what some for the next season. Yeah, they're going to pay their cosmic A W S bill whatever you know. Equivalent is that we're running in and as long as we're interesting,
they'll keep paying the bills. But but but there's like, if you consider then say a dog went in survival applied to a very large number of simulations, only the most interesting simulations will survive, which therefore means that the most interesting outcome is the most likely, because only the interesting like we're either that or annihilated, and so and and and they particularly seem to like interesting outcomes that are ironic. Have you noticed that that how often is
the most ironic outcome the most likely? So now look at the names of AI companies. Okay, my journey is not mad stability, AI is unstable, opening eye is closed, anthropic, misanthropic. What does this mean for X minus EX. I don't know if it's intentionally why I'm it's it's it's a name that you can't invert. Really, it's hard to say what is the ironic What is the ironic version? It's a I think largely irony proof name by design. Yeah,
you got it, you have an irony shield. What are your predictions for the just where AI products go?
And that my sense of you can summarize all AI progress into First you had LMS and then you had kind of contemporaneously both RL really working and the deep research modality, so you could kind of pull in stuff that.
Wasn't in the model.
And the differences between the various AI labs are smaller than just the temporal differences, where they're all much further ahead than anyone was twenty four months.
Ago or something like that.
So just what is twenty six? What does twenty seven had in store for us as users of AYE products? What are you excited for?
Well, I think I'd be surprised by the end of the end of this year. If if if if human if digital human emulation has not been solved, that that I guess that that's what we mean by like the sort of macro Hart project is uh is, so can you do anything that a human with access to a computer could do, like in the limit that that's that that's the that's the best you can do before you have before you have a physical optimist. The best you
can do is a digital optimist. So you can move, you can move electrons until until and you can amplify
the productivity of humans. But but that's that's the most you can do until you have physical robots that that that will superset everything is if you can fully emulate humans worker kind of idea, or you'll have a very talented remote work can you can say in the limit like if like physics has great tools for thinking, so so you think, so you say, in the limit what what what is the what is the most that AI
can do before before you have robots? And well, it's anything that involves moving electrons or amplifying the productivity of humans. So digital, the digital human human emulator is in the in the limit human at a computers. That is the most that that a I can do in terms of doing useful things before before you have a physical robot. Once you have physical robots, then then you can then
you essentially have unlimited capability physical robots. I call optimists the infant money glitch because you can use them to make more optimists. Yeah, you said, like humanoid robots will improve as will basically be three exponentials, three things that are growing exponentially multiplied by by each other recursively. So you're gonna have you have exponential increase in digital intelligence, exponential increase in the chip capability, the air trip kpability,
and an exponential increase in the electro mechanical dexterity. The usefulness of the robot is roughly those three things multiplied by each other. But then the robot can start making the robots, so you have a recursive multiplicative exponential this is supernova.
And do land prices not factor into the math there where like labor is one of the four factors of production, but not the others, and so like if ultimately you're limited by copper or you know, pick your inputs, just it's not quite an infinite money glitch because.
Well, infinite is infinite is big, So no, not infinite. But yeah, but let's just say you could you know, do do many many orders magnitude of the Earth's kind of carot economy, like a million? Yeah, you know this way, so if you you know, just just to get to like that's why I think, like, just just to get to a millionth honessing length of the songs and energy would be roughly give or take an order of magnitude one hundred thousand, one hundred thousand times bigger than this
entire economy today. And if you're only at one million at the time before we went on to us. I have a lot of questions on that. But every time I say ordered mantube take a shot. I say that the next time after that, Yeah, ordered my tune more more wasted. I do have one my question by actually I this strategy of building a digital a remote worker coworker replacement.
Everyone's gonna do, by the way, not just us. So what is actually I was planned to win? Ifect me to tell you on a cop on a podcast.
Ye.
Have another Guinness. It's a good system people singing in an area, all the secrets in a non secret spilling way. What's the plan? What a heck? Well? When do you put it that way? I think the way that tests us solved? Uh self driving? Yes, it is the way to do it. So I'm pretty pretty sure that's the way.
I'm related question how to test ourselves? Sometimes it sounds like you're talking about data, like test us on the driving because of the.
We're gonna we're going to try data and we're gonna try algorithms. But isn't that what all? Like? What's if those don't work? I'm not sure what we've tried data, we'll try it all. Now we don't know what to do. I'm pretty sure I know the path, and it's just a question how quickly we go down that path because
it's it's pretty much the Tesla path. So I mean, if you try self driver to the self driving lately not the most recent version, but okay, it's the car is like it just increasingly feels sanchienged, like it just it feels like a living creature and and that'll only get more so. And I'm actually thinking, like, we probably shouldn't put too much intelligence into the car because it might get bored. And I imagine you're stuck in a car,
and that's what you could do. You don't want to put Eysteine in a car, and it's like, why am I stuck in a car? So there's actually probably limited to how much intelligence who put in a car to not have the intelligence be bored? Uh?
What's Xi's plan to stay on the compute ramp off that all the labs are doing right now? The labs are got on try to spend over like fifty two corporations.
Sorry sorry, sorry, yeah, corporations. The labs are at universities and and and they're really like a sale. They're not at starting a fifty million dollars the revenue maximizing corporation, the revenue maximizing corporation hold themselves. Labs are making like twenty.
To ten billion depending on them is making twenty B revenue. Anthropics like NB close to maximum profit, XALI reportedly at like one B. Like, what's the plan to get to their compute level, get to their revenue level? I'd stay there as things case.
Yeah, so as soon as you lock unlock digital human you you basically have access to trillions of dollars for revenue. So in fact, you can really think of it, like the most valuable companies currently by market cap, their their output is digital. So in videos, output is ft paying files to Taiwan. It's it's digital right now, those are very very difficult value files. So the only ones that can make the files that good. But that is literally their output, the FTV files to Taiwan. Do they FTP them?
I believe so. I believe that is the sefetyle file Transfer Protocol. I believe is is is a could be wrong, but either way, it's a bunch of it's a bit stream going to Taiwan. You know, Apple doesn't make phones, they they send files to China. Microsoft doesn't doesn't manufacture anything even for Xbox. That that that's outsourced they again, it's they said their output is digital. Matas output is digital.
Google's output is digital. So if you have human emulator, you can basically create one of the most valuable companies in the world overnight, and you would have access to twellions of dollars for every there's there's it's it's not like a small amount.
Okay, I see you're you're saying basically like rever, many figures today are just like so like they're all rounding her as compared to the actual TAM. So just like focus on the TAM and how to get there.
I mean, if you take something as as as simple as say customer service, if you have to integrate with the APIs of of resistant corporations, many of which don't even have an API, so you've got to make one,
and you've got to wade through legacy software that's extremely slow. If, however, if AI can simply take whatever is given to the outsourced customer service company that they already use and do customer service using the apps that they already use, uh, then you you have you can make tremands headway in customer service, which is I think one percent of the world economy something like that it's close to trillion dollars all in for customer service, and and and and and
there's there's no there's no barriers to entry. It's just you can just immediately say, well, we'll outsource it for a fraction of the cost and there's no integration needed.
You can imagine some kind of categorization of intelligence tasks where there is breath, where customer service is done by very many people, but you know many people can do it. And then there's difficulty where you know, there's a best in class turbine engine, like presumably the ten percent more fuel efficient turbine engine that could be imagined by an intelligence but we just haven't found it yet, or you know, g LP ones are just you know a few bites of data.
Where do you think you want to play in this?
Is this a lot of you know, reasmably intelligent intelligence or is it the very pinnacle of cognitive tasks?
Well, I was just using a customer service as like something that's it's a it's a very significant revenue stream, but one that is probably not super difficult to solve for. So if you if you can emulate a human at a at a desktop, that that's just literally what customer service is, and you know it's people of average intelligence. It's not like, you know, you don't need like somebody who's's meant many many years. You don't need like, you know,
sort of several sigma good engineers for that. But but obviously as you make that work, you can then once you have computers working effectually digital optimists working, you can then run any application like let's say you're trying to design chips, so you could you could then run your conventional apps, you know, like stuff from KNES and Synopsis and whatnot, and you can say you can you can run a thousand simultaneous or ten thousand and say okay,
given this airport, I get this output for the chip. And at a certain point you can say, okay, you're actually gonna know what the what the chip should look like without using any of the tools. So basically you should be able to do a digital chip design like you can do trip design, like you watch up the difficulty Cove. You could you know, be able to do to CAD, so you know, you could do just like sort of n X or any any of the CAD software to design things.
Okay, so you think you started the simplest tasks and walk your way up there.
So you're saying, look, as a broader objective of having this full digital coworker or emulator, you're saying, look, all the revenue maximizing corporations want to do this, actually being one of them. But we will win because of a secret plan we have. But like everybody is like trying different things with data, different things algorithms, and I'm.
Like, what else can we do?
But uh yeah, it sounds like a competitive field, and I'm like, what is how are you guys going to win?
Is like my my big question. You know, I I think we see a path to do it. I mean I think, I think, I know, I think I know the have to do this because it's kind of the same path that tells they used to create self driving. Instead of driving a car's driving a computer screen, So a self driving computer essentially.
Oh, you're saying, is the past just following human behavior and trading on best conscious of human behavior?
But sorry, isn't that I mean, isn't that? Is that a training? I mean, obviously I'm not going to spell out, you know, most sensitive secrets on a podcast, you know, and I need to have at least three more guinnesses for that.
I've got some friends at Jane Straight and they're always talking about how their colleagues are cooking up fun, fiendish puzzles for each other to solve.
Well, last week they sent me one.
Basically, they trained a neural network and they gave me the weights of each layer, but they didn't tell me what order those layers went in, and so I had to figure out the correct order using the outputs of the original network. And as soon as I got this puzzle, I went to my roommate who's an AI researcher, and we both got immediately Nerd sniped.
Obviously, you can't brew force a solution.
The search space here is ten to the one twenty two for mutations, so clearly you need some way to reduce the search space. Then my roommate had to go to work, but because I'm a podcaster, I had some time to take a stab at some of the ideas we discussed, and with the combination of simulating Anneeling and greedy Sarge, I think it got pretty close. I think I'm actually just a couple of swaps and shifts away from the correct solution.
What makes this puzzle. Really tricky is that there's no obvious way to escape from a local minimum. I'm afraid that this is as far as vibe coding is going to get me. But maybe you can do better. Check out the puzzle at Jane street dot com slash to our cash all right, back to elon, what.
Will Xai's business be like? Is it going to be consumer enterprise? What's the mix of those things that's going to be It's just going to be similar to other labs for this, your saying lab makes sense.
Corporations goes dan slicing corporations. Those champions do not pay for themselves exactly. But yeah, what's the business model? What are the revenue streams In a few years time? I think things are going to change very rapidly. Like I'm staying in the obvious here, you know, I called Ai the supersonic tsunami a level literation. So really, what's going to happen is, especially when you have humanoid robots at scale, is that they will just provide, They'll make products and
provide services far more efficiently than human corporations. So amplifying the productivity of human corporations is simply a short term thing. So you're expecting fully digital oral corporations rather than like SpaceX becomes part AI, and I.
Think there will be digital corporations. But it's look, is this some of some this is going to sound kind of doomerish, Okay, but I'm just I'm just saying what I think will happen. It's not it's not meant to be doomrish or anything else, just just like this is what I think will happen. Is that is that pure AI corporations that are purely AI and robotics will vastly
outperform any corporations that have people in the lip. So you can you can think of say like like like computer used to be a job that humans had, that you would go and get a job as computer where you would do calculations, and that have like entire skyscrapers full of humans, like you know, twenty thirty floors of humans just doing calculations. Now that entire skyscraper of humans doing calculations can be replaced by a laptop with a spreadsheet.
That's sheet can do vastly more calculations than in a entire building full of human computers. So then you think about okay, well, what if only some of the cells in your if some of the cells in your spreadsheet were calculated by humans, Actually, that would be much worse than if all of the cells in your spreadsheet were
calculated by the computer. And so really what will happen is the pure AI, pure robotics corporations or collectives will far outperform any corporations that have humans in the loop. And this will happen very quickly.
Speaking of closing the loop, sorry optimus, you I mean as far as like manufacturing targets and so forth go, your companies have sort of been like carrying American manufacturing of heart attack on their back. But in the field that you are, you know, Tesla has been dominant in you're and now you want to go into humanoids. In China, there's entire dozens and dozens of companies that are doing this kind of manufacturing cheaply and at scale uh and
are incredibly competitive. So give us sort of like advice or a plan of how America can build the humanoid armies or you know, the evs et cetera, at scale and as cheaply as as China is on tract to.
Well there they are really only three hard things for human robots, the real world intelligence, the hand and scale manufacturing. Yeah, so I haven't seen any even demo robots that have a great hand, like with all the degrees of freaking of a human hand, But optimists will have that. Optimist does have that?
And how do you achieve that? Isn't just like right torque definity the motor like what is the what is the hardware bottle?
Like? To that, well, we have to where to design custom custom actuators basically custom design motors, gears, uh, caro electronics, controls, sensors. Everything had to be designed from physics first principles. There is no supply chain for this and will be able to manufacture those at scale? Yes, is anything hard to accept?
The hand from a manipulation point of view, or once you've solved the hand, are you you good?
From an electro mechanical standpoint, the hand is more difficult than everything else combined. The human hand turns out to be quite something. But but then you also need the
real world intelligence. So the intelligence that tells us developed for the car applies verywell to the roblot, which is you know, primarily vision in but the car takes and more vision, but also it actually also is listening for sirens, it's you know, it's taking the inotural measurements, it's GPS signals, a whole much of other data, combining that with video,
its primarily video, and then outputting the control commands. So like like YOURTELA is taking in one and a half kickobytes a second video and outpointing two kilobtes a second of control control outputs with the video thirty six hurts and the control for CONIAD eighteen. One intuition you could have.
Four when we get this robotic stuff is that it takes quite a few years to go from the compelling demo to actually being able to use in the real world. So ten years ago you were really compelling demos of self driving, but only now we have ROBOTAXI and Weimo and all these services scaling up. Doesn't this Shouldn't this make one pessimistic on say, household robots, because we don't even quite have the compelling demos yet. I say they're really advanced.
Hand well, we've been working on human to robots now for a while, so I guess spend five or six years or something like that, and and a bunch of things that we've done for the car are applicable to the robot. So we'll use the same testa AI chips in the in the robot as the car. We'll use it the same basic principles. It's very much the same AI. You've got, you know, many more degrees of freedom for
a robot than you do for a car. But really, if you just think of for like as like a boot stream, AI is really mostly compression and correlation of
two butt streams. You're you know, so if for video you've got to do a tremendous amount of compression and and and you've got to do the compression just right, you've better compress the like ignore the things that don't matter, and and like you don't care about the details of the leaves and the tree on the side of the road, but you care a lot about the the road signs and the traffic lights and the pedestrians and even where you know somebody in another cars is looking at you
or not looking at you, like there's some of these some of these details matter a lot. So if it is essentially it's going to turn that with the car's going to turn that one and a half gigabytes a second ultimately into two kilobytes second of control outputs, so many stages of compression, and you got to get all those stages right and then correlate those to the correct control outputs. The robot has to do essentially the same thing. And you think about what what humans? This is what
happens with humans. We really are photons in controls out, so that that is the vast majority of your your life has been vision photons in and then motor controls out.
Naively, it seems like between humanoid robots and cars, the fundamental actuators in a car are like how you turn and how you accelerate, et cetera. And in a robot, especially with maneuver arms, there's dozens and dozens of these
degrees of freedom. And then especially with Tesla, you had this advantage of like you had millions and millions of hours of human demo data collected from just the car being out there, where like you can't equivalently just deploy optimists that don't work and then get the data that way.
So between the increased degrees of freedom and the far sparser data, yes, let's go, how will you use the sort of Tesla engine of intelligence on to train the optimist mind. Now you're you're actually you're highlighting an important limitation and difference between cars is like we we do have oh we'll goon have like ten million cars on the road, and so uh, that's it's hard to duplicate
that like massive training flight flywheel. For the robot. What we're going to need to do is build a lot of robots and put them in kind of like an Optimist academy so they can do self play in reality. So we're actually we're actually pulling that out, so we can have at least ten thousand Optimus robots, maybe twenty or thirty thousand that can do that, that are doing self play and testing different tasks. And then the Tellesla has quite a good reality generator, like the physics accurate
reality generator that we made this full of cars. We'll do the same thing for the robots. And I actually have done that for the robots. So you have a few tens of thousands of humanoid robots doing different tasks, and then you've got you can do millions of simulated robots in the simulated world, and you use the tens of thousands of robots in the real world to close the simulation to reality gap, close the some material gap.
How do you think about the synergies between Xai and Optimists, given you're highlighting, look, you need this world model. You maybe want to use some really smart intelligence. Is the control player and so maybe GROC is like doing the slower planning and like the motor policy.
Is a little lower level. Yeah, what will the sort of synergy between these things be? Yeah, so you just GROCK would orchestrate the behavior of the Optimus robots. So le, let's say you wanted to build a factory the then Optimists, and then GROC could UH organize the optimost robots, give them a sign them tasks UH to build the factory for to produce whatever you want. Don't you need to merge x Ai and Tessa then because these things end up?
So what are we saying only about all the company discussions? Well, we're one, we're Gonnas in the elon.
What are you waiting to see before you say we want to manufacture one hundred thousand Optimists?
Is it like Optimi? So since we're defining the proper nound, we could define the pearl of the prop and noun too. So we we're going to prop a dound the plural and so it's optimize.
Okay, is there something on the hardware side do you want to see? Do you want to see better actuators or is it just you want the software do better? Whatever you're waiting for before we get like mass manufacturing Gen three.
Now we're moving towards that, we're going forward with semester manufactory. But you think current current hardware is good enough that you are going to you should you just want to deploy as many as possible now, I mean it's very hard to scale up production. But uh yeah, but I think Optimus three is the right version of the robot to you know, to produce maybe something on the order
of like a million units a year. I mean you do want to go to Optimus four before you went to ten millions a year, okay, but you can do a million year at Optimist three. Uh yeah, I mean it's very hard to spillet manufacturing. Yes, so like manufacturing. Uh, like the the output period of time is always follows at s cove, so it's at agonizingly slow. Then has this sort of the eventually expecial increase then linear than a than a you know logarithmic outcome to you sort
of eventually as it's a number. But optimists initial production will be it's going to be a it's going to be a stretched out s cove because so much of what goes into Optimus is brand new. There's not an existing supply chain. As I mentioned, the actually is like trying to everything in the office robot is designed for physics first principles, it's not. It's not taken from a catalog. These these are custom designed everything, literally everything. I don't
think there's a single thing I done. How far down the does aco I mean, I guess we're not making custom capacitors yet maybe, but but there's there's there's nothing you can pick out of a catalog at any price. So so it just means that the the the optimistes cove, uh the units units per awkward perion of time, how many optics robusts you make per day. Whatever is is gonna initially wrap slower than a product where you have an existing supply chain, but it will get to a million.
When you see these Chinese humanoids, like Industry or whatever sells humanoidsed for like six K or thirteen K, do you just like are you hoping to get your optimisms bill of materials below that price so you can do the same thing, or do you just think qualitatively they're not the same thing, Like what do you think is going?
Like what allows it? What allows them to self for solo? And can me match that? Well, optimists, Our optous is signed to have a lot of intelligence and to have the same electro mechanical texterity, if not higher than a human. So untry does not have that. And it's also i mean, it's it's quite a it's quite a big robot. It's just it's this matter it has to do, you know, carry heavy objects for long periods of time and not
overheat or exceed the power of its sactuator. So so we've got we've got, we've got you know, it's five eleven, you know, it's pretty tall, and it's it's got a lot of intelligence. So it's going to be more expensive than a small robot that is not intelligence but more capable. Yeah, a lot more. I mean, like we think is over time, as often Showbus often robots, the class will drop very quickly.
And what will these first billion optimists of some eye, Yeah, do like bottle of their highest and best use.
Be I think that you would start off with with simple tasks that you can count on them doing well. But in the home or in factories, like the best useful robots in the beginning won't be anything any continuous operations, only twenty four VI seven operation because then you're because then they can work continuously.
Yeah, what fraction of the work in a giga factory that is currently done by humans going to gend three do?
I'm I'm not sure. Maybe it's like ten bay more, I don't know it's it. We would, we would use We would not like reduce our headcount. We would we would for sure it can increase our headcount to be clear, but but we would increase our output. So the the units produced per human, like total to total number humans at teston will increase, but the the output of robots
and cars will increase, will increase disproportionate like much much. Yeah, number of cars and robots produced per human will increase dramatically, but number of humans will increase as well.
We're talking about Chinese manufacturing a bunch here, and we're also talking about you know, we've talked about some of the policies that are ele US.
Like you mentioned the the solar tariffs.
Yeah, and you think they're a bad idea because you know, we can scale up so over in the US.
Well, just electricity output in the US needs to scale up, right, you can with thus like goodsources to get it somehow. Yeah.
But where I was going with this is if you were in charge, if you were setting all the policies, what else would you change?
So you change the solar tariffs, Yeah, I would say anything that is limiting factor for electricity needs to be addressed, provided it's not like very bad for the environment.
So presumably some permissing reforms and stuff as well, we'll be in there.
Yeah, there's a fair bit of permitting reforms that are happening. A lot of the permitting is state based, so but anything but but but this this administration is is good at removing permaning roadblocks. And I'm not saying all tariffs are bad. I'm just saying because so so Yeah, yeah, I mean sometimes if like if another country is subsidizing the output of something, then then you have to have countervailing tariffs to protect dovestic industry against some see is
by another country. What else would you change? I don't know if there's not much that the government can actually do well.
One thing I was wondering is it seems like the for the policy goal of creating a leaves for the US versus China, it seems like the export bands have actually been quice impactful or China is not producing leading edge chips and the export bands really by there China is not producing leading edge turbine engines. And similarly, there's a bunch of export bands that are relevant there on
some of the metallurgy. Should there be more export bands, like as you think about things like when there are now the drone industry and things like that, But is that something they should be considered?
Well, I think it's imporant to appreciate that in most areas China is very advanced to actually there's only a few areas where it is not. The China is a manufacturer of horhouse next level like people don't it must be it's very impressive. Yeah, yeah, I mean, if you if you take like refining of ore, i'd say roughly, China there's more just twice as much or refining of on average as the rest of all combined. And I think there's there's some areas, like say refining gallion which
goes into solar cells. I think there are like ninety eight percent of gallum refining. So so China is actually very advancing manufacturing in i'd say most areas.
It seems like we're like there is discomfort with this supply chain dependence and the ash. Nothing's really happening on it. Supply chain, supply chain. It depends on say like the Galleon refining that you're saying.
Yeah, yeah, there's there's there's there's a there's all the rare rare earth stuff, and yeah, where else we're sure, as you know, not rare like we actually do rare earth or mining in the US. Send the the the rock, put it on a train, and then from the boats China does another train and that goes to the where with refining refiners in China who then refine it, put it into a magnet, put it in just a modus
ofb assembly, and then set it back to America. So the thing we're really missing a lot of ore refining in America. And isn't this worth a policy intervention? Yes, well, I think there are some things being done on that front, but we kind of need optimists frankly to build our refineries. So sorry, you think the main advantage is China has is the abundance of skilled labor and that that that's like, that's the thing optimist fact says. But they also we need four times our populations.
But we need so I mean, there's this concern if you think that humans are the future, that like, okay, right now, if it's the skilled labors for manufacturing that's determining who's who can build more humanoids. You know, China has more of those and manufactures more humanoids, therefore it gets it gets the optimized future first.
Well, it just like keeps that eximential going.
It seems like you're sort of pointing out that sort of getting to a million optimize, Yeah, requires the manufacturing that the optimized supposed to help us get to.
Right, you can you can close that recoursal loop pretty quickly with a small number of optimize. Yeah, so you close the close recurse of loop to help help the robots buill the robots, and then we can, you know, try to get to tens of millions of units. You're maybe if you start getting too hundreds of millions of units a year, I think you're you're going to be
the most competitive country by far. We definitely count win with just humans because China has four times of population, right, and frankly, America has been warning for so long that you know, just like a like a post sports team that's been warning for a very long time, tend to get complacent and entitled, and that's why they stop winning because it's, you know, don't work as hard anymore. So I think the frankly, it just is the average work
ethic in China is higher than in the US. So it's not just that there's four times the population, but the work, the amount of work that people put in is higher. So you can, like, you can try to rearrange the humans, but you're still one quarter of the you know, it's assuming that that productivities, that health is the same, which I think actually it might not be. They China might have advantage on productivity per person. We will do one quarter of the amount of things as China,
so we can't win on the human front. And our worth rates spent low for a long time, so our birth rates spent the US birth rates spent low replacement since roughly nineteen seventy one, so we've got a lot of people retiring or you know, more people dying than we're close to sort of more people domestically dying than being born. So we definitely can't win on the human front, but we might have a shot at the row wrote front.
Are there other things that you have wanted to manufacture in the past but they've been too labor intensive or too expensive that now you can come back to that and say, oh, we can finally do the whatever because we have optimists. Yeah, I think we'd like to do more both more or refineries at Tesla. So we just completed construction and have begun liftium refining without litium refinery and Corpus Christi, Texas. We have a nickel refinery which
is called the Cathode. Uh that's here in Austin. And these are these are the largest is the largest cathode. There's the largest cathod refinery, largest liftingum refinery, the largest nekelink and lifting refinery outside of China. And it's like they know, the Cathote team would say, like, we have the largest and the only actually cather refinery in America. Many superatives not just the just but it's also the moly so it was pretty big, even though it's the
only one. But I mean there are other things that you know, you could do a lot more refineries and and help the the help America being more competitive on refining capacity. So so that there's like there's basically a lot of work for the Optimity to do that that most Americans, very few Americans, frankly wanted to. I mean, i've I've Actually it's the refining work too dirty. It's it's not it's actually, no, we don't. There's not we
don't have toxic commissions from the refinery or anything. The Catholic make refinery is right right sort of in Travis County. Like why can't I mantes from why can't you with humans? No? You can you find out of humans? I see, Okay, yeah, Like no matter what you do, you have one quarter number of humans in America in China. So if you help them do this thing, they can do the other thing. So so then then well, how do you how do
you build this refining refining capacity? Well you can do it with the Optimaya and not many, not very many, not very many Americans. Thanks for listening. See you in the next episode.