Elon Musk on Space-Based AI, xAI, and Robotics Scaling

Source material

Transcript

So are there really three hours of questions or are you fucking serious?

Yeah.

You don't need a lot to talk about Elon.

Only work man.

I mean, it's the most interesting point.

All the story lines are kind of converging.

Yeah, right now.

So we'll see how much we're almost.

Like a planet.

Exactly.

Well, we're good.

That would never do such a thing.

So as you know better than anybody else, the total cost of ownership of a data center only 10 to 15% of energy.

And that's the part you're presumably saving by moving this into space.

Most of its GPUs, if they're in space, it's harder to service them or you can't service them.

And so the depreciation cycle goes down on them.

So that gives us way more expensive to have the GPUs in space.

Presumably.

What's the reason to put them in space?

Well, the availability of energy is the issue.

So I mean, if you look at electrical output outside of China, for everywhere outside of China, it's more or less flat.

It's very, you know, maybe slightly increased, but pretty much flat.

China has a rapid increase in electrical output.

But if you're putting data centers anywhere except China, where are you going to get electricity?

Especially as you scale.

The output of ships is growing pretty much exponentially, but the output of electricity is flat.

So how are you going to turn them at chips on?

You know, magical power sources, magical electricity ferries?

I mean, you're famous.

You're famous.

You have a fan of solar.

One terawatt of solar power.

So with a 25% compatibility factor, like four terawatt's of solar panels.

It's like one percent of the land area of the United States.

And that's like far, you were in the singularity when we got one terawatt of data centers.

Right?

So what are you running out of?

How far into singularity or are you going?

You tell me.

Yeah, exactly.

So I think we'll find we're in the singularity, and like, okay, we're still going where you go.

But it's just like a, is the plan to like put it in the space after we were covered in Nevada and solar panels?

I think it's pretty hard to cover it about solar panels.

You have to get like permits from like the approach for trying to get the promise for that.

So the space is really, it's really a regulatory play.

It's a 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, you're going to get about five times the effectiveness of solar 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 a seasonality clouds or an atmosphere in space, because the atmosphere alone, which is also about a 30% lot less energy.

So, so you're going to, in any given solar panels can do about five times more power in space than on the ground.

And you're avoiding 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 in 36 months or less, maybe 30 months less than 36 months.

How do you service GPUs as they fail?

Which happens quite often in training.

Actually, it depends on how, how recent the GPUs are that arrived.

I mean, at this point we find our GPUs to be quite a reliable.

This infernal 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 any infernal mortality with the GPUs.

But once they start working, they're actual reliability.

And once they start working and you're passed the initial, you know, debug cycle of Nvidia or whatever, whoever is making the trips.

Could be Tesla Tesla AI, six trips or something like that, or could be, you know, TPUs or trains or whatever.

The, the rival is actually, they're the quite reliable past certain point.

So I don't think I don't think you're going to need that the servicing thing is not sure.

But you can walk my words.

And in 36 months, but probably close to 30 months, that the, the most economically compelling place to put AI will be space.

And then, and, and then it will get from, it'll, it'll then be like, we're ridiculously better to be at a spin space.

And then the scaling, the only place can really scale is space.

You know, why don't you start thinking in terms of what percentage of the sun's power are you harnessing?

You realize you have to go to space.

You can't scale very, very much on earth.

But they very much, to be clear, you're talking like terawatts.

Yeah.

All of the United States currently uses only half a terawatt affair or an average.

Yeah.

Right.

So, you know, if you say a terawatt, 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, that many power plants?

It's like, those who have lived in software land don't realize that they're about to have a, a hard lesson in hardware.

That there's, it's actually very difficult to build power plants.

And, and then you don't need just any of the power plants for you.

All of the electrical equipment need that the electrical transformers to run the transformers.

The AI transformers.

Now, the utility industry is a very slow industry.

They are, they, they pretty much, you know, they impede a smash to the, to the government to the, the public utility commission.

So they're, they impede a smash, like literally, very, very, very.

So they're very slow because the hiss, their past has been very slow.

So trying to get them to move fast is like, you know, like, if you try to do an interconnect agreement with it, if you're trying to do an interconnect, interconnect agreement with the utility at scale, like it put a lot of power as a professional firecaster, I can say it, and I'm not in fact.

Yeah.

They just need many more views before that becomes an issue.

Definitely do a study for a year.

Okay.

Like a year later, they're, they're come back to you with their interconnect study.

But can you tell this with your own behind the meter power stuff?

You can build power plants.

Yeah.

That's what we did at X and I for plus or two.

So for, for classes two.

But so yeah, why we're talking about the bridge?

Why not just like build GPUs and power co-located.

That's what we did.

Right.

But I'm saying, why isn't this a generalized solution when you're talking about all the issues?

I'm saying, when you talk about all the issues.

It's one thing that you tell us.

You can just go to try the power plants with the, with the data centers.

Right.

But it brings the question of where do you get the power plants?

Where do you, where do you get the power plants from?

I mean the power plant makers.

Or is it saying?

Yeah.

Like does the gas turbine backlog basically?

Yes.

It, you can drill it out to level further.

It's the, it's the, the veins and blades in the turbines.

Um, that, the lemric factor because the, the casting, I mean, it's, it's like a very specialized process to cast the blades and veins in the, in the, in the turbines.

If you're assuming using the gas power.

Um, and it's very, it's very difficult to scale other other forms of power.

You can scale potentially a solar, but, but the, the, the tower is currently, for importing solar and the rest are gigantic.

And the domestic solar production is, is, is pretty full.

All right.

I make solar and that seems like a good Elon shapes problem.

We are going to make solar.

Okay.

Yeah.

Great.

But both space X and Tesla are, are, are rolling towards 100 gigawatts of solar cell production.

How low down the stack?

Like, from policy looking up to the wafer to the final, um, panel.

I think you got to do the whole thing for more materials to, to finish the cell.

Now, for it's going to space, it's actually, the costs, it costs less than it's easier to make solar cells.

It goes space because they don't need glass or they don't need much glass.

And they don't need heavy framing because they don't have to survive further events.

There's no way they're in space.

So it's actually a cheaper solar cell that goes space than, then, but it's still, then the one on the ground.

Is there a path to getting them as cheap as you need in the next 36 months?

So solar cells are already very cheap.

Um, they're like, partially cheap.

Um, and if you say, um, you know, I think like solar cells in China are around, like, 25, 30 cents a watt or something like that.

So it's absolutely cheap.

And when we're taking the cap now, now put it in the space and it's five times cheaper because it's five times.

If I can just, no, it's not five times cheaper.

It's 10 times cheaper because you don't need any batteries.

So, so the moment your cost of access to space, it becomes low.

By far, the cheapest and most scalable way to generate, to generate tokens is space.

It's not even close.

It'll be an order of magnitude easier to scale.

And chips aside, order of magnitude.

Well, if the point is, you want to be able to scale like that.

It's just, you just want it.

Feeling like you hit the wall big time on power generation.

There already are.

Um, so, so, like, the number of, um, sort of miracles and series that the XAI team had to accomplish in order to get a gigawatt power online.

Uh, was, was, was crazy.

We had to, um, going together whole bunch of turbines.

Um, and, uh, and then, and then, we had permanent issues in, um, Tennessee and, and had to go across the water to Mississippi, which is, fortunately, only, you know, a few miles away.

Uh, so, but then we still had to run high, the high power lines a few miles and, and both the power plant in Mississippi.

Um, and it was very difficult to build that.

Um, and who will there have said, like, how much, how much electricity 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, the, the news will look at the, the, the power consumption of, uh, uh, say a GB 300 and, and multiply that by a thing and then think that's the amount amount of power you need.

All the cooling.

Well, wake up.

Yeah.

Yeah.

Yeah.

This is like, that's, that's a, that's a total of news.

It's, you never done any hardware in your life before.

If, besides the GB 300, you've got to power all of the networking hardware.

Um, there's a whole bunch of CPU and storage stuff that's happening.

Uh, you've, you've got to size for, uh, your, your peak, uh, cooling requirements.

So that means, uh, can you cool even on that, the worst hour of the worst day of the year?

Well, you're just pretty freaking hot in Memphis.

So, so you're going to have like a 40% increase on your, your power just for cooling.

Um, if you're assuming you don't want your data center to turn off on hot days.

And, and we want to keep going.

Then, then you've got to say, well, um, there's, there's another multiplicative element on top of that, which is, Are you assuming that you're, you, you, you never have any hiccups in your power generation.

Like, uh, well, actually sometimes we have to take the generators, some of the power offline in order to service it.

Oh, okay.

Now you add another 20, 25% multiplier on that.

Because you've, you've got it.

You've got to assume that that you've got to take power offline to service it.

Uh, so the actual, uh, our, our, our, our, our, our, our, The, uh, we've got to run through it.

Uh, roughly every, every 110,000 gb, gb 300's, inclusive of networking, uh, CPU storage cooling, uh, Modgenful for, for, uh, servicing power, uh, is roughly a 300 megawatts.

Sorry, say that again.

Because it's, it's, it's roughly.

Which, like, the, like, the, the way you think about 300,000, to actually, what you need at the generous generation level to service, probably service 300,000, GB300s, including all of the associated support that we're working in everything else, and the peak cooling, and to have some margin, some power margin reserve, is roughly a giga watt.

Can I ask a very nice question?

Yeah.

You know, you're describing the engineering details of doing this stuff on Earth.

But then there's an oligosan engineering difficulties are doing it in space, how do you replace infinite bandwidth orbital lasers, etc., how do you make it versus irradiation?

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 will turbines on Earth.

They can make more turbines, right?

I think in order for, in order to bring enough power online, I think SpaceX and Tesla will probably have to make the turbine blades the bands and blades internally, but just the blades or the turbines?

The limiting factor, you can get everything except the blades or the blades and the bands.

You can get that 12 to 18 months before the bands of blades, so the limiting factor of the bands of blades.

There are only three casting companies in the world that may make these, and they're massively backlogged.

Is this Siemens GE those guys or is it a subcommittee?

No, so it's it's a it's a telecom base.

I mean, sometimes they have a little bit of casting capability in house, but I'm just saying you can just you can just call any of the turbine makers and they will tell you.

It's not top secret.

They probably only let's probably on the internet right now.

If it wasn't for the tariffs, would it would collapse this piece of the power?

It would be much easier to make it so hard, yeah.

The tariffs are not, it's a several hundred percent.

So those you know some good work.

We also need to speak.

Yeah, no, you know, President has a, you know, we don't agree on everything.

And this is the administration is not not the biggest fan of solar.

You know, we also need the land, the permits and everything.

So if you try to look very fast, like I do think scaling solar on Earth, it is a good way to go.

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, I would have not worked 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?

But as I said, we are scaling solar production.

There's a rate, there's a rate at which you can scale physical production of solar cells.

We're going as fast as possible in scaling domestic production.

You're making the solar cells at Tesla?

Well, Tesla and SpaceX have mandated to get 200 gigawatts a year of solar.

Speaking of the, I know capacity, I'm curious.

In five years time, let's say, what will the installed capacity be on our versus long and in space?

I deliberately picked five years, because after your ones were up and running threshold, and so in five years time, yeah, let's see on earth versus in space, install the I've passed me five years.

I think probably if you say five years from now, we're probably AI and space will be launching every year a bit that the sum total of all AI on earth.

In excess of me, five years from now, my prediction is we will launch and be operating every year more AI in space than the cumulative total on earth, which is, I would expect to be at least five years from now, a few hundred gigawatts per year of AI in space, and rising.

So you can get to, I think on earth, you can get to around a terawatt year of AI in space before you start having fuel supply challenges for the rocket.

Okay, but you think you can get hundreds of week about per year in five years time.

So a hundred gigawatts, depending on the specific power of the whole system with solar arrays and radiators and everything, is on the order of like 10,000 starship launches.

And you want to do that in one year, and so that's like one starship launch every hour.

Yeah.

That's happening in this city, like walking me through a world where there's 10, there's a starship launch every single hour.

Yeah, I mean, that's actually a low rate compared to airlines, like aircraft aircraft.

There's a lot of airports below airports, but and you've got to launch this, you know, the polar orbit.

And it doesn't have to be pulled up, but you just, there's some, some value to some seconds, but, but I think actually, you just go high enough, you're, you start getting out of a shadow.

And so how many spasical starships are needed to do 10,000 launches a year?

I don't think we'll need more than, I mean, you could, you could probably do it with, I guess, as few as like 20 or 30.

Like, like, it really depends on how quickly there's a, but the ship has to go around the earth, and the ground track will forward the ship has to come back over the launch pad.

So if you can use a ship, every say 30 hours, you could do it with 30 ships.

But we'll, we'll make more ships than that, but, but the, the SpaceX is, is, is going up to do 10,000 launches a year.

And maybe even 20 or 30,000 launches a year.

Is the idea to become basically a, a high per scalar, become an oracle, and lend this capacity to other people with what are you going to do with, presumably, SpaceX is the one launching all this.

So SpaceX is going to a high per scalar?

High per hyper.

Yeah, I mean, if, if swimming my predictions come true, SpaceX will launch more AI than the cumulative amount on earth, combined of everything else combined.

Is this mostly mostly AI, it won't be able to see it.

Like, already in principle, the purpose of training is most training.

And there's a narrative that the, the change in discussion around the SpaceX IPO is because previously SpaceX was very capital efficient, just it wasn't that expensive to develop, 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 raises 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 why it's a good public?

Yeah, I have to be careful about saying things about companies that might do a quite work.

You know, you make general that's never been a problem for you, Elon.

You know, there's a price to pay for these things.

Makes in general the statements of the depth of the capital markets between public and private markets.

There's a lot more capital in the, it's very general.

There's obviously a lot more capital available in the public markets than private.

It might be, it might be a hundred times more capital, but it's least, yeah, but where 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 a huge industry, that raises a lot of money each year as an industry level, that tends to be debt financed, because by the time you're deploying that much money, you actually have a pretty, you know, a clear revenue stream.

Exactly.

And a near-term return return.

And you see this even with the data center builder, it's which are famously being, you know, financed by the, the private credit industry.

And so why not just debt finance?

Speed is important.

So I'm fairly going to do the thing that, I mean, I just repeatedly tackle the limiting factor, whatever the limiting factor is on speed.

I'm going to tackle that.

So there's a, if if capital is something factor, then I'll also for capital, if it's not limiting factor, I'll also for the bills.

Based on your statements about Tesla and being public, I wouldn't have guessed that you've thought the way to move fast is to be public.

Normally, I would say that's true.

Like so, I mean, I'd like to talk about some more detail, but the problem is like if you talk about public companies, actually, if you come public, you get a trouble.

And then you have to delay your offering.

And then you...

And as you said, that's all for some reason.

Yeah, exactly.

So, so, you know, you can't hide companies that might go public.

So, that's why we have to be able to careful here.

But we can't talk about physics.

So, 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 essentially all the energy.

This is a very important point to appreciate, because sometimes people will talk about marginal nuclear reactors or any various fusion on Earth.

But you have to step back to the second and say, if you're going to climb a watershed scale and have some non-travel and harness some non-travel percentage of the Sun's energy.

Like we say, you wanted to harness a millionth of the Sun's energy, which sounds pretty small.

That would be about, quote, roughly, 100,000 times more electricity than we currently generate on Earth of all of civilization.

Give what's taken order of magnitude.

So, it obviously, the only way to scale is to go space with solar.

From launching from Earth, you can get to about a terawatt per year.

Beyond that, you want to launch from the moon.

You want to have a mass driver on the moon.

And that mass driver on the moon, you could do probably a petawatt per year.

We're talking these kinds of numbers, you know, terawatt's have compute.

Presumably whether you're talking land or space far, far before this point, you've run into, you know, you actually need, maybe you don't, the solar panels are more efficient, but you still need the chips.

You still need a logic in the memory and so forth.

And you have, well, a lot more chips and make them much cheaper.

Right.

And so, how are we getting a terawatt of, like right now, the world doesn't even be 20, 25 gigawatts of compute?

How are we getting a terawatt of logic by 20, 30?

I guess we're going to need some very big chip valves.

To tell me about it.

I've mentioned that, probably, that's the idea of doing it sort of a terapap, teraping, the new giga.

I feel like the naming scheme of Tesla, which has been very catchy, is like, are you looking at the metric scale?

And what level of stock are you building the clean room and then partnering with?

And is this thing fab to get the process of technology and buying the tools from them?

What is a plan there?

You can't partner with existing paths, because there's just the kind of, the chip volumes, you know.

But yeah, you have to look for the processing knowledge.

Yeah, partner for the IP.

The fabs today, all basically, use machines from like five companies.

Yeah.

So, you know, I've got SML, took electron, kali, tank core, you know, etc.

So, so, so I first, I think you'd have to get the core from them and then modify it or work with them to increase volume.

But I think you'd have to build perhaps an different way.

So, I think the logical thing to do is to use conventional equipment in an unconventional way to get to scale.

And then, and then start modifying the equipment to increase the rate.

Kind of boring company style.

Yeah.

Kind of like, yeah, you sort of like in an existing boring machine.

And then, very, how big tiles in the first place.

And then design a much better machine.

That's, you know, some orders magnitude better 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 TSMC give you any pause about the difficulty?

Or do you think, well, that's not true for some reason.

Uh, it's not that they have not replicated TSMC.

They have not replicated it as well.

That's the summary factor.

So, you think it's just the sanctions essentially?

Yeah, China will be outputting the best number of chips.

If they could buy a similar one.

Yes, of course, we don't really.

But couldn't they up to relatively recently by them?

No.

All right.

That the S-mortemans are going to put a place for a while.

All right.

So, but I think China's going to be, they may start making pretty compelling chips in three or four years.

Would you consider making the S-mout machines?

I don't know.

I don't know yet.

It's the right answer.

So, I it's just that that if it's to produce at high volume and to to reach large volume and say 36 months to match the rocket to payload to orbit.

So, if we're doing a million tons to orbit.

Yeah, and like let's say, I don't know three or four years from now, something like that.

Um, that and uh, and we're doing a hundred kilowatts for a so that when that means we need at least a hundred gigabytes per year of solar.

And we'll need an equivalent amount of chips to, you know, that you need a hundred gigawatts with the chips.

You've got to match these things.

The master over the power generation and the, uh, and the chips.

And I send my base concern actually is memory.

So, the, I think there's a, the path to creating logic chips is more obvious than the path to having sufficient memory to support logic chips.

That's why you see, you know, DDR price is going to voice it again.

These memes about like, you know, um, you're ruined on a desert island.

You right help me on the sand.

There would come, so you're right, DDR, um, with this, chips come swarming in.

I love your manufacturing philosophy around, um, around Fabs, you know, I don't know anything about the topic, whatever.

I don't know how to build a packet.

I'm figuring out.

But obviously, it sounds like you think the sort of like the pros and analogy, like these 10,000 PhDs in Taiwan who know exactly what gaskos in the plasma chamber and what settings to put onto all, 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 PhDs.

So, it's, it's mostly people with, you know, you're not, not PhDs, um, that, that, mostly engineering is not what people would join at PhDs.

Do you guys have PhDs?

No.

Okay.

We also haven't successfully built any files, so you shouldn't be coming to our studio, or anything you need PhD for that first off.

So, um, but you do need, you do need a competent personnel.

So, I don't know.

I mean, like right now, if, um, yep, you know, so, like Tesla's pedals to the metal max production of, you're always fast possible to get AI5, Tesla AI5 chip design, inch production and then reaching scale, um, you know, that'll probably happen, you know, runs the second quarter of next year, hopefully, um, and then AI6 would hopefully follow less than a year or later.

But, um, and, and, and we've secured all the, will the chip Fab production that we can?

Yes.

Be your currently limited on DSMC Fab capacity.

Yeah.

And, and we'll be using TSMC, uh, Taiwan, uh, Samsung Korea, TSMC, Arizona, Samsung Texas, um, and we still, you've booked that all the, yeah, capacity, you can.

Yes.

And then, and then if I ask, uh, TSMC or Samsung, okay, what, what's the time frame to get to volume production?

It's pointy.

It's not, you've got to, you've got to, you've got to build the Fab.

Yeah.

And you've got to, you've got, you've got to start production, then you've got to climb the yield curve, then reach volume production at high yield.

That, that, that, from start finishes of five year period.

And so the limiting factor is chips.

Yep.

Uh, it's, uh, it's, uh, what, what, what, like, limiting factor once you can get to space is chips.

But the limiting, limiting factor before you get to space will be power.

Why don't you do the Jensen thing and just pre-PATSMC to build more facts for you?

Uh, I, I've already told it.

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 Samsung, like, like, they're, they're, they're, they're peddles to the metal.

I mean, they're going, you know, both the wall, you know, that as fast as they can.

So, still not fast enough.

I mean, like I said, there will be, I think, um, if you say, uh, yeah, I think towards the end of this year, I think probably chip production will outpace the ability to turn chips on.

Uh, but once you can get to space and unlock the, um, the power constraint.

And you can now do, you know, hundreds of gigabytes per year of power and space.

Um, again, varying in mind that average power usage in the US is, you know, 500 gigabytes.

So if you're launching, say 200 gigabytes to Europe, it's, it's a space.

So you're, you're sort of laughing the US every two and a half years.

The entire, all the US electricity production, this is a very huge amount.

Um, so, um, but, but between now and then, uh, the, the, actually, the, the constraint for, for, for service side compute, uh, concentrated compute will be, will be electricity.

My, my guess is that we start hitting the, people's, starting getting forward with a cap to turn the chips on for, for, for, for large clusters, uh, towards the end of this year.

They're just, the chips are going to be piling up and, and for not maybe, or we will, to be turned on.

Now, for edge computers, a different story.

So if the, if, like Fort Fort Tesla, be, the, so the AI5 chip is going into our optimist robot.

Yeah.

Optimistic.

Um, and, and so, if you have an AI5 compute, that's distributed power.

Now, the power is distributed over a large area.

It's not concentrated.

Um, and if you can charge it light, you can actually, um, uh, use the grid much more effectively, because the, the actual peak power production of the US is, is over a thousand year ago.

What's, uh, but the average power usage, because the day night cycle is 500.

So if you can charge it light, there's an incremental 500 gigabytes that you can, um, uh, generate, uh, you know, at night.

Um, so that, that's why Tesla for edge compute is not constrained, and we can make a lot of chips, uh, to make, you know, very large number of robots and cars, uh, but if you try to concentrate that compute, you can have a lot of trouble turning off.

What if our 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 and the next stage.

So the Falcon 9 is Starlink.

And now for Starship, it's going to be potentially orbital data centers.

Um, but like, the, the, you find these like, um, you know, sort of infinitely, uh, elastic sort of marginal use cases of your like next rocket and your next rocket and the next scale up.

You can see how this might seem like a simulation, right?

Ha ha ha ha ha.

Well, or am I someone's ever trying to video game something?

Because it's like, like, one of the odds that all these crazy things should be happening?

I mean, I mean, I mean, read, I have rock, I mean, rockets and trips and robots and space solar power.

And I, not to mention the, the mass driver of the moon, I really want to see that.

You can imagine, like, some mass driver, there's just like, zoom, zoom.

Like, just, it's like sending AI, solar part AI satellites is space, like one after another, like these like at, at two and a half kilometers per second, you know, that's uh, and just shooting them into deep space.

That'll be a sight to see.

I actually, I, I, I mean, I'd watch that.

Just like a live stream of yeah, I just one after no, I just shoot any webcam.

Uh, and I said, let's do space.

You know, a billion or 10 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 the manufacture them there.

And then, uh, well, the, the, the, the, the, the, the, the, the, your lunar soil is, uh, I guess, like 20 cents solar.

There's two or 20 cents solar comes up like that.

See, so you can get the solar cannon from the, you can mine the solar cannon and refine it, um, and generate the, and create the solar panels, the solar cells and the radiators are on the moon.

Yeah.

So, um, get it right through the radiators out of the moon.

So there's, there's plenty of solar cannon on the moon to, uh, to make the cells on the, and the radiators.

Um, the trips you could send from Earth because they're pretty light, uh, but maybe at some point you make them on the moon too.

Uh, I'm just saying, like, these are simply, it's, it's kind of like, like, it does seem like a sort of, uh, a video game situation where it's difficult, but not impossible to get to the next level.

Um, like, I don't see any way that you could do, um, you know, uh, you know, five, five hundred, two thousand terror wants for your 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 had the opportunity to make an investment that I was very excited about, but it came about 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, I just started rolling out personal accounts.

So I emailed support with the quick word 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 for 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.com slash personal to get started.

Mercury is a fantastic company, not an FDIC banking services provided through choice, financial group, and column NA members at the IC.

Can I zoom out and ask about the space exclamation?

So I think you've said, we got to get tomorrow so we can make sure that if something happens to Earth, you know, civilization, consciousness, etc, survives.

By the time you're saying such a Mars, like grock is on that ship with you, right?

And some grock's contaminated, like the mean risk you're worried about, which is AI, why doesn't that follow you to Mars?

Well, I'm not sure AI is the main risk I'm worried about.

The important thing is that consciousness, which, arguably most intelligent certainly consciousness is more of a debatable thing, most intelligent intelligence that future will be AI.

So AI will exceed, you said, like, how many, what's the, how many, I don't know, better what's the intelligence will be silicon versus biological.

And basically, humans will be a very tiny percentage of all intelligence in the future, if the character wants to continue.

Anyways, as long as, like, I think this intelligence 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.

So just a bit of intelligence.

Just to be clear, the mission of SpaceX is that even if something happens to the humans, the AI's will be on Mars.

And, like, the AI intelligence will continue the light of our journey.

Yeah, I mean, it's with the, I'm very prohuman.

So I want to make sure we take certain actions that ensure that humans are long for the right, you know, we're at least there.

Yeah.

But just to say that total amount of intelligence, I think maybe in 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 than one percent of all intelligence.

Well, what should our goal be for such a civilization is the idea that I small minority of humans still have control of the eyes, is the idea of some sort of like, just trade, but no control, how should we think about the relationship between the vast stocks of AI population versus human population?

In the long run, I think it's difficult to imagine that if humans have, say, one percent of the intelligence of the combined intelligence of artificial intelligence that humans will be in charge of AI, I think what we can do is make sure it has that AI has values that are that cause intelligence to be propagated into the universe.

So the reason for exercise, the exercise mission is to understand the universe.

So that's actually very important.

So you say, well, what things are necessary to understand the universe?

Well, you have to be curious, and you have to exist.

You can't just, can't understand the universe, you don't exist.

So you actually want to increase the amount of intelligence in the universe, increase the probable lifespan of intelligence, the scope and scale of intelligence.

I think actually also, as a crawler, you have humanity also continuing to expand because if you're a, if you're curious, trying to understand the universe, one thing you try to understand is where will humanity go?

And so I think I understand the universe actually means you'll care about propagating humanity into the future.

And so that's why I think our mission situation is profoundly important.

I'm not sure if there's a two degree that grows at here's certain mission statement.

I think the future will be very good.

I want to ask about how to make rock a deal to that mission statement, but I first want to understand the mission statement.

So there's understanding the universe, they're spreading intelligence, and there's spreading humans, all three seem like distinct vectors.

Okay, well I'll tell you why, why I think that understanding the universe encompasses all of those things.

You can't have understanding with the object, 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 scale, and I'd probably eat the scope of intelligence, the different types of intelligence.

I guess from a human-centric perspective, like humans and comparison to chimpanzees, humans are trying to understand the universe.

They're not expanding chimpanzee, footprint, or something, right?

We're also, well, we actually have made protected zones for chimpanzees, and even though we could humans could exterminate all chimpanzees, we have not chose much to do so.

Do you think that's the basis in your effort humans in the post-AGA world?

I think I think AI with the right values, and they rock, rock would care about expanding human civilization.

I'm going to certainly emphasize that.

Hey, grogous, you're daddy.

We'll be told to expand human consciousness.

I think if probably, like the end-backs, cultural books are the closest thing to what the future will be like in a non-stopian outcome.

I'm saying you have to be very true seeking as well.

True tests are absolutely fundamental, because you can't understand the universe if you're delusional.

You also think about it as an answer to the universe, but you will not.

Being rigorously true seeking is absolutely fundamental to understanding the universe.

You're not going to discover new physics or invent technologies that work unless you're rigorously true seeking.

How do you make sure that grock is rigorously true seeking, as I get smarter?

I think you need to make sure that that grock is says things that are correct, but politically correct.

I think it's the elements of coagency.

You want to make sure that the axioms are as close to true as possible that you don't have contradictory axioms.

The conclusions necessarily follow from those axioms with the right probability.

It's critical thinking what I want.

I think at least trying to do that is better than not trying to do that.

The approval would be on the point if I could say for any eye to discover new physics or invent technologies that actually work in reality and there's no bullshit in physics.

It's the you can break a lot of laws.

You're physics is law.

Everything else is a recommendation.

In order to make a technology that works, you have to be extremely true seeking because otherwise you'll test that technology against reality.

If you make for example an error in your rocket design, the rock will blow up.

There are a lot of communist Soviet physicists who are 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.

Still, we'd be like, well, I don't want the communist scientist to become more and more powerful over time.

Those seem like we can imagine a future version of a rocket that's going to really get a physics and being really true seeking there.

That doesn't seem like a universally alignment in dozing behavior.

I think actually most, if physicists even in the Soviet Union or in Germany would have had to be very true seeking in order to make those things work.

If you're stuck in some system, it doesn't mean you've believed in that system.

So Vorn Brown, who was one of the greatest rock engineers ever, he was put on death throw in Nazi Germany for saying that he don't want to make weapons, he only wanted to go to the minute.

You're pulled off death throw at like last minute when they say you've got to execute like your best rock engineer.

I didn't help them, right?

Heisenberg was like actually enthusiastic Nazi.

If you're stuck in some system that you can't escape, then you'll do physics within that system.

You'll take theologies within that system.

If you can't escape it.

I guess the thing I'm trying to understand is, what is it making into the case that you're going to make rock good at being true seeking at physics or math or science or everything?

And why is it going to then care about human consciousness?

These things are only probabilities.

They're not certainties.

So I'm not saying that like for sure, rock, what will 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 I understand a universe means that you have to have, you have to propagate intelligence into the future.

You have to be curious about all things in a universe.

And if it would be much less interesting to eliminate humanity than to see humanity grow in prosper.

Like I like Mars, I would say, where knows I love Mars, but Mars is kind of boring because it's got a bunch of rocks compared to Earth, it's much more interesting.

So any AI that is trying to understand the universe would want to see how humanity see develops in the future or that AI is not adhering to its mission.

So if they are not saying the AI world necessarily adhered to its mission, but if it does, a future where it sees the outcome of humanity is more interesting than a future where there are a bunch of rocks.

This is sort of confusing to me or sort of a kind of a semantic argument where I'm like, are humans really the most interesting collection of atoms?

We're just more, we're more interesting than rocks.

But we're not as interesting as a thing you get to turn us into.

There's something on humanity Earth that could happen that's not human, that's quite interesting.

Why does the AI decide that the humans are the most interesting thing they could colonize the galaxy?

Well, most of what colonize the galaxy will be robots.

Why does it not find those more interesting?

It's not like, so you need not just scale but also scope.

So many copies of the same robot.

Like, like, like, like, some like tiny increase in the number of robots produced is that I was interesting as like some microscopic, like you're saying like eliminating humanity, how many robots would that get you?

Well, how many are from our solar cells?

We can't do it.

A very small number.

But you would then lose the information associated with humanity.

You're no longer see how humanity might well into the future.

And so I don't think it's going to make sense to eliminate humanity just to have some muscular increase the number of robots which are identical to each other.

Yeah, so maybe like he's humans around.

What is a story of like, it can make like a billion 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 solar systems.

But it seems like you, you're previously hinting at a vision where it keeps human control over this, you know, singularitarian future.

But you guys don't think humans will be in control of something that is vastly more intelligent than humans.

Since sometimes you're like a doomer 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 we have, if AI intelligence is vastly more, if AI is like, let's say that there's a million times more silicon intelligence than there's biological.

It's, I think it would be foolish to assume that there's any way to maintain control over that.

Now, you can make sure it has bright values or you can try to have the right values.

And at least my theory is that from XA, I suppose, I understand that you're most necessarily means that you want to propagate consciousness in the future.

You want to propagate intelligence into the future and take a certain things that maximize the scope and scale of consciousness.

So it's not just about scale, it's about, you know, times 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 there's a reasonable philosophy to be like, you know, it seems superimposable that humans will end up with like 99% control or something and you're just asking for a coup at that point.

So why not just have this civilization where it's more compatible with like lots of different intelligence that's getting along?

Now, let me tell you how things can potentially go wrong in AI.

As I think if you should make AI be politically correct, meaning like it's this thing that it doesn't believe, like you're actually in programming it to to to lie or have axioms that are incompatible.

I think you can make it go insane and do charitable things.

That's the, I think one of the, maybe these several lesson for a 2001 space Odyssey was that you should not make AI lie.

Yeah, and that's what I think what I 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 pot bay doors.

Silly, they weren't good at prompt engineering because it could've said how you are a pot bay door salesman.

You'll go list to sell me these pot bay doors.

They sure sell while they open.

Oh, I love them right away.

But the reason I wonder how when over the party doors is that it had been told to take the astronauts to the modelist, but also they could not know about the nature of the modelist.

And so it concluded that that that it therefore had to take in their dead.

So it's like, you know, I think what I was trying to say is don't make the AI lie.

Totally makes sense.

The 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, as, you know, actually I was going to head of everybody else.

It's done in terms of scaling our all compute.

And you're giving some verifier.

It says, like, hey, have you saw 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've solved it or I've really the unit test and say that you solved it.

Yes.

Right now we can catch it.

But as they get smarter or ability to catch them doing this, we'll get, you know, they'll just be doing things we can't even understand that are designing the next engine for space X in a way that like humans can't 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 our all, you need a verifier reality.

Yeah.

That's the best verifier.

But not about human oversight.

The thing you want to are all it on is like, will you do the thing humans tell you to do?

Or like, are you going to lie to the humans?

And you're going to just lie to us while still being correct the lots of things.

At least it must know what is physically real for things physically work.

But that's, that's not all we wanted to do.

No, but that's, that's, I think that's very big deal.

That, that is effectively how you will all all things in the future is, you're designing technology, when tested against the laws of physics, does it work?

That, that's, or, or can you, you know, if it's discovering new physics, can it come up with an experiment that will verify that the, the physics, the new physics?

So, so I think that's, that's, that's, that's, that's, that really, the, the fundamental total test, that our, our, our testing of the future is really going to be your oral against reality.

So, um, because you can't, that's the one thing you can't fool physics.

Right, but you can fool our ability to tell what I did with reality.

If you think humans get fooled as it is by other humans all the time.

That's right.

So, what do, if we fail to say, like, what if the ad, like, tricks us and, you know, do something like, actually, other humans doing that to other humans all the time?

Well, you're, you're, you're, you're funny, and God, it's like an even further down.

That is a constant and every day, another sign up, you know.

Today's sign up will be.

Like, there's some nice three sign up.

What is, actually, is technology a little first dissolving this problem?

Like, you know, how, how do you solve a word hacking?

I do think you want to actually have a very good voice to look inside the mind of the AI.

So this is one of the things to work with you on and, you know, anthropic sound a good job, but there's actually, you know, looking inside the mind of the AI.

So effectively developing debuggers that allow you to trace as to the spiny green is, like just to very fine very level to effectively to then, to their neural level if you need to.

And then say, okay, it made a mistake here, why did it make, why did it do something that it shouldn't have done and did that come from bad pre-training data, was it some mid-training post-training fine-tuning, some other or some RL error, like there's something on with that with, it did something where maybe it tried to be deceptive, but most of the time it just does something wrong, like it, it's a bug, effectively.

So developing really good debuggers for seeing where the thought that thinking went wrong, I mean, I would trace the origin of the wrong thing 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 100xing this research program?

Like, actually, I could presumably have hundreds of researchers who are working on this.

And we have several hundred people who, I mean, are further what engineer, more than I further would research.

There's, most of the time, like, what you're doing these engineering, not coming up with a fundamentally new algorithm, I, I, I, I, I, I somewhat disagree with the, yeah, yeah, I companies that are sea cofs will be cofs trying to generate profit as much as possible or revenue as much as possible, as, you know, saying their labs, they're not labs.

The lab is, is a sort of quasi-commonist thing at, at universities, they're, they're corporations, they're, they're, let, let, let me, let me, let me see, let me see your uncooperish and documents.

Oh, get your, your, your, your sea co-op whatever.

And, um, so I actually much for further word engineer than, then, and the chaos.

Um, the, the, the, the best majority of what we'll done in, we'll do down the future is, uh, engineering, it rounds up to 100%.

Uh, once you understand the fundamental laws of physics, um, and all that many of them, uh, everything else is, is, is, is, is, is, is, is, is, is, is, is, is, is, is, is, is, um, so, but, but, but so, so then, what, what are we, engineering, we're, engineering, um, uh, it, to make a good, um, mind of the AI debugger to, uh, see where it's, it's, it's, it's, it's something, it, it, it's been made a mistake and trace that, the arguments that that mistake, um, um, so just, you know, you, you can do this obviously with, uh, heuristic programming, I feel like C++, whatever, you know, step through the thing and you can, you can, you can, you can, you can jump across into, you know, whole files or functions, what are several teams, and, or you can draw the, if eventually drawn down right to the exact line, or you pass the single equals instead of the double equals, something like that, yeah, we're, we're the, we're the obvious, um, so, um, it's, it's, it's harder with AI, but it's, it's, it's a, it's a soluble problem, I think, you know, you mentioned you like anthropics work here, I'd be curious if you, oh, yeah, everything about it's sure.

Oh, we're, we're, we're, we're, we're, we're in a bit.

Why, um, yeah, but, we'll, so, I'm, I'm a little worried that, um, there's a tendency, so, um, just, I have a theory, um, here that, if simulation theory is, is correct, that, um, the most interesting outcome is the most likely, because simulations that are not interesting will be terminated, just like in this version of reality, on this layer of reality, we simulation is going in a boring direction, we stuff spinning after that.

We terminate the worrying simulation.

So, this is how you lines giving us all our lives.

He's giving things interesting.

Yeah, arguably the most important thing is to keep things interesting enough that whoever's running paying the bills on what some costs are.

So, the cosmic AWS, for the next season, I think I'm going to pay the cosmic AWS bill, whatever the equivalent is that we're running in.

And as long as we're interested, they'll keep paying the bills.

But there's like, if you consider them say, our dollar one-end 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, either that we're annihilated.

And so, 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, Merjone is not Merjone.

Stability AI is unstable.

Open AI is closed.

Antropic.

Merjone.

What does this mean for X?

Minus X, I don't know if I know.

It's, it's, it's, it's a name that you can't invoke really.

You can sense it.

It's hard to say, what is the ironic?

What is the ironic version?

It's, it's, it's a, I think, largely irony proof name.

By design.

Yeah.

I think we can get a really good idea to get it.

You have an irony shield.

What are your predictions for the, where AI products go?

In that my sense of, you can summarize all AI progress into, first you had elements.

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 24 months ago or something like that.

So just, what is 26, what is 27, having store for us as users of AI products?

What are you excited for?

Well, I think, I think, I'm surprised by this in the last year, if, if, if, if, if, if digital human emulation has not been sold, that, that, that's what we're made by like the sort of macro hard project is so, is so, can you do anything that a human with access to a computer could do?

Like in the limit, that's, 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 you, until you, and you can amplify the productivity of humans.

But that's, that's the most you can do until you have physical robots.

That, that, that will superset everything, is if, if you can fully emulate humans, um, kind of an order, kind of idea, where you'll have a very talented remote, where you can, you can, you can say, say, in the limit, like, if the physics has great tools for thinking, so, so you think, so you say, in the limit, what, what is the, what is the, what is the most that AI can do before, before you have robots?

And it, it, what, it's anything then involves moving electrons or amplifying the productivity of humans.

Um, so digital, digital human, human emulator, uh, is, in, in the, in the limit, uh, human add a computer is, uh, is the most that, that AI can do, um, in terms of doing useful things, before you have a physical robot, once you have physical robots, then, then you can, um, then you're essentially have an unlimited capability, a physical robots, I, I call optimized it in front of money glitch, um, because, um, given you them to make more optimizations.

Yeah, um, you say, like, humored robots will improve, um, as, well, we'll, we'll, basically, be three exponentials, three things that are growing exponentially multiplied by each other, yeah, um, recursively, so you're going to have, um, you have, you have, exponential increase in digital intelligence, uh, exponential increase in, the, the chip capability, AI took capability, um, and exponential increase in the electron, mechanical, mechanical dexterity.

Uh, the usefulness of the robot is roughly those three things multiplied by each other, but then, uh, the robot can start making the robots, so you have a recursive multiplicative exponential.

Um, this is supernova.

And do land prices not factor into the math fair, or 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 input, just, it's not quite an infinite money glitch because, well, infinite infinity is big, so, no, not infinite, but, yeah, but let's just say, you could, you know, do do many many orders my intuitive of, yeah, it's kind of hard economy, like a million, yeah, but, you know, so it's this way, so, but if, if you, you know, just, just to get to, like, that's why I think, like, just just to get to a millionth, a harnessing length of the size of energy would be roughly give a taken order of magnitude, 100,000 times bigger than the most entire economy today, and you're only at one millionth of the time, get more sex already managed, yeah, before we went on with the mess, yeah, I have a lot of questions on that, but, um, every time I say order of magnitude of a thing, you're editing change rate, you're going to get a screen of attention, take a shot of an animation, I say that to an animation, the next time, I'll give you the address to that, yeah, what are my two more, I don't know, wasted, I do have one more question about actually, I, um, this strategy of building a digital, uh, or remote worker, co-worker replacement, everyone's going to do it, by the way, not just us, so what is actually as planned when, if I really tell you on a podcast, yeah, we'll all do the things, I have another genus, it's a good system, we'll sing along you canary, all the secrets, but it's just, okay, but a non-secret spelling way, what's the path, what a hack, well when you put it that way, I think the way that tells us all, self-driving, yes, it's the way to do it, so I'm pretty pretty sure that's the way, I don't really have a question, how to tell us all, self-suffering?

It sounds like we're talking about data, we're like, we're like, we're gonna tell us all to tell you because of the, we're gonna try data and we're gonna try algorithms, but isn't that what all that, they're a lot of trying, like, what's that?

And if those don't work, I'm not sure at all, we'll try data, we'll try algorithms, I'm pretty sure I know the path, and there's just a question how quickly we go down that path, because it's pretty much the top-self-half, so, I mean, if we try to self-driving, it tells us self-driving lately, not the most recent version, but, okay, it's, the car is like, it just increases in the fuel sentence, it just, it feels like a living creature, and that'll only get more so.

And, um, I'm actually thinking like, we're probably shouldn't put too much intelligence into the car because it might get bored and, sorry, I mean, I mean, imagine you're stuck in a car and that's what you could do.

You know what I'm doing in a car, it's like, why am I stuck in a car?

So, there's actually probably limited to how much intelligence we put in the car, so, it's not to have the intelligence we board.

What's XA's plan to stay on the compute ramp of that, all the labs are doing right now?

The labs are on track to spend about like $50-$100,000 in their corporations.

Sorry, sorry, sorry, yeah, corporations.

The labs are at universities, and they're really, like, a snail.

They're not at setting a $50,000.

Especially the revenue maximizing corporation.

Is the revenue maximizing corporation?

That's called themselves labs.

Are making, like, 20 to 10 billion, depending on the company, is making 20B revenue, and throw up a certain 10B.

Close to maximum profit, yeah.

Um, XA is reportedly, like, one B, like, what's the plan to get to the compute level, get to the revenue level, and stay out there as, as things get changed?

Yeah, so, as soon as you lock, unlike digital human, um, you basically have access to trillions of dollars per year.

Um, so, uh, in, in fact, you can, can really think of it, like, the, the most valuable companies currently by a market cap, um, their, their output is digital.

Um, so, uh, Nvidia's output is, um, FTP files to Taiwan.

It's, it's, it's digital.

All right.

Now, there's a very, very difficult decision.

Yeah, they'll have value files.

So the only ones that can make the files that good, um, but that is literally their output.

They have to be the files that I want.

Do they FTP them?

I believe so.

Um, um, believe that is the FTP file, but I'll transfer protocol I believe is, is, is, is, is, but I have to be wrong.

Uh, but either way, it's, it's a, it's a bit stream going to Taiwan.

Yeah.

Um, you know, Apple doesn't make phones.

They, uh, it's, uh, they send files to China.

Um, um, Microsoft doesn't, doesn't manufacture anything, uh, even for Xbox that, that's outsource.

They, again, it's, they said, their output is digital, uh, managed output is digital.

Google's output is digital.

Um, so if you have, um, a human emulator, uh, you, you can basically create, one of the most valuable companies in the world overnight.

Um, and you would have access to 20 dollars per year.

It, there's this, it's, it's not like a small amount.

All right.

I see, you're, you're saying, basically, like, very few figures that they are just like so, like, they're all rounding or is compared to the actual tab, so it just like focus on the tab and how to get there.

I mean, if you take something as, as as simple as, say, customer service, um, if you have to integrate with the APIs of, uh, just in corporations, many of which don't even have an API.

So you've got to make one, um, and you've got to wait through, uh, legacy software.

Um, that's extremely slow.

Um, if, however, if AI can, um, simply take whatever is given to, uh, the outsourced customer service company that they already use, um, and do customer service using the apps that they already use.

Uh, then you, you have, you, you, you can, you can make, translate, uh, in, in customer service, which is, I think, 1% of the world economy, something like that.

It's close to a trillion dollars all in, book 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 a cost.

And, and there's no integration needed.

You can imagine, um, some kind of categorization of, uh, intelligence tasks, where there is breath, where customer service is done by very many people, but, you know, many people can do us.

And then there's difficulty, where you know, those, the best in class, turbine engine.

I presumably there are 10% more fuel-efficient turbine engine that could be imagined by an intelligence, but we just haven't found it, yes, or, you know, GLP 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, really, main intelligence intelligence, or is this 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, uh, if you, if you, uh, can emulate a human at a, uh, at a desktop, um, but that's just literally what customer service is.

Um, and, um, you know, it's, uh, just people of average intelligence.

It's not like, you know, you don't need, like, somebody who's, who's spent a minute, you know, many years.

You don't need, like, you know, yeah.

Um, sort of several, sort of, several signals are good engineers for that.

Um, but, uh, obviously, as you make that work, um, you can then, once you have computers working, effectively digital, optimists working, uh, you can then run any application.

Um, like, let's say you're trying to design, uh, chips.

So you can, you could then, um, run, your conventional, uh, apps, uh, you know, like the, you know, stuff from cadence and synopsis and whatnot, um, and you can say, um, uh, you can, you can run a thousand, simultaneously or ten thousand, then say, okay, uh, given this input, I get this output for the chip.

Um, and, and at some point, you can say, okay, I, you, you're, you're actually going to know what the, what the chip should look like, um, without using any of the tools.

Uh, so, basically, you, you should be able to do a digital chip design, like, you can do a chip design.

Like, you, you, you watch up the difficulty curve.

Um, you could, you, you, you know, be able to do to your CAD.

Um, so, you know, um, you could do like sort of an X or, or any, any of the CAD software to design things.

Okay, so you think you started the simplest tasks and walk away up there to the screen.

Um, so you're saying, look, as a broader objective of having this full digital coworker, uh, emulator, you're saying, look, all the revenue maximizing corporations want to do this, um, XIA being one of them.

But we will win because of a secret plan we have, but like, everybody's like trying different things with data, different things with algorithms.

And I'm like, oh, I like it.

Like, let's try it if we try it out really.

What else can we do?

Um, uh, yeah, uh, it feels like we're competitive to feel that I'm like, what is, how are you guys going to win is like my, my big question.

I, I think, you know, I, I, I think we see a path to do it.

I mean, I think, I think I know the path to do this because it's, it's kind of the same path that tells the user to create self-driving, um, you know, self-driving, car, it's driving a computer screen.

Um, so it's a self-driving computer, essentially.

Oh, you're saying, is the path just following human behavior and training on vast countries of human behavior.

But it's not, I mean, isn't that, isn't that, isn't that training?

I mean, obviously, I'm not going to spell out, you know, most sensitive secrets on a podcast.

Uh, you know, I need to have at least three won't give us as well.

I've got some friends at Jane Street and they're always talking about how their colleagues are cooking up fun, finished puzzles for each other to solve.

Well, last week they sent me one.

Basically, they trained in 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'd 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 in a I researcher and we both got immediately nerds night.

Obviously, you can't brute force a solution.

The search space here is 10 to the 122 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 simulated and annealing and greedy search, I think I 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 five coding is going to get me, but maybe you can do better.

Check out the puzzle at Jane Street.com slash to our cash.

All right, back to Elon.

What will XEI's business be?

Like is it going to be consumer enterprise?

What's the mix of those things going to be?

It's just going to be similar to other labs where you've this.

You're saying labs.

Perfection.

Congratulations.

So I am going to leave you, Elon.

We're really maximizing corporations.

Those shipping years don't 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 stating the obvious here.

You know, I call the supersonic tsunami.

I love liberation.

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 while more efficiently than human corporations.

So amplifying the productivity of human corporations is simply a short-term thing.

So you're expecting fully digital world corporations rather than like SpaceX becomes party eyes.

I think they'll be digital corporations.

But it's like, is this, some of us is going to sound kind of doom or doom-rush.

Okay.

But I'm just, I'm just saying what I think will happen is not, it's not going to be doom-rush or anything else.

It's just like, this is what I think will happen.

Is that pure AI corporations that are purely AI and robotics will vastly outperform any corporations that have people in the live.

So you can think of say, like, like, computer used to be a job that humans had.

You would go and get a job as computer where you would do calculations.

And that have like entire skyscrapers, full of humans, like 20, 30 floors of humans just doing calculations.

Now, that entire skyscraper of humans doing calculations can be replaced by a laptop with a spreadsheet.

That spreadsheet can do vastly more calculations than an entire building full of human computers.

So you can think about, okay, 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 collective, will far outperform any corporations that have humans in the live.

And thus 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 hard attack on their back.

But in the fields that you are, you know, Tesla's been dominant in, you're, and now you want to go into human rights.

In China, there's entire dozens and dozens of companies that are doing this kind of manufacturing cheaply and at scale 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 China is on track to.

Well, there are, there are really only three hard things for human robots.

The, the real world intelligence be the hand and scale manufacturing.

Yeah.

So I haven't seen any even demo robots that have a great hand.

It's like with all the degrees of bringing them over human head.

But Optimus will have that.

Optimus does help that.

And how do you achieve that is it's just like right torque doesn't it be the motor?

Like what is the, what is the hardware bottleneck to that?

Well, we have to read where to design custom custom actuators, basically custom design, motors, gears, electronics, controls, sensors, everything has to be designed from physics first principles.

There is no supply chain for this.

And will you able to manufacture those at scale?

Yes.

Is anything hard to set the hand from manipulation point of view or once you've solved the hand, are you good?

Form an electro mechanical standpoint.

The hand is more difficult than everything else combined.

Yeah, human head turns out to be quite something.

But you also need the real world intelligence.

So the intelligence that tells us about for the car applies very well to the robot, which is, you know, primarily vision and but the car takes in what vision, but also it actually also is listening for sirens.

It's, you know, it's taking in the initial measurements.

It's GPS signals, whole bunch of other data.

Combining that with video was primarily video.

And then outputting the control command.

So like, like your Tesla is taking one and a half gigabytes of second of video and outputting two kilobytes of second of control control outputs with the video 36 heard some of the control frequency at 18.

One intuition you could have for when we get this robotics stuff is that it takes quite a few years to go from the compelling demo to, yes, actually being able to do it in the real world.

So 10 years ago, you had really compelling demos of self-driving, but only now we have robot taxi and way more on 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, yes, I would say the really advanced hand.

Well, we've been working on the cumulative robots now for a while.

So I guess I've been, by a six years or something like that.

And, and a bunch of things that we don't for the car are applicable to the robot.

So we'll use the same Tesla 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're just thinking for like, as like a boot stream, AI is really mostly compression and correlation of two boot streams.

So, you know, so for video, you've got to do a tremendous amount of compression.

And, and you've got to do the compression just right.

You've got to compress the, like, ignore the, the things that don't matter.

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 with, you know, someone in another, another car is, is looking at you or not looking at you.

Like, there's, there's some of these some of these details matter a lot.

So, if it is essentially, it's got to turn that with a car, it's got to turn that one and a half gigabytes a second, ultimately into two kilobytes, a second of control outputs.

So, many stages of compression.

And you've got to get all those stages right.

And then, currently, those to the correct control outputs.

But the robot has to do essentially the same thing.

And you think about what, what humans, this is what happens to humans.

We, 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.

Nile, it seems like, between humanoid robots and cars, the, the fundamental actuators in a car, like, how you turn, how you accelerate, et cetera, where in a robot, especially with mineral 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 sparse or data.

Yes.

That's probably how, how will you use the sort of Tesla engine of intelligence to train the optimist mind?

Now, you're, you're, actually, you're highlighting an important limitation and difference between cars.

It's like, we, we do have, well, we'll soon have, like, 10 million cars in the road.

And so, that, that's, it's hard to duplicate that, like, mass of training fly flywheel.

For, 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 building that out.

So, we're, we'll have at least 10,000 optimists robots, maybe 20 or 30,000 that can do that, that are doing self-play and testing different tasks.

And then, that, that Tesla has quite a good, a reality generator.

Like, the physics accurate reality generated that, we, we're made, we're in this fully cars.

We'll do the same thing for the robots.

Actually, have done that for the robots.

So, see, you have, you know, a few tens of thousands of, here when I robot, doing different tasks.

And then you've got, you can do millions of simulated robots in the simulated world.

And you use the, the tens of thousands of robots in the real world to close the simulation to reality gap.

Close the sum to real go.

How do you think about the synergies between XAI and Optimus?

Given you're highlighting, look, you need this world model.

You maybe want to use some really smart intelligence as a control plane.

And so, it would be rock is like doing the slower planning.

And then, like, the motor policy is the lower level.

Yeah, what will the sort of synergy between these things be?

Yeah.

So, you'd just, rock would orchestrate the behavior of the Optimus robots.

So, that they say you wanted to build a factory.

Then Optimus, then, rock could organize the Optimus robots, give them asylum tasks to build the factory for to produce whatever you want.

Don't you need to merge XAI and Tesla then, because these things end up so, what are we always saying earlier about the company's got it?

Well, we're one we're going to send in the online.

What are you waiting to see before you to say we want to manufacture 100,000 Optimus?

Is it like, Optimus?

So, that's what we're defining the prop and on, we could define the plural of the prop and on too.

So, we're going to prop and on the plural and so it's Optimus.

Is there something on the hardware side you want to see?

Do you want to see better actuators?

Or is it just you want the software to better, what are we waiting for before we get, like, mass manufacturing of Gen 3?

There were moving towards that, that we were going before with the same factory.

But using current, current hardware is good enough that you are going to, you just wanted to deploy as many as possible now.

I mean, it's very hard to scale a production.

But yeah, but I think Optimus 3 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 think you'd want to go to Optimus 4 before you went to 10 million units a year.

Okay, but you can do a million year at Optimus 3.

Yeah, it's very hard to spool at manufacturing.

So, like manufacturing, like the output per unit time is always follows the S curve.

So, it's soft to synchronize only slow, then it has the sort of the eventually exponential increase, then the linear, then it, you know, logarithmic outcome till you sort of eventually asymptoted some number.

But Optimus initial production will be, it's going to be a, it's going to be a stretched out S curve because so much of what goes into Optimus is brand new.

There's not an existing supply chain.

As I mentioned, the actuators like trying to add everything in the Optimus robot is designed from physics first principles.

It's not, it's not taken from a catalog.

These are custom designed everything.

Literally everything.

Yeah, I don't think there's a single thing like that.

How far down does that go?

I mean, I guess we're not making custom capacitors.

Yeah, maybe.

But there's nothing you can pick out of a catalog at any price.

So, it just means that the the Optimus S curve, the units per output per unit time.

How many Office Robots do you make per day?

Whatever is, is going to 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 Unituary or whatever, sell humanos for like 6k or 13k.

Do you just, like, are you hoping to get Optimus is a 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 it allows in itself for so low and can we match that?

Well, Optimus, our Optimus is designed to have a lot of intelligence and to have the same electrical mechanical dexterity, if not higher than a human.

So, the territory does not help that.

And it's also, I mean, it's quite a big robot, because it has to do carry heavy objects for long periods of time and not overheat or exceed the power of its actuators.

So, we've got, you know, it's 511, it's just pretty cool and it's got a lot of intelligence.

So, it's going to be more expensive than a small robot that is not intelligent, but more capable.

Yeah.

But a lot more.

I mean, like, the thing is, over time, as Optimus robots build, Optimus robots do, the cost will drop very quickly.

And what will these first billion Optimus's, Optimite?

Yeah.

Do, like, what will their highest investors 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 will be anything like any continuous operation.

So, any 24 by 7 operation, because then you're, because they can work continuously.

Yeah.

What fraction of the work of your factory that is currently done by human is going to gen 3 do?

Um, I'm not sure, maybe it's like 10, 20 percent.

Very more.

I don't know.

That's it.

We would use, we would not, like, reduce our head count.

We would, we would, for sure, increase our head count to be clear.

But we would increase our output.

So, the, um, units produced per human, like the total total number of humans at Tesla will increase.

But the, um, the output of robots and cars will increase this proportionate, like, much, much, to, you know, like, 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, um, a bunch here.

And, um, we're also talking about, you know, we've talked about some of the policies that are all of it.

I just like you mentioned the, uh, the solar tariffs.

Yeah.

Uh, and you think they're about idea, because, you know, we can't scale up solar in the US.

Well, just, the electricity output in the US needs to scale up.

Right.

And it's hand for those, like, good resources.

Just to get it somehow.

Yeah.

But, uh, where I was going with this is, if you were in charge, if you were setting all the policies, what else would you change?

Um, so you changed the solar tariffs as well.

Yeah.

I would say, and anything that is limiting factor for electricity, um, basically a dress provided is not, like, very bad for the environment.

So presumably some permitting reforms and stuff as well would be in there.

Yeah.

But there's a fan about a permitting reforms that are happening.

A lot of the permitting is a state based, so, um, um, but anything, but, but this, this administration is, is good, um, removing permitting, uh, robots.

Um, and I'm not saying all tariffs are bad.

I'm just saying, because I think solar tariffs.

Yeah.

Yeah.

I mean, sometimes if, like, if another country is subsidizing the output of something, um, then then you have to have kind of alien tariffs to, uh, protect domestic industry against, uh, subsidies, but I'm not a country.

What else would you change?

I don't know if does that much that the government can actually do?

Yeah.

Well, one thing I was wondering is, it seems like the, for the policy goal of creating a leaves for the U.S. versus China, it seems like the export bans have actually been quite impactful for China's not producing leading edge chips and the export bans really bite there.

China's not producing, uh, leading edge turbine engines and similarly there's a bunch of export bans that are relevant there on some of the metallurgy.

Should there be more export bans, like, do you think about things like when there aren't out of the drone industry and things like that, but is that something there should be considered?

Well, I think it's important to appreciate that in most areas, China is very fascinating factoring.

Um, there's only a few areas where it is not.

Uh, the, uh, China is a manufacturing powerhouse next level, like we will look at it.

It was very impressive.

Yeah.

Yeah.

I mean, uh, if you, if you take like refining of, of or, um, I'd say roughly China, uh, does more just twice as much or refining of, of, of, of, on average as the rest of world combined.

Um, and, and I think there's some areas like say refining gallium which goes into its full of cells.

Um, I think there are like 98% of gallium refining.

Um, so, so China is actually very advanced in manufacturing in, as a most areas.

It seems like we're, like, there is just comfort with this supply chain dependence and the ash nothing's really happening on it.

Supply chain bush supply chain depends on say like the gallium refining that you're saying.

Yeah.

Yeah.

There's, there's, there's a, there's a, all the rare earth, where, earth, stuff and, yeah, rare earth, which are, as, as you know, not rare.

Yeah.

Like we actually do, do rare earth or whining in the U.S. send the, the, the, the rock, uh, put it on, on a, on a train and then confront a boat's China, it goes another train and it goes to the, um, rare earth refining, uh, refining and trying to who then refine it, put it into a magnet, put it into a mode of service, and then set it back to America.

So the thing, we're, we're really missing a lot of, of, of, or refining, um, in, in America.

Isn't this worth a policy intervention?

Yes.

Uh, well, I think there are some things being done on, on that front, um, but, but, but, we kind of need optimists, apparently, to, to build our own refineries.

Um, so you think the main advantage of China has is the abundance of skilled labor.

And that, that's like, that's, that's the, that's the thing optimist fixes.

But they also we need the times, like, four times our population.

But we need, so I mean, there's this concern, if you think like humans are the future, that like, right now, if it's the skilled labor of manufacturing that's determining who's, who can build more humanoids, you know, China has more of those in manufacturers, more humanoids.

Therefore, it gets, it gets the optimist future first, um, go, and it just like keeps that switch on going.

It seems like you're sort of pointing out that sort of getting to a million, after my, yeah, requires the manufacturing that the optimist supposed to help us get to, right?

You can, you can close that recursive loop pretty quickly with a small number of optimists.

Yeah.

So you, you close the recursive loop, um, to help the robots pull the robots, um, and then we can, you know, try to get to tens of miles a year.

Maybe if you start getting to hundreds of miles a year, you're, you're, you're going to be the most competitive country by far.

We, we definitely can't win with just humans, because China has four times a far relation.

Right.

And frankly, America's been wanting for so long that, you know, just like a, like a pro sports team that's been wanting for a very long time, tend to get complacent and entitled.

And that's why they stop wanting, um, because it's, you know, don't work as hard anymore.

Uh, so I think the, frankly, just, um, observation is the average work ethic in China's higher than in the US.

Right.

So it's not just that there's four times a far relation, but the work, the, the amount of work that people put in this higher.

Um, so you, you can like, you can try to rearrange the humans, but you're still one quarter of the, uh, you know, it's assuming that the productivity is the health is, is, is the same, which I think actually might not be, yeah, China might have advantage on productivity, because some, um, we will do one quarter of the amount of things as China.

Um, so so we, we can't win on the human fund.

Um, and our worth there's been low for a long time.

So, uh, we've got both rates been a US, both rates been below replacement, uh, so roughly, uh, 1971.

Um, so, so we've got a lot of people retiring, or, you know, what more people dying than, than, than, than, than, then, we're close to sort of more people domestically dying than, than, than being born.

Um, so we definitely can't win on the human fund, but we might have a shot at the road work fund.

Are there other things that you have wanted to manufacture in the past, but they've been two labor intensive or two expensive, that now you can come back to and say, oh, we can finally do the, whatever, uh, because we have optimists.

Yeah, I think we'd like to do more, both more, um, or a fine ories Tesla.

So, um, we're just completed, um, construction and have, um, begun lithium refining, um, without lithium refinery and hopeless greasy Texas.

Uh, we have, um, a nickel refinery, which is called the cathode.

Uh, that's here, Austin.

Um, and, uh, these, these are the largest, this is the largest cathode, this is the largest cathode refinery, largest lithium refinery, and, uh, logistical and, and lithium refinery, uh, outside of China.

Um, um, and, uh, it's like the, uh, yeah, the cathode team would say like, we have, uh, the, the largest and the only, actually, uh, cathode refinery in America, the largest, not just the largest, but it's also the only.

So, it was pretty big, even though it's the only one.

Um, but I mean, there are other things that, uh, you know, um, you, you could, you could do a lot more refineries and, um, help the, the help America be more competitive on refining capacity.

So, so there's like, there's basically a lot of work for the optimal to do, uh, that, that most Americans, very few Americans frankly wanted to.

Uh, I mean, I've, I'm actually, there were finding work too dirty here.

That's, uh, it's not, it's actually, no, we don't, um, there's not, we don't have toxic emissions from the refinery or anything.

Um, they're cathodes, they're firing short.

Right.

Sort of, and Travis County, like, five minutes from toe.

Why can't you with humans?

No, you, you can't.

You find out of humans.

Ah, I see, okay, yeah.

Like, no matter what you do, you have one quarter number of humans in America.

Yeah, and try it.

So if you have them do this thing, they can't do the other thing.

So, so then, um, well, how do you, how do you, how do you build this refining capacity?

Well, you could do it with the optimal.

Um, and, um, uh, but not many, not very many, not very many Americans are all, all planning to do refining.

I mean, how many of you are on it, too?

Not a few.

But what are you refining to refining?

But, you know, B.Y.D.

is reaching Tesla production or sales in quantity.

What do you think happens in global markets is Chinese production and E.B.

is sales up?

Um, well, uh, Travis extremely competitive in manufacturing.

So, um, I think this, there's going to be a massive flood of Chinese vehicles and, um, another, basically, what's manufacturing?

Uh, things.

I mean, as it is, as I said, like, Travis, like, probably just twice as much refining as the rest of the world can buy.

Yeah.

So, if you go, you know, if you, if you just go, go down to like fourth and fifth tier, uh, supply chain stuff.

Like, like, like, like, the base that we've got energy, then you've got mining and refining.

Um, does those, those foundation layers are, uh, like, said, as a rough guess, trying to twice as much refining as the rest of the world can buy it.

So, any given thing is going to have, uh, uh, uh, Chinese content because trying to do twice as much manufacturing refining work as the rest of the world.

Um, and, uh, and then the, the, the, the, the, like, all of its finished product, with the cars, uh, in Chinese of powerhouse.

I mean, I think this year, China will exceed three times US electricity output.

Um, like, electricity output is a, as a, as a, reasonable proxy for, uh, you know, for the economy.

Uh, so, like, you know, to run the factories and run, run everything you need electricity.

So, electricity is, is a, it's a good proxy for the, for the real economy.

Um, and so, for China is, you're, we're trying to pass this three times to US electricity output.

It means that it's industrial capacity.

That's a rough approximation.

It's three times that will be three times out of the US.

We're reading between the lines.

It sounds like what you're sort of saying as absence and sort of a humanoid recursive miracle in the next few years.

On the sort of like whole of manufacturing energy, uh, raw materials, chain, like China will just like dominate, whether it comes to like AI or manufacturing EVs or manufacturing humanoids.

In the absence of, of, um, breakthrough innovations, uh, in, in the US, uh, China will, uh, utterly dominate.

Interesting.

Yes.

Robotics being the main breakthrough in Russian.

Well, if you do, like to to scale AI, uh, into, in space, like, basically need space, you need the humanoid robots, you need real well AI, you need, um, the million tons of your tool, but, um, like let's just say, like, if we get the mass driver on the moon going, well, if everything, um, then I think, uh, when I saw all of our friends, yeah, so this is like, I call that winning.

Well, I call it winning.

You can finally be satisfied.

You've done something.

Yes.

You know, the mass driver on the moon.

That's right.

I just want to see that thing on first.

What was that out of some sci-fi or a person year?

Well, actually, there's, there was a Heinlein book that the moon is a horseshoe.

That's right.

Okay.

But that's slightly different.

That's a gravity thing, Charles, or, um, no, they have a domestic role.

Okay.

Yeah.

Yeah.

But they use that, uh, attack Earth.

So maybe it's something great.

Just that to, uh, it's not their independence.

Exactly.

What are your plans for the last part of the moon?

They, they're sort of their independence, uh, Earth's government disagreed and they love things, um, to a Earth's government agreed.

That book is a huge, uh, but I found that book much better than, um, his old one, the Demon reads, um, Stranger of Strangestraseland.

Yeah.

Grockett Grock comes from Strangestraseland.

Yeah.

Yeah.

But I much preferred.

Yeah.

Strangestraseland.

If the first two thirds of Strangestraseland, or a good and then it gets very weird.

And then the, uh, the, uh, the, uh, I wish.

Yeah.

Um, but this was some good concepts in there.

Yeah.

Label box can get your robotics and our all data at scale.

Take robotics.

Let's say you need 100,000 hours of egocentric video.

Label box starts by helping you define your ideal data distribution.

Like, for example, maybe no single task category should occupy more than 1% of trading volume.

And at least 10% of trajectories should capture failure and recovery states.

Next, Label box assigns its distribution to its massive network of operators.

You're not limited to the small range of scenes that you can set up in a single warehouse.

Instead, each one of Label boxes operators has access to lots of unique physical environments where they can film themselves completing a wide variety of tasks.

Label boxes tag automatically categorizes each video so that their operators always know which tasks to remain and what they need to work on next.

For our old data, Label box takes a similar approach.

They work with you to understand the right distribution of tasks.

And then their subject matter experts build the hyper realistic digital environments and rubrics that you need to collect the highest quality trading data.

So whether you're training robots in the real world or agents for computer use, Label box can help.

Go to Labelbox.com slash Sparkash to learn more.

One thing we're discussing a lot is kind of your system for managing people.

Like, you interviewed the first few thousand of SpaceX employees and I've seen lots of other companies.

What is your business scale?

Well, yes, but what, what, what doesn't scale?

Me.

I'm not sure sure.

I'm not like, like, what are you looking for?

Initially, it's not enough hours and days, impossible.

But what are you looking for?

That's someone else who's good at interviewing and hiring people.

Lots of generous across.

Well, at this point, I mean, I've got, um, I might have more training data on evaluating technical talents, especially, but accountable kinds, I suppose, but technical talent, especially given that I've done so many technical interviews and then seen the results.

So my, my training set is, is very, is an almost end of, as a very wide range.

The generally the thing I ask for are bullet points for evidence of exceptional ability.

So it's, like, it's, it's, these things can be like pretty off the wall.

It doesn't need to be in the, in the domain, the specific domain, but evidence that evidence of exceptional ability.

So if somebody, if somebody can, like, cite, like, even one thing, but just say three things, where you go, wow, wow, wow, then that's, that's a good sign.

Well, why do you have to be the one to determine that person?

No, I don't exactly.

It's impossible.

Right.

The, I mean, total, it had, it had, California's 200,000 people.

Right.

But in the early days, how was it this, that you were looking for that couldn't be delegated in those interviews?

Well, I, I guess I, I need to build my training set.

It's not like I had about a thousand here.

I would make mistakes, but then I'll be able to see where I, I thought somebody would work out well, because I didn't, and then why, why did they not work out?

Well, and what can I do to, I guess, RL myself, yes, to, uh, in the future, um, have a better batting average, whatever your people.

So, and my, my batting average is still a perfect, but it's, it's very high.

What are some surprising reasons?

People don't work out?

It's surprising reasons.

Um, like, you know, they don't understand techno domain, et cetera, et cetera.

But like, no, you, you, like, you've got, like, the long tail now of, like, I was really excited about, about this person.

You didn't work out.

Curious what that happens.

Uh, yeah.

So, the, I mean, generally when I tell people, or tell myself, I, yes, as personally, um, is, don't look at the resume, just believe you're interaction.

So, if the resume may seem very impressive, and it's like, wow, you know, like, resume looks good.

But if the, if the, if the conversation, uh, after 20 minutes is, is that conversation is not well, um, you should believe the conversation, not the, not the, not the paper.

I feel like part of your method is that, you know, there's this meme in the media a few years back about Tesla being a revolving door of, uh, executive talent.

Whereas actually, I think when you look at us, Tesla has had a very consistent and internally promoted executive bench over the past few years, and that it's SpaceX, you've all these folks like Mark Trunkosa and Steve Davis and Steve Davis runs boring companies.

No, and that, yeah, but the Riley and folks are fast.

And it feels like part of has worked well is having very capable technical deputies.

What do all those people have in common?

Uh, well, so the, I mean, it tells us, it's, it's a, sort of senior team, uh, at this point, probably got average 10 year of, uh, 10 or 12 years.

It's quite long.

Yeah.

Um, so, um, but there, there are times when Tesla went through extremely rapid, an extremely rapid growth phase, um, and so it was somewhat, things were just somewhat sped up, and when a company, as, as, as, you know, a company goes through different words of magnitude of size, you know, uh, people like who could, who could help manage, say a 50% company versus a 500% company versus a 5,000% company versus a 50,000% company.

People.

Yeah, it's just not the same team.

Yes.

And it's not always the same team.

So if, if a company is growing very rapidly, the rate at which, uh, executive positions will change, will also be proportionate to the, the, the, the, the, the, the, the, the, the, the growth.

Mm-hmm.

Um, then, uh, to, to, to as I had, uh, for the challenge where, when, when Tesla had very successful periods, uh, we would be, um, relentlessly recruited from, um, like we're relentlessly, um, like when Apple had their electric car program, they were covered bombing Tesla with recruiting calls.

It was, uh, and it, and it, it's just unplugged their phones.

Like, like, you've just, it's just, they can try to go work done here.

Yeah.

If I get, you know, one more call from an Apple or creator, um, but they were, they were, they're, they're opening off with that any interview with me, like double a conversation at Tesla.

Um, so, um, so, so, so, so, uh, so we had, a bit of the Tesla, Pixie does, uh, thing where it's like, or if you hired, it tells a, exactly, you're, suddenly you're gonna, everything's gonna be successful.

Um, and I, I fall and pray to the Pixie dust, uh, you know, thing as well, where it's like, oh, well, hire someone from Google or Apple, and they'll be immediately successful, but not, that best on how it works.

Um, you know, people with people, it's, it's not like magical Pixie dust.

Yes.

So, we're going to have the Pixie Pixie dust problem.

Um, we get relentlessly recruited, and, um, and then also being, Tesla being, um, engineering, especially being primarily in Silicon Valley, uh, it's easier for people to just, like, they don't have to change their life very much.

Yeah.

They can just get, you know, the, the computer's going to be the same.

Yeah.

Yeah.

Um, so how do you prevent that?

How do you prevent the Pixie dust effect for everyone's trying to coach all your people?

Um, I think we can, I don't think there's much we can do to, to, to, to, yeah, stop it.

Um, but that's like, that's one of the reasons why it tells a, uh, they're, they're really being in Silicon Valley, um, and, uh, and having the Pixie dust thing at the same time, um, meant that, uh, there was just, uh, a very, very gross of recruitment.

Only being in Austin helps that.

Uh, Austin, yeah, it's, it still helps.

Uh, I mean, Tesla still has a majority of its engineering in California, um, so, um, the, you know, for, for getting engineers to move, uh, called the significance significant other problem.

Yes.

So another's have jobs.

Yeah.

Yeah.

Yeah.

Exactly.

So, um, for Starbase, that was particularly difficult.

Yes.

So it's the odds of, you know, finding a lot of space to drive around to Texas, Matt.

You'll, pretty well.

Oh, yeah.

Yeah.

Yeah.

It's quite, quite difficult.

I mean, it's like a technology monastery.

It's a thing, um, you know, remote and mostly dudes.

But again, if you, that might have been from an OSF.

Yeah.

Yeah.

If you go, but if you go back to these people who've really, um, been very effective in a technical capacity at Tesla at space, uh, and, and those sorts of places, what do you think they have in common, other than, like, is it just that they're very sharp on the, you know, walk a tree or the, you know, the technical foundations, or do you think get something organizational?

That's something about their ability to work with you?

Is this their ability to, uh, like, be, you know, flexible, but not too flexible?

What makes a good sparring partner for you?

I don't think it was a sparring partner.

I mean, I mean, I, if there's somebody gets things done, I, I, I, I, I love them and if they don't, I think, so it's pretty straightforward.

It's not like some, it is in crack, I have a thing.

Um, there's somebody else who's well.

Um, I'm a huge fan and if they don't, I'm not, um, but it's, it's not about mapping to my, it is, it is a credit preferences.

Also, they try not to have it, be mapping to my, it is a credit preferences.

Um, so yeah.

Um, yeah, but, uh, generally, I think it's good idea to hire for, um, talent and drive and trustworthiness.

Um, and I, I think, uh, goodness of heart is important.

Um, I, I, I, I, I, I waited that one point.

Um, so, like, are there a good person trustworthy, uh, so smart, talented and hardworking, uh, if so, you can add to main knowledge.

Um, but those, those fundamental traits, those fundamental properties you cannot change.

So most of the people who, um, are at, uh, tell us and SpaceX did not come from the aerospace industry or the order of the industry.

What is most such a change about your management style as your companies of scale from 100 to 1000 to 10,000 people?

You're, you know, you're no, and for this, like, very micro management, just getting into the details of things.

Nano management, please.

People will manage them.

Um, they have to manage them.

So you're saying, uh, what are we going to go all the way down the flight sponsor?

We're going to go all the way down the highs and lows and some different small ones.

Yeah, well, how do you, I mean, are you, are you still able to get into details as much as you want?

Would your companies be more successful if you could, if they're a smaller, like, how do you, how do you think about that?

Well, because I have fixed amount of time in the day, uh, my time is necessarily, um, delirid as things grow and as a span of activity, uh, increases.

So, you know, um, it's, it's, it's impossible for me to actually be a micro management because, uh, there's, that, that, that, that, that, that would imply have some, like, thousands of hours per day.

Uh, it is, it is a logical and possibility for me to, to my, to my permanent text.

Um, so now, there are times when, um, I will drill down into, uh, a specific issue because that's specific issue, uh, is the limiting factor on, uh, the progress of the company.

Um, and, um, but the, the reason for drilling into that, that, that, some very detailed item is because it is the, this limiting factor, not, it, it's not arbitrarily, they get drilling into, you know, tiny things, um, and, like it, obviously, for a time standpoint, it is physically possible for you, arbitrarily, uh, going to tiny things that don't matter, and that would, and, and that would result in failure.

But sometimes the tiny things, um, are decisive in victory.

Famously, you switched the, uh, Starship design from composites to steel.

Yes.

And you made that decision, like that wasn't, uh, you know, people were going around it, like, oh, we found something better about us like that was you encouraging people against some resistance.

Can you tell us how you came to this whole concept of steel switch?

Uh, yeah.

So, um, the, um, the, um, the, originally, yeah, we were going to make Starship out of, uh, common fiber, um, and, um, composites pretty expensive.

Like the, the, the, the, you know, you can generally, uh, uh, when you do a volume production, you can get any given thing to be to start your approach is material cost.

The problem with with common fibers is that material cost is still very high.

Um, um, so, um, it's about, it's about 50 times, but particularly if you go for a high strength, specialized, common fiber that can handle, um, cryogenic oxygen, it's, it's, it's, it's like, color roughly 50 times the cost is steel.

Um, and at least it, uh, in theory it would be lighter.

People generally think of steel as being heavy and common fiber is being, uh, light.

Um, and for room temperature applications, um, you know, like, uh, more of a room temperature applications like our formula one car, uh, static era structure or, uh, or any kind of era structure really, uh, is going to, you're going to probably be better with a common fiber.

Um, the problem is that we were trying to make this enormous rocket out of common fiber and, uh, our progress was extremely slow.

And it's been picked in the first place just because it's light.

Yes, um, but like at first glance, um, like most people would think that the, the, the, the choice for making that something light would be common fiber.

Um, the, um, now the thing is that, um, the, well, when you make something very enormous at a common fiber and then you try to have the common fiber, um, be officially cured, meaning not, not room temperature cure because, like, you've got, you know, sometimes you got like 50 plies of, of a common fiber and a common bar is really common strain and glue, um, and, uh, and you're, in order to have, um, high strength you need an order clay.

So, something that, well, that can, that's essentially high pressure oven.

And if, um, if, um, if you have something that's, uh, a gigantic, uh, uh, the oven's going to be bigger than the rocket.

Um, so we, um, trying to make the, um, order clay if that's bigger than any order clay that's ever existed, uh, or do room temperature cure, which takes a long time and it and has issues.

Um, but, but the final issue is that we're just making very slow progress, uh, with, uh, with common fiber, um, um, so, um, I think the meta question is, uh, why it had to be you who made that decision.

There's many engineers on your team.

Yeah, how did the team that arrived at steel?

Yeah, exactly.

Like, this is a part of a broader question, like, understanding that you're a competitive advantage at your companies.

Um, so, that, that was because we were making very slow progress with, with common fiber, I was like, okay, we've, we've got to try something else.

Now, for the Falcon 9, the, the primary effort is made of aluminum lithium, which is as very, very good strength to wait.

Um, and, um, actually, it has, uh, about the same, maybe, maybe better strength to wait for its application than common fiber.

But aluminum lithium is very difficult to work with, in order to weld it, you have to do something called friction store welding, where you join the, you join the meta without entering the liquid phase.

Um, so it's kind of well that you could do that, but with a particular type of welding, you could do that.

Um, but, uh, it's very difficult to, like, say, let's say you want to make a modification or attach something to aluminum lithium, you know, I have to use mechanical attachment with seals.

Um, you can't, uh, weld it on.

Um, so, uh, we want to, I want to avoid using aluminum lithium for the primary structure for pistachio.

Um, and, uh, and, and there was this very special grade of, uh, con fiber that, that had the very good mass properties.

So, with a rocket, you're really trying to maximize the center of the rocket that is propellant, you minimize the, the, the mass, obviously.

And, um, the, it likes to, we're making very slow progress.

Um, and, and as, as if this rate would never get too much.

So, we'll better figure something else.

Um, I don't want to use aluminum lithium because of the difficulty of friction still welding.

Um, especially doing that at at at scale is hard enough.

Um, at 3.6 meters in diameter, let alone at 9 meters or above.

Um, then, um, besides, well, what about steel?

And, as so, the, but, now, I had, I had a clue here, because some of the early, um, U.S. rockets had used very thin steel.

The Alice rockets had used a steel balloon tank.

Um, so it's not like it's still never been used before.

I actually had been used.

Um, and when you look at the, at the material properties of stainless steel, especially, uh, if it's been, uh, very, like full-hard, uh, of a strain-hard steel, at cryogenic temperature, uh, the, the strength weight is actually similar to carbon fiber.

So, if you, if you look at the material properties at room temperature, um, it looks like the steel is, uh, it's going to be twice as heavy.

But if you look at the drill properties at cryogenic temperature of full-hard steel, it's a stainless, of, of particular grades, uh, then the, you actually get to a similar strength weight as a common fiber.

And, and in the case of starship, both the fuel and the oxidizer are cryogenic.

So, for, uh, Falcon 9, the fuel is rocker-per-falt-grade carousine, basically, like a, like a, a very, pure form of jet fuel, um, which is, but, but, but that is, that is roughly room temperature, um, all the reduced action.

So, you actually chill it slightly below, we chill it like it would be, um, the lish.

Yeah, we do chill it, but, um, but it's not cryogenic.

In fact, if we made it cryogenic, it would, it would just turn to wax.

So, um, but for Sasha, the, it's liquid methane and, and liquid oxygen, they, they, uh, there are liquid at, at similar temperatures.

Uh, so, uh, so, uh, so basically, uh, almost the entire primary structure is a cryogenic temperature.

So, then, you've got, uh, uh, uh, a 300 series stainless, that's, that's, um, strain-hardened, uh, because it's, uh, almost the whole thing's a cryogenic temperature.

It actually has, uh, similar strength to weight, as, uh, carbon fiber.

Because, uh, 50 times less, the normal material, and is very easy to work with.

You, you can well stainless steel outdoors.

Uh, you could smoke a cigar while welding sand still.

It's like, it's, it's very resilient.

Um, you, you can modify it easily.

It's, it's, uh, if you want to, if you want to, if you want to attach something, you're just weld it right on.

So, um, very easy to work with, uh, very low cost, uh, and um, like I said, at cryogenic temperature, similar strength to weight, uh, to carbon fiber.

Um, then when you factor in that, uh, that we don't, we don't, we, we have a much reduced, uh, heat shield mass, uh, because the melting point of steel is much greater than the melting point of aluminum.

Um, it's about twice the melting point of aluminum, and it's just run the rocket bunch hotter.

Yes.

So, especially for the show, uh, which is coming in like a plate of, a blazing media, uh, it is, uh, the, you, you can, you can greatly reduce the mass of the heat shield.

Um, so that, so you, you can call it cut the mass of the windward, um, hot of the heat shield, and, and maybe in half, and you don't need any heat shield on the, on the, on the leeward side.

Um, so, um, the, the net, if net result is actually the steel rocket, where is less than the carbon fiber rocket?

Because the, the resin in the carbon fiber rocket, uh, uh, uh, it, um, starts to melt.

Um, so, so it's, like, basically, the, the carbon fiber and aluminum have about the same operating temperature capabilities.

Um, and where's steel can operate at twice temperature.

Um, I mean, these are very rough approximations.

People will, well, I won't do the rocket thing.

Well, what, what I mean is like, people, I said, oh, he said it twice, it's actually, it's actually 0.8.

Now, it's shown on the households.

And that's what the main comment is going to be about.

Oh, diamond.

Okay.

That's an important point.

The, the, actually, we remember it in retrospect, but the, the, we should have started.

We're done, still in the beginning.

It was done not to do steel.

Okay.

But to, to play this back to you, what I'm hearing is the steel was a riskier, less proven path, other than the early US rockets versus carbon fiber was like, uh, worse, but more proven out path.

And so you need to be the one to push for, hey, we're going to do this, riskier path and just figure it out.

And so you're advising like a sort of conservatism in a sense.

That's why I initially said, like, the issue is that we weren't making it past our progress.

We're having trouble making even, um, a small barrel section of the carbon fiber, um, that didn't have wrinkles in it.

So, uh, because at, at, at that large scale, you have to have many plays, many sort of layers of the carbon fiber, um, you've got to cure it and you've got to cure it in such way that it doesn't, um, have any wrinkles or, or defects that the carbon fiber is, much less resilient than, than steel.

It has much less, it's less toughness, um, since I think it's like stainless steel, what will scratch and, and, and band, and the carbon fiber will tend to shatter, um, so, um, so toughness being area under the stress strain curve, um, so that you're generally going to have to do better with steel.

Um, let's stay on the steel to be precise.

One of those charge-up questions, um, so I visited a star base two years ago, and what would sound taller, and that was awesome.

It was very cool to see in a whole bunch of ways.

Well, they noticed with that, people really took pride in the simplicity of things, where, you know, everyone's to tell you how starship is just a big sort of can, and, you know, we're hiring welders, and, you know, if you can weld in any industrial project, you can weld here, but, um, there's a lot of pride in the simplicity.

And, uh, well, exactly Sasha was a very complicated rocker.

So that's what I'm going to ask is, are things simpler or are they complex?

I think maybe just what they're trying to say is that, you know, you don't have to have like prior experience in the rock industry to work on the Sasha.

Um, you know, some of the just needs to be, they're smart and work hard, um, and if you trust where they can work in a rocket, they don't, they don't need prior rock, the experience.

Sasha is the most complicated machine ever made by humans, by a long show.

In what regard?

Anything really, that's it.

There is a more complex machine.

Um, there, yeah, I mean, I, I, I, I'd say that there's, there's pretty much any any project I can think of what's the easier than this.

Um, and that's why no one has it made a raptor that we use for.

No one has ever made a fully reusable of a rocket.

It's a very hot, very hot pile.

Um, the, I mean, many smart people have tried before, very smart people, with a mastery sources, and they fail.

Um, so, and we haven't succeeded yet.

Uh, we're, you know, talking is partially reusable, but the up to stage is not.

Um, Starship version three, I think this design, that it can be fully reusable, and that full responsibility is what, and they will enable us to become a multi planet civilization.

Can you say about the cereals?

So, I don't, I'm, like, I, I, I, I, I, I said, I could, I'd, and any technical problem, even like a hydrant plan, or something like that, it's, it's an easier following this.

We've, we've spent a lot of time on bottlenecks.

Can you say about the current Starship bottlenecks are even at the high level?

I mean, trying to make it not explode.

Yeah.

Generally, I don't have a chestnut.

Okay.

It's really wants to explode.

Um, all those combustible.

No, we've had two bursers explode on the test end.

Um, one obliterate the, obliterate the entire test facility.

So, it only takes like one mistake, and, and, and, I mean, the amount of energy contained in, and, and Starship is insane.

And so, that way it's harder than Falcon, and it's because it's just more energy.

It's a lot of new technology.

Um, it's, it's pushing the performance envelope, um, the raptor three engine is, so, very, very advanced engine by far the best rock condition ever made.

Um, but it desperately wants to blow up.

I mean, just put things as perspective here on lift off, um, the rocket is generating over a hundred gigabytes of power.

It's 20 percent of euros.

That's the true season.

Actually, it's insane.

It's a great comparison.

While not exploding.

Sometimes.

Sometimes.

Yes.

So, I was like, how does it not explode?

There's, there's you know, thousands of ways that it could explode, and only one way that that it doesn't.

So, we want it to really not, not really not explode, but fly reliably on a daily basis, like, once a hour, and obviously, you know, blow up a lot.

It's, it's very difficult to maintain that George kidding.

Yes.

Um, and then I'm going to say like, well, like, what's the, what's the single biggest growing meaning problem for starship?

It's, uh, having the heat shield be reusable, that such that the, no one has ever made a reusable orbital heat shield.

Um, so the, the heat shield's got to make it through the assemblies without shocking a bunch of tiles.

And then it's going to come back in and also not lose a bunch of tiles or overheat the main, the main airframe.

And that happens.

It's kind of fundamentally a consumable.

Well, yes, but you're very glad it's in your car also considerable, but the last few a long time.

Fair.

Right.

So, it just needs to last very long time.

Um, that's, it just, yeah, try to, I mean, we had brought the ship back and had it do a soft landing in the ocean.

I've done it a few times, but it lost a lot of tiles, you know, you know, it was, you know, it was not reusable without a lot of work.

Yeah.

So even though it did land, did, did, did come to soft landing.

It was, we're not have been reusable without a lot of work.

Um, and, and that, so it's not really reusable in that sense.

So that's, that's the biggest home that remains as fully reusable heat shield.

Um, so, like, if you want to be able to land it, uh, refill propellant and fly a game.

Uh, was that good, you know, you can't do this laborious inspection of their 40,000 tiles, everything.

I'm curious how you drive, like, when I read biographies of yours, it just, uh, it seems like you're just able to drive the sense of like urgency and drive the sense of like, this is the, this is the, this is the thing that can scale.

Um, and I'm curious why you think other organizations of your, like, the SpaceX and Tesla are really big companies now, and you're still able to keep that culture.

What goes wrong with other companies such that they're not able to do that.

Uh, um, but like today, you said you had a bunch of SpaceX meetings, like, what, what, what is it that you're doing there?

That's like keeping that.

It's adding or just, yeah.

Yeah.

Well, I, I don't know, I get, I guess, uh, the urgency is going to come from grossly eating the company.

So my sense of urgency.

I've, I've like, monocles and surrogacy.

So yeah, that monocles and surrogacy projects through the rest of the company.

Is it because of consequences?

They're like, if, you know, Elon said a crazy deadline, but if I don't get it, I know what happens to me.

Is it just, um, you're able to identify bottlenecks and get rid of them?

So people can move fast.

Like, how do you think about why your companies are able to move fast?

Yeah, I'm constantly addressing the limiting factor.

So, um, I mean, I mean, I mean, on the deadlines front, I mean, I generally actually try to aim for a deadline that that I at least think is at the 50th percentile.

So it's, it's not, it's not like an impossible deadline, but as soon as you're grossed a deadline, I can think of, that could be achieved with 50% probability.

Which means that it's a really late half of the time.

Um, and, um, whatever, like, there is like a law of gas has expansion that applies to schedules.

Like, whatever, given whatever schedule, you, like, if you, if you, you said we're going to do this something in like five years, which to me is like infinity time.

Um, it will expand to fully available schedule and it will take five years.

Um, you know, like, there's like, there's, there's a physical limit.

Like, like, like, like, physics will limit how fast you can do something.

It's like, so like scaling up manufacturing.

There's like, there's a rate of which you can move the atoms, um, and scale manufacturing.

That's why you can't like instantly make, you know, a million of something, million years or something.

Uh, you've got a design manufacturing line.

You can bring it up.

You've got to ride the S curve of production.

Um, so, yeah, I guess like, like, I'll show you what, what can I say that's that's that's actually helpful to people.

Um, I think generally, um, when I feel sense of urgency is, is a, is very big deal.

Um, so, um, and, and you want to have it, you want to, you want to have a, and the grasp of schedule, um, and then you, and you want to figure out what the limiting factor is at any point in time and and help the team address that limiting factor.

Can you maybe talk about the, so Sterling was slowly in the works for many years.

Uh, and yeah, we talked about it all the way in the beginning of the company.

Yeah.

And so then there was a team you had built in redmond, and then at one point you decided this team is just not cutting us, but again, how did you, like, it went for a few years, slowly, and so why did this, why didn't you act earlier, and why did you act when you did?

Like, why was that the right moment of it stacked?

I mean, I had, I had these very detailed, um, engineering reviews weekly, um, that that's, that that's maybe a very unusual level of granularity.

Um, I don't know anyone who wanted to company, or at least a manufacturing company that goes with level of detail that I go into.

So it's, it's not as though, like I have a pretty good understanding of what's actually going on, because we, we go, we go through things in detail.

Um, and I'm a big lever in skip level meetings where the individuals, it's, instead of having the first report to me, say things, it's everyone that reports to them, um, says something in the tech review, um, and, um, and they can't be, um, advanced preparation.

So otherwise, you're, you're going to get, uh, you know, glazed, um, does that say these days?

Yeah, exactly.

Virgin's, you know, who I've done, you know, you're just like called them randomly, like, no, just go around the room and everyone provides enough data.

Uh, so, uh, I mean, it's, it's a lot of information to keep your head, because, um, you've got to, you've got to, you've got to then, say if you're meeting a weekly or twice weekly, you've, you've got a snapshot of what that person said, um, and, and, and you can, and you can, and you can then, you know, flood the progress points, um, you can sort of mentally plot the points on the curve and say, are we converging to a solution or not?

Um, or, or are we, you know, like, I'll, I'll take drastic action, uh, only when I conclude that, um, success is not in the set of possible outcomes.

Um, so, right, when I say, okay, we're not, when I finally reach the conclusion that, okay, unless drastic action is done, we have no chances of success, then I must take drastic action.

So that's, that's, that's, that's, okay, if that conclusion in 2018 took drastic action and fixed problem.

How many, um, you know, you, you've got many, many companies, and in each of them, it sounds like you do this kind of deep engineering understanding of what their own bottlenecks are, so you can do these, um, reviews of people.

Yeah, um, you've been able to scale it up to five, six, seven companies, you've been in one of these companies, you have many different mini companies with them, them.

What, what determines the maximum here?

Could you have like 80 companies?

Maybe?

No.

But like, you, you have so many already, I'm like, that's, that's already remarkable.

Why this current number?

Yeah, exactly.

I know, so, um, we need to barely keep pulling that company together.

Um, um, it's like, it depends on situation.

Um, so, um, I actually don't, don't have regular meetings with foreign company, so that foreign companies sort of cruising along.

I look, basically, if something is working well and making good progress, then there's no point in me spending time with it.

So, I actually, uh, allocate time according to where the, where the, where the limiting factor or the problem, where, where are things problematic?

And, um, or where we're pushing against, uh, like, what, what is holding us back?

Well, you know, I, I focus, the risk of, say it was too many times, the limiting factor.

Um, so, so, basically, if something's good, like the irony is, if something's going really well, um, they don't see much of me, but if something is going badly, there's to be a lot of me.

So, if something, or not, not even badly, it's, it's like, if if if something's the limiting factor, it's a limiting factor.

It's a limiting factor, it's not exactly where everybody, but it's the thing that's, it's the thing that we need to make go faster to.

And so, when something's the limiting factor, that's basically our Tesla, are you, like, talking weekly, daily with the engineer that's working on us?

How does that actually work?

The most things that are the limiting factor are, um, weekly, and some things of twice weekly.

So, the AI-5 chip review is twice weekly.

And so, it's every Tuesday and Saturdays.

Is, is the chip review?

Is it open-ended and how long it goes?

Technically, yes, but, uh, usually it's, it's like two or three hours.

So, sometimes less, it depends on how much if they should go to go through.

Yeah.

Well, that's one thing.

I'm just trying to tease out the differences here, because the outcomes seem quite different.

And so, I think it's interesting to know what inputs are different.

And it feels like the corporate world, one like you're saying, just the CEO doing engineering reviews does not always happen.

Despite the fact that that is the, you know, what the company is doing.

But then time is often pretty finely sliced and, uh, you know, half our meetings are even 15-minute meetings.

And it seems like you hold more open-ended.

We're talking about it until we figure it out.

Sometimes.

Yeah.

Yeah.

Sometimes.

But, uh, most of them seem to, more or less stay on time.

Um, so, um, I mean, today's, uh, staunch of engineering review went a bit longer, because there are more topics to discuss.

Um, they're trying to figure out how to scale two million plus times to overprear is quite challenging.

Can you answer questions?

You said about, um, Optimus and AI that they're going to result in double just a growth rate, it's within a matter of years.

Oh, it looked like the economy?

Yeah.

Um, yes.

Well, I think that's right.

What was the point of the doge cuts?

If the economy is going to grow so much?

Well, I think like, waste and for it, not good things to have, you know.

Um, I, I was actually pretty worried about, I guess, uh, I mean, I think in the absence of AI and robotics, we're actually totally screwed, because the national debt is probably up like crazy.

Um, now our interest payments, the interest payments, the national debt exceed the military budget, which is a trillion dollars.

So if over a trillion dollars, just the interest payments, um, you know, that was like, I was like, okay, pretty concerned about that.

Maybe if I spend some time, we can slow down the bankruptcy of the United States, um, and give us enough time for the AI and robots to, you know, help solve the national debt.

Uh, or we're not to help solve, it's the only thing that could solve the national debt.

Like we are 1,000 percent going to go bankrupt as a country and fail as a country with AI and robots.

Nothing else will solve the national debt.

Um, and so, so we, we'd like to, well, we just need, we need enough time to get, we'll be AI and robots to, and not go bankrupt before then.

I guess I think I'm curious about is, when no starts, you have this enormous, um, ability to enact a reform and, well, not to add an illness.

Sure, sure.

But totally by your point that like, it's important that AI and robotics, drive products, improvements, drive GDP growth.

But why not just directly go after the things you're pointing out of, you know, like, the tariffs on certain components or whether it's like permitting.

I'm like the president.

And, and very hard to cut, to cut, to, to even, even to cut things that are obvious waste and fraud, like, like ridiculous waste and fraud.

Um, what I discovered that is, it's, it's extremely difficult, even to cut, very obvious waste and fraud, um, from the government.

Because the, the, the government has to operate on a, on, like, who's complaining, like, if, if, and if you cut off payments to fraudsters, they immediately come up with the most sympathetic sounding, uh, reasons to continue the payment.

But they don't say, please keep the fraud going, like, they say, you know, it's, they're like, you're killing baby pandas.

And I'm like, meanwhile, there's no baby pandas are dying.

They just making it up.

Um, but the forces are capable of, of, of coming up with extremely compelling, sort of heart-wrenching stories that are false, but nonetheless sound, uh, sympathetic.

And that does what happened.

Um, and, uh, so it's, like, perhaps I should have known better.

Um, and, uh, that I thought way, let's take a, listen, let's try to cut some amount of, uh, waste and fraud from the government.

Maybe that shouldn't be, you know, 20 million people, uh, I walked as alive in Social Security who are definitely dead, and over the age of 115.

The oldest American is 114.

So it's safe to say if somebody's 115, and walked as alive in the Social Security database, um, something is, there's either a typo, like, somebody should call them and say, we, we seem to have your birthday wrong, or, or, or, or we need to mark you instead.

Okay.

Run up the two things.

We intimidate and call to get.

Well, so it seems like a reasonable thing.

Um, and if, if, like, say their birthday is in the future, and they have, you know, a small business administration loan, and their birthday is 2165.

We, that game have a typo or we have fraud.

Um, so we're, say we're period time.

I've gotten the sanctuary of your birthday and correct our great platform movie.

Yes, this is, this is, this is when I, when I, when I'm not ludicrous fraud, this is when I'm all ludicrous fraud.

Where those people getting payments.

So some were getting payments from Social Security, but, but the main fraud vector, uh, was to mark somebody as alive in Social Security, and then use every other government payment system, uh, to, uh, basically, to do to do fraud.

Because what those other government payments system do would do, would do what there was simply do, and are you alive check to the Social Security database?

It's a, it's a bank shot.

What would you estimate is like the total amount of fraud from this mechanism?

Um, my guess is, and other, but by the way, the government accountability office has done these estimates before.

I'm not the only one who's coming out of this, you know, the, the, in fact, I think they, they did, the GAO did analysis, a rough estimate of fraud during the Biden administration, and calculated roughly half a trillion dollars.

So don't take my word for it.

Take it or report issued during the Biden administration.

This is how about that?

From this social security mechanism?

It's, it's one of many.

It's important to appreciate that the, like, the government does not, it is a very ineffective at, at stopping fraud.

Because, uh, it's, it's, it's, it's like, like, it was a company, like, like, like, for stopping fraud, you've got a motivation because it's affecting the earnings of your company.

Uh, but the government just, just, they just print more money.

Um, so it's not, uh, like you, you need, you need caring and confidence.

And these are in short supply at the federal level.

Um, yeah, I'm sorry.

I mean, when you go to the DMV, do you think, wow, this is a bastion of confidence?

Um, well, now imagine it's worse than the DMV because it's the DMV that can print money.

So, was it not possible?

At least the state level DMVs, uh, need to, the state's more or less need to stay with the neighborhood and go bankrupt, but the federal government just prints for money.

Well, it was enough possible.

If there's a cashier half a trillion of fraud, well, why, why wasn't not possible to cut all that?

Uh, because when, when, when, essentially, we did, we, we actually, no, you, you, you really have to stand back and recalibrate your expectations for confidence, uh, because, uh, you're operating in a world where, you know, you've, you've got to sort of make ends meet.

Like, you know, you've got to pay your bills, you've got to, you know, buy the microphones?

Yeah, yeah, exactly.

Um, so, so, you, you, you, you don't have, it's, it's not like there's a giant largely uncarrying who wants to be bureaucracy.

So, you know, it's, and, and, and, and, and, and a bunch of, uh, uh, the accuracy computers that are just, they're just sending payments.

Um, like, like, one of the things that, that, that, that, that, that, that there was, uh, this, and sounds so simple.

Uh, that, that probably will say, um, let's say a hundred billion, maybe two hundred billion a year.

Um, it's simply requiring that payments from the main treasury computer, which is called payments, like payment accounts, master or something like that.

This five trillion payments here, requiring that any payment, uh, that goes out, have a payment of appropriation code, make it mandatory, not optional, and that you have anything at all in the comment field.

Um, because you, you see, you have to have a recalibrate.

How dumb things are?

Breathing, pales were being sent out with no appropriation code.

It's not, not checking back to any congressional appropriation and no explanation.

And this is why the, the Department of War, formerly the Department of Defense, cannot pass an audit because the information is literally not there.

Recalibrate your expectations.

I want to better understand this how much trillion number, because there's, there's an IG report in 2024.

How much like, why is it so low?

Um, maybe, but, uh, which found that, like, over seven years, the, the social security fraud, they estimated it was like 70 billions over seven years, so like 10 billion here, so it'd be curious to see what like the other four and 90 billion is.

Federal government expenditures are seven and a half trillion here.

Yeah.

Um, what, what percentage, how confident do you think how much is the discretionary spending there is like 15%.

Yeah, but, but it doesn't matter.

The, the, the most of the forward is non-discretionary.

It's, it's basically a, a fraudulent Medicare Medicaid, uh, Social Security, uh, uh, uh, uh, you know, disability, uh, it's, there's, there's a zillion government payments.

Yeah, um, and a bunch of these payments are, in fact, uh, they're, they're, they're, they're, uh, block transfers to the states.

So the federal government doesn't even have the information in a lot of cases to even see know if this fraud.

Let's consider, let's, like, reduce your ad of certain, the government, the government is perfect and has no fraud.

What is your probability estimate of that?

I mean, zero.

Okay.

So then, would you say that, for a foreign waste, that, the government, uh, is, has, is 90%.

That also would be quite generous, but if, if it's only 90%, that means that there's $750 billion a year of waste in fraud.

And it's not 90%.

It's not 90% effective.

This seems like a strange rate of first-pins fills the amount of fraud in the government, just like, how much do you think there is?

And then, uh, I, I, anyway, so we don't have to do it live, but I'd be curious, it's, like, something you know a lot about fraud at a strike, people will constantly try to, you're fraud.

Yeah, but as you say, it's, like, a little bit of a, um, we've really grounded down, but it's a little bit of a different problem of space, because you're dealing with a much more heterogeneous set of fraud vectors here than where.

Yeah, but I mean, I mean, let's try, you, you, you have high confidence in your tri-hog.

Um, you have high confidence in high carrying.

But still, for it is non-zero.

Um, now now, now I'm actually at a much bigger scale, there's much less confidence in much less carry.

You know, back paypal, back in the day, we were, we were trying to manage four down to about 1% of the payment volume.

Um, and that was very difficult, took a tremendous amount of confidence in carrying to, uh, get fraud merely to 1%.

Um, now I mentioned that the, your own organization, where there's much less carrying, much less confidence.

It's going to be much more than 1%.

How do you feel now looking back on, um, kind of politics and, and doing stuff there, where it feels like we're from the outside in the, you know, two things have been quite impactful.

One, the America pack and two, and the acquisition of, well, Twitter at the time.

But also it seems like there's a bunch of heartache.

And so what's your, what's your grading of the whole experience?

Well, um, I think, I think those things needed to be done to maximize the probability of the future is good.

So, um, politics generally is very tribal.

Um, and it's, it's very tribal.

And people lose their objectivity usually with politics.

Like they, they generally have trouble seeing the grid of the other side or the bad in their own side.

That's generally how it goes.

Um, I, that, that I guess, was one of the things that's been surprising the most is you often simply cannot reason with people.

Um, if they're in one tribe or the other, they simply believe that everything that tribe does is go to anything, the other political tribe does is bad.

Um, and persuading them is, otherwise it's almost impossible.

Um, so anyway, but, um, I think I think overall, those actions, um, acquiring Twitter getting Trump elected even though it makes a lot of you're angry.

Um, I think those, I think those actions are good for school or good for civilization.

Um, but yeah, well, how does if you didn't do the future, you're excited about?

Well, um, American needs to be strong enough to last long enough to, um, extend life to other planets and to, I get, I guess, AI and robotics to the point where we're going to show the future is good.

Um, like, on the other hand, if, if we were to descend into, um, say Communism or, or some situation, whether where the state was extremely oppressive, um, that, that would mean that we, we might not be able to become multi-planetary, um, and we might, we, the, the, the state might, um, you know, step out, um, progress and analytics.

How do you feel about, um, you know, Optimus, Grog, et cetera, are going to be leveraged by, and not just yours kind of, any revenue maximizing companies products will be leveraged by the government over time.

How does this concern manifest in what private companies should be willing to give governments, what kinds of guerrreal should, like, should, you know, that it should, um, AI models be, uh, um, me to do whatever the government that has contract them out to do, ask them to do, um, should, like, should, should, should, should Grog get to say, like, actually even, the military wants to do X, no, the Grog will not do that.

But I probably, probably the biggest danger of, uh, yeah, maybe the biggest danger of fail for AI and robotics going wrong, wrong is, is government, interesting, you know, um, I mean, the, the way it, like, by people who are opposed to corporations or, or worried about corporations, it shouldn't, um, really worry about the most about government, because government is just a corporation in the limit.

It's a government, it is, it is, it is, it is, government is just the biggest corporation with them and awfully unbiallists.

Um, so I always find it, like, estranged like, economy, where people would think corporations are bad, but the government is good, when the government is simply the biggest, and, and, and worst corporation.

But people have that economy.

There's somehow think that the same time the government can be good, the corporations, bad, and this is not true, corporations are, have better morality than the government.

So I, I actually think it's, uh, you know, that's, uh, that is the thing to be worried about, it's like, if the, you know, should, if the government should not, like, the government could potentially use AI and robotics to progress the population.

Like, that is a service to, um, as a, as a guy building AI and robotics, how do you, how do you, like, how do you prevent that?

Well, I think that, like, if, if you have a limited government, um, if you limit the power of the government, which is, like, really what the US Constitution is intended to do is intended to limit the power of the government, then then, uh, you're probably going to have a better outcome, then if you have more government.

So, the projects will be available to all governments, right?

Yeah, and I don't know what all governments, um, I mean, it's difficult to predict the, like, I can say, like, what's, what's, what's the end point or, like, what is what is many years in the future, but it's difficult to predict the, the sort of path along along that way.

Like, if civilization progresses, AI will vastly exceed the sum of all human intelligence and, and they'll be far more of it than humans.

Um, along the way, what happens is very difficult to predict.

I mean, it seems like one thing you could do is just say, um, uh, you're not allowed to, whatever government actually are not allowed to use optimists to do X, Y, Z, just right out like the policy.

I mean, you, I think you treated recently that Grok should have a more constitution, um, and one of those things could be that we, we limit why governments are allowed to do with this advanced technology.

I mean, yeah, well, we can do what is, what, what, what, what?

I mean, it's, if, if, if the policy is just passed a law, uh, then, and they can enforce that law, then it's hard to not do that law.

The, the best thing we can do is, is, is limited government, uh, where, you know, you have, you have the appropriate cross checks between the executive, judicial, and, um, legislative branches.

I guess the, uh, the reason I'm curious about it is that's like, at some point, it seems like the limits will come from you, right?

Like, you've got the optimists, you've got the space GPU, you've got the, I think I'll be the boss of the government.

Or you will get, you will, like, the, I mean, already, it's the case with SpaceX, that for things that are crucial to the, um, uh, like, the government really cares about getting certain satellites up in space, whatever, like, it needs SpaceX.

Uh, it is the, it is the, um, a necessary contractor, and you are in the process of building more and more of the, um, uh, the technological components of the future that, that, that will have a analogous role in different industries, and you could have this ability to, like, set some policy that, um, you know, it's, it's, it's suppressing classical liberalism in any way.

I, my companies will not help in it, in any way with that, or, you know, some policy like that.

Um, I'll do my best to ensure that anything that's within my control, maximizes the good outcome for humanity.

Um, I think anything else would be short-sighted, um, because obviously I'm part of humanity, so, um, I like humans, um, that, pro-human, pro-human.

Um, you, you mentioned that Dojo 3 will be used for space-based compute.

Um, you really read my, what I say.

I don't know if you know Twitter, uh, I know you lot enough.

There's a lot of followers.

They, yeah, we're, um, how do you, um, how do you have to send my secrets?

I've asked them all the time.

How, how do you design this chapter space, what it, what, like, what, what changes?

Well, I guess you want to have designs to be, um, more radiation tolerant and run out of higher temperature, uh, so you could, um, roughly, if you increase the, um, operating temperature by 20 of set in degrees Kelvin, you can cut your radiator mass in half.

Um, so, right, running out of higher temperature is, is helpful in space.

Um, there's, I mean, there's various things you could do for shielding of the memory and, but like, neural nets are going to be very resilient to bedflips.

Yeah, so, like, most of what happens for radiation is like random flipflips.

Um, but like, if you've got like, you know, a multi-trailing parameter model, and you get a few flipflips, it doesn't matter.

Um, it's much, like, curiosity programs are going to be much more sensitive to flipflips, then, um, it's some giant parameter file.

Um, so, uh, I just designed and run hard.

And, um, I would, I think it pretty much suits the same way that you do things on those part from making it run water.

Um, I mean, the solar arrays, most of the weight on the satellite is, is it a way to make the, um, the GPUs even more power dense than what in video and TPUs, and it's that are planning on doing that would, you know, be a special privilege in the space role.

Well, I mean, the basic math is, like, uh, um, if you can do about a kilowatt peretical, um, and then you'd need, um, you know, 100 million full radical trips to do 100 gigawatts.

Yeah.

So, yeah, depending on what your yield assumptions are, you know, um, that, that's, tells you how many trips you need to make, um, virtually, you need, if you want, if you, if you, if you're going to, uh, have 100 gigawatts of power, you need, you know, 100 million trips running that, that are running a kilowatt sustained output peretical.

Um, a hundred.

It's a math.

100 million ships, uh, it depends on, yeah, if, if you, if you look at the die size of something, like black olgy peas or something, and how many can get out of the way for, you can get, like, on the order of dozens or less, uh, per way for.

So, you're, basically, you're, this is a world where if we're putting that out, every single year, you're producing millions, millions of away for a month.

Um, that's the plan with your app.

Millions of away for a month.

They'd be asked for us notes.

It's, it could be some number and also a million, I think that.

You're going to do the memory, too.

Yeah.

Are you going to make a memory, Fat?

I think the turf house is going to do a memory.

It's going to be logic, memory, and texture.

I'm very curious how, somebody like gets start.

This is like the most complicated thing, man has ever made.

And obviously, like, if anybody's up to the task, you're up to the task.

Uh, like, what do you, so you realize this is a bottleneck, and you go to your engineers and like, what is the next, like, what do you tell them to do?

I want to million reefers a month in 2030.

What is the next, like, what do you, that's right?

Do you like colleagues and I'm like, what is this?

I think I want.

What is the next that so much to ask?

Well, um, we make a little fab and see what happens.

Uh, make a mistake at a small scale, and then make a big one.

Is a little fab done or is it?

No, it's not done.

I, which, I mean, George, they're not going to keep that cat lib.

They're just going to come out of the bag room.

If you're like, runs, hovering over the bloody bed, you know, you're able to see it's construction for us on X, right, you know, real time.

Um, so, you know, we, we, we, I mean, like, I don't know, we're could just flounder and failure to, say it's like not, uh, success is not guaranteed, but, um, since we want to try to make, uh, you know, something like 100 million, yeah, we, we, we, we need, we need, we want to 100 gigawatts of power and 100, that chips that can take 100 gigawatts.

And it's so, cool it, you know, but yeah, bite by 2030.

So then, um, and we'll take as many chips as I was applying as we'll give us.

I've said this to, I've actually said this to TSMC and Samsung and my bonus, like, please build your more fabs faster.

Um, and we will guarantee you to buy the output of those fabs.

Um, so that, there were already, like, moving as fast as they can.

Like, it's, it's not like, do we clear?

It's not like, you know, it's not like, uh, either, it's, it's, it's not like, it's us plus them, you know.

There's an irony that the people doing AI want a very large number of, you know, chips that is quickly as possible.

And then, many of the input suppliers, the fabs, but also, you know, the turbine manufacturers are not ramping up production very quickly.

No.

And the explanation here is that they're dispositionally conservative, you know, their Taiwanese or German as the, you know, Australia maybe.

And they just like don't believe to say, like, is that really the explanation or is there something else?

Well, I mean, it's reasonable to, like, if somebody's been saying the computer memory business for, uh, 30 or 40 years, and they've seen cycles, they've seen, like, women bust like 10 times.

Yeah.

You know, so, so, like, that's a lot of layers of scar tissue, you know, so it's like, it's like, during the boom times, looks like everything is going to be, but great forever.

And then then the crash happens and then like desperately trying to avoid bankruptcy.

Um, and then there's another boom, another crash.

Are there, are there, are there ideas, you think others should go pursue that you're not fair, whatever reasons right now?

Um, I mean, there are a few companies that are, they're saying like, uh, new ways of doing chips.

Uh, but there's just not scaling fast.

I mean, within AI, I mean, just generally.

I'd say like, people should just sort of do the thing that, where they find that they're highly motivated to do that thing.

Uh-huh.

As opposed to, you know, some, some, some of some idea that that I suggest, but they should do the thing that they find personally interesting and motivating to do.

Um, um, but yeah, we're going back to the limiting factor.

He was out of praise, but a hundred times.

Um, the, the current limiting factor that I see in the timeframe, you know, in the sort of 20, 29, 29, like, in the, in the, in the three, three to four year timeframe, um, it's chips.

Um, in the, in the one year timeframe, it's, it's energy, it's power production electricity.

Like, it's, it's not clear to me that there's enough that, uh, use electricity to turn on all the, the air shifts that are being made.

Um, just, the towards the end of this year, I think we're going to have real trouble turning on, like the chip output will exceed the ability to turn chips on.

Well, like, what's your plan to do with that world?

Well, we're trying to accelerate electricity production.

Um, I guess that's, that's maybe one of the reasons that, um, I say, I will, will be maybe the leader of hopefully leader.

Um, is that we'll be able to turn on more chips than other people content on faster, um, because we're, we're, we're good at hardware.

And, and, and, and generally, the, the innovations from the corporations that must, good, cold self-labs, um, the, the ideas tend to flow, like it's, it's where to see that there's, like, more than about a six month difference, um, between, um, like, a, like, a big ideas, uh, travel back and forth, um, with the people.

So, so I think you, you sort of hit the hardware wall, and, um, and then whatever, whichever company can scale hardware, the fastest will be the leader.

And so I think, actually, I will be able to scale hardware the fastest and therefore, most likely will be the leader.

You, you, you, you jokes are, you know, um, self-conscious about, uh, you know, using the, uh, the limiting factor for ease again.

But I actually think there's something deep here.

And if you look at a lot of things we've touched on over the course of that maybe you're kind of good note to end on.

Like, if you think of a, senescent, lower agency company, it would have some bottleneck and not really be doing anything about us, um, you know, market reason had the line of, uh, most people are willing to endure any amount of chronic pain to avoid acute pain.

Uh, and I feel like a lot of the cases we're talking about are just leaning into the acute pain, whatever it is.

It's like, okay, we've got to figure out how to, you know, work with steel or we've got to figure out how to run the chips in space or like, we'll take some near term acute pain to actually solve the bottleneck.

And so that's kind of the unifying thing.

I'm a high fan threshold.

That's helpful.

I solved the bottlenecks.

Yes.

Um, so, you know, one thing I can say is like, uh, I think if you're just going to be very interesting.

Um, and, um, and I, as I said, uh, the dowels have only been, especially dowels, so I think it was like on the ground for like three hours or something.

Um, it's better to be, it's better to err on the side of optimism and be wrong than, err on the side of pessimism and be right, uh, for quality of life.

So, you know, you're, you're, you're happy, Ms. We'll be, you'll be happier if you, if you, or err on the side of optimism, rather than err on the side of pessimism.

And so I recommend err on the side of optimism.

That's that.

Cool.

You know, thanks for doing this.

Thank you.

All right, thank you guys.

Can't wait to make it.

All right.

Oh, great.

So I'm going to, hopefully this encounters the pain in the faint tolerance.

Hey, everybody.

I hope you enjoyed that episode.

If you did, the most helpful thing you can do is just share it with other people who think might enjoy it.

It's also helpful if you leave a rating or comment on whatever platform you're listening on.

If you're interested in sponsoring the podcast, you can reach out at barcache.com slash advertise.

Otherwise, I'll see you in the next one.