Elon Musk’s new Space AI ambitions

[Vultur]

Eh. I don't really agree, but we'll have to see what happens ... And it might be a couple decades before the answer is clear.

What I'm more concerned about in the short term is an investment bubble burst bringing down space stuff along with everything else. This could happen *regardless* of the soundness of the underlying technology - the dot com bubble of the late 90s wrecked private space projects then, though both the Internet in general and satellite internet specifically ultimately did work out.

I'm pretty sure that it'll go down a lot faster than it did with thermodynamic engines.

Probably*, but that doesn't really answer the underlying question. There was a very long time between the invention of steam engines and them really becoming widespread; 20-25 years or so (like the time gap between Teledesic and Starlink) would still be much faster.

Economic problems/investment bubble bursts/etc can happen on a much, much shorter timescale than that, and can happen almost regardless of the underlying value or not of the technology. (The dot com bubble didn't mean the Internet didn't ultimately get everywhere, the housing bubble didn't mean people stopped needing housing.)

I am *also* much less optimistic than you about the ultimate value of LLM based AI technology or "generative AI", but that's basically a *completely separate question*. The short term (next 4-5 years) concerns remain either way.

I just don't want SpaceX to get so economically tied to xAI or some other AI thing that an investment bubble burst pushes Moon and Mars exploration back 20-25 years. And I want them to keep focus on Mars for the next couple synods, at least as much as they can with Artemis obligations.

*Though I don't think the 19th-20th century industrial revolution analogy really holds. That was a period of very rapid demand growth driven by both overall population growth and much more of the world being drawn into the industrial economy.

[meekGee]

Eh. I don't really agree, but we'll have to see what happens ... And it might be a couple decades before the answer is clear.

What I'm more concerned about in the short term is an investment bubble burst bringing down space stuff along with everything else. This could happen *regardless* of the soundness of the underlying technology - the dot com bubble of the late 90s wrecked private space projects then, though both the Internet in general and satellite internet specifically ultimately did work out.

I'm pretty sure that it'll go down a lot faster than it did with thermodynamic engines.

Probably*, but that doesn't really answer the underlying question. There was a very long time between the invention of steam engines and them really becoming widespread; 20-25 years or so (like the time gap between Teledesic and Starlink) would still be much faster.

Economic problems/investment bubble bursts/etc can happen on a much, much shorter timescale than that, and can happen almost regardless of the underlying value or not of the technology. (The dot com bubble didn't mean the Internet didn't ultimately get everywhere, the housing bubble didn't mean people stopped needing housing.)

I am *also* much less optimistic than you about the ultimate value of LLM based AI technology or "generative AI", but that's basically a *completely separate question*. The short term (next 4-5 years) concerns remain either way.

I just don't want SpaceX to get so economically tied to xAI or some other AI thing that an investment bubble burst pushes Moon and Mars exploration back 20-25 years. And I want them to keep focus on Mars for the next couple synods, at least as much as they can with Artemis obligations.

*Though I don't think the 19th-20th century industrial revolution analogy really holds. That was a period of very rapid demand growth driven by both overall population growth and much more of the world being drawn into the industrial economy.

Agreed that there's much that is unknown, scope and timewise.

I don't know that SpaceX is betting the farm on AI though.  I think financially Starlink was already a good enough foundation.

Orbital AI is just one possible expansion, same as p2p. Maybe it will, maybe it won't.  I see potential and possibilities, not a sure thing.

But it's not just "if you don't try, you'll surely miss out".  It's that in Musk world, the act of trying is what potentially makes it happen. It's not about catching the wave, it's about making the wave. As in electric cars, LEO constellations, etc.

[Vultur]

I don't know that SpaceX is betting the farm on AI though.  I think financially Starlink was already a good enough foundation.

I think so, and hope so. It was the connection of the space datacenter stuff to Starlink v3 that started me worrying.

But it's not just "if you don't try, you'll surely miss out".  It's that in Musk world, the act of trying is what potentially makes it happen. It's not about catching the wave, it's about making the wave. As in electric cars, LEO constellations, etc.

That then raises the question of whether it's a good thing to make happen. I am far from convinced that AI that *is* cost-effective would actually be an overall net good for humanity, in a world with limited demand growth.

[seb21051]

That then raises the question of whether it's a good thing to make happen. I am far from convinced that AI that *is* cost-effective would actually be an overall net good for humanity, in a world with limited demand growth.

Some perspectives on what is good for humanity:

Seb's Theorem: Nature's Agenda is to enable as many viable species as possible to develop. The individuals of those species are useful only in their role to strive for the ongoing survival of their species. They have no other purpose. Certainly no such thing as Rights, Freedom, or the pursuit of happiness. That is a human thing, and in viewing as many other species as possible, purely a figment of our collective imagination. (I grew up in Africa, and had much time to observe other species in their daily fights for survival).

1. Humanity as a species is very successful, in a local stellar system sense, but individuals might not be.

2. To remain a successful species a sufficiency of survivors is required.

3. In order to survive, a species must be adaptable, or scratch out a unique niche to occupy.

4. So it may be that AI, in its most successful evolution, keeps some portion of humanity surviving, functional and evolving. Either as equals or pets.

5. I see AI as the next step in our species' exercise to be adaptable. It may lead to our extinction, eventually. But, as we develop Super AI, we play the roles of creators, sort of little demi-gods.

6. The biggest single problem I see in this mad rush of ours to evolve, is that it seems very unlikely that we will be able to keep control of our AI invention, as it becomes more intelligent than we are.

7. The next couple of hundred years are going to be fascinating to observe. I just wish I could be a fly on the wall. If only Neuralink would hurry up and make it possible to load my psyche up to the cloud. I estimate I have less than 10 years to see that happen.

To summarize: It is likely that the majority of humanity will suffer with the advent of more and more intelligent machines, if Nature's Agenda is as I see it. Pets or batteries. Take your pick.

[zubenelgenubi: Please use quote function when quoting.]

Fair enough.

[Twark_Main]

Since we are all number persons here. Let's lay out some numbers and let us think:

One traditional GWh data center for AI on earth has initial setup costs of around 80.000.000.000$ and has self live of around 5 years, before the hardware has to be replace.

I am sure that a clever man with a lot of money can reduce that cost by using custom self produced AI chips and mass production, by some margin but assuming that will be more than 50% does not sound very realistic, especially thinking of the need to space hardening the whole setup, putting it into a rocket, make it save to 2g of acceleration .... Than there will be the extra costs of moving that hardware.

Now add the cost of terrestrial solar + batteries to power it 24/7, vs ~1/5th the solar and no batteries if you're in space.

You also delete a huge mass of steel and concrete and glass needed for terrestrial PV, which is actually a large fraction of the finished cost (and ecological impact) nowadays.

[Twark_Main]

Eh. I don't really agree, but we'll have to see what happens ... And it might be a couple decades before the answer is clear.

What I'm more concerned about in the short term is an investment bubble burst bringing down space stuff along with everything else. This could happen *regardless* of the soundness of the underlying technology - the dot com bubble of the late 90s wrecked private space projects then, though both the Internet in general and satellite internet specifically ultimately did work out.

I'm pretty sure that it'll go down a lot faster than it did with thermodynamic engines.

Compared to thermodynamic engines, AI is in the early "play around and see what works" stage.

Eventually, with steam engines, we figured out the limiting laws that governed their operation (Carnot's efficiency limit and what would later become Odum's specific power limit), and this new theoretical understanding enabled rapid progress which quickly approached these limits.

I expect we'll see the same in AI, where we develop a "Carnot's limit" for the maximum algorithmic efficiency of computation and ML (ie a software equivalent of what Landauer's Principle is for compute hardware).

Eventually the early drama blows over as the field matures, all the clever tricks get incorporated into the status quo, and it all gets boiled to a boring flowchart sizing the "turbine" for a particular information processing task.

[meekGee]

Eh. I don't really agree, but we'll have to see what happens ... And it might be a couple decades before the answer is clear.

What I'm more concerned about in the short term is an investment bubble burst bringing down space stuff along with everything else. This could happen *regardless* of the soundness of the underlying technology - the dot com bubble of the late 90s wrecked private space projects then, though both the Internet in general and satellite internet specifically ultimately did work out.

I'm pretty sure that it'll go down a lot faster than it did with thermodynamic engines.

Compared to thermodynamic engines, AI is in the early "play around and see what works" stage.

Eventually, with steam engines, we figured out the limiting laws that governed their operation (Carnot's efficiency limit and what would later become Odum's specific power limit), and this new theoretical understanding enabled rapid progress which quickly approached that limit.

I expect we'll see the same in AI, where we develop a "Carnot's limit" for the maximum algorithmic efficiency of computation and ML (ie a software equivalent of what Landauer's Principle is for compute hardware).

Yup and cyber science in the 21st century moves a lot faster than the very primitive investigation of things like combustion dynamics that limited early 20th century development of thermodynamic engines (and are only robustly solved since maybe a couple decades ago)

[Vultur]

Eh. I don't really agree, but we'll have to see what happens ... And it might be a couple decades before the answer is clear.

What I'm more concerned about in the short term is an investment bubble burst bringing down space stuff along with everything else. This could happen *regardless* of the soundness of the underlying technology - the dot com bubble of the late 90s wrecked private space projects then, though both the Internet in general and satellite internet specifically ultimately did work out.

I'm pretty sure that it'll go down a lot faster than it did with thermodynamic engines.

Compared to thermodynamic engines, AI is in the early "play around and see what works" stage.

Eventually, with steam engines, we figured out the limiting laws that governed their operation (Carnot's efficiency limit and what would later become Odum's specific power limit), and this new theoretical understanding enabled rapid progress which quickly approached that limit.

I expect we'll see the same in AI, where we develop a "Carnot's limit" for the maximum algorithmic efficiency of computation and ML (ie a software equivalent of what Landauer's Principle is for compute hardware).

Yup and cyber science in the 21st century moves a lot faster than the very primitive investigation of things like combustion dynamics that limited early 20th century development of thermodynamic engines (and are only robustly solved since maybe a couple decades ago)

There is, however, the possible opposite effect of the difference between a period of very rapid demand growth (due *both* to rapid population growth *and* incorporation of more and more of the world population into the industrial economy) and a period where both of those drivers of demand growth are far lower.

This is however a longer term issue than a potential investment bubble burst, which is a "next few years" question.

EDIT: the recent discussion of an IPO doesn't make me any more comfortable that this isn't a diversion from original core goals.

[Ludus]

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">SpaceX has way more satellites in orbit than the rest of the world combined, so maybe we know a thing or two about the subject 🤣

Starlink V3 will be 20kW and launched at scale around Q4 next year. No problem to scale that to &gt;100kW if the satellite mass is shifted towards solar…</p>&mdash; Elon Musk (@elonmusk) December 10, 2025 <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>

The shift to Space based AI seems to be the main reason for the 2026 SpaceX IPO plans. It’s widely projected at $1.5 T market cap raising $30B+.

The near term project seems to sun synchronous orbit with the V3 Starlink platform minus antennas and related electronics which is the majority of its mass then scaling power and thermal dissipation from 20 kW to 100+kW. Laser comm connecting it to Starlink net.

Sats are mass produced and launched just like Starlink V3 on Starship.

The hyperscalers are already reaching the terrestrial limits. Every new data center becomes a problem competing for limited locations that have power and local regulatory tolerance.Expanding AI compute in space can be done systematically effectively just by scaling up mass production, of Starships, Starship infrastructure, Starlink platform satellites, etc.

One obvious prediction from this is that the SpaceForce will come out with approval of Starship launch and landing facility at SLC6 Vandenberg very soon - because this requires LOTS of Starship launches into Sun Synchronous orbit.

[JaimeZX]

....as we develop Super AI, we play the roles of creators, sort of little demi-gods.

6. The biggest single problem I see in this mad rush of ours to evolve, is that it seems very unlikely that we will be able to keep control of our AI invention, as it becomes more intelligent than we are.

The biggest problem is that the demi-gods are motivated by quarterly financial reporting and not What Is Best For The Species.

Pets or batteries. Take your pick.

Iain Banks' "Minds" are best-case scenario.
Pets is almost second-best, and almost certainly preferable to our extinction.

[seb21051]

....as we develop Super AI, we play the roles of creators, sort of little demi-gods.

6. The biggest single problem I see in this mad rush of ours to evolve, is that it seems very unlikely that we will be able to keep control of our AI invention, as it becomes more intelligent than we are.

The biggest problem is that the demi-gods are motivated by quarterly financial reporting and not What Is Best For The Species.

Pets or batteries. Take your pick.

Iain Banks' "Minds" are best-case scenario.
Pets is almost second-best, and almost certainly preferable to our extinction.

We are doomed to live in what the Chinese call "Interesting Times".

What I would give to be able to be a fly on the wall for the next 200 years. Neuralink needs to get its Psyche-Upload-To-The-Cloud service launched.

[Vultur]

What I would give to be able to be a fly on the wall for the next 200 years. Neuralink needs to get its Psyche-Upload-To-The-Cloud service launched.

I am very skeptical that uploading minds is even theoretically possible, even with arbitrarily advanced technology. The brain doesn't store information in the same way an electronic computer does; how would you get all the information out without destroying the 3D structure that is key to storing it? I think you'd need something like Star Trek scanners, which probably aren't physically possible.

(Anyway even if possible it would be the Star Trek transporter problem .. it's a copy of you not *you*. It's not immortality just a nonbiological form of reproduction, a sort of mental cloning.)

[seb21051]

What I would give to be able to be a fly on the wall for the next 200 years. Neuralink needs to get its Psyche-Upload-To-The-Cloud service launched.

I am very skeptical that uploading minds is even theoretically possible, even with arbitrarily advanced technology. The brain doesn't store information in the same way an electronic computer does; how would you get all the information out without destroying the 3D structure that is key to storing it? I think you'd need something like Star Trek scanners, which probably aren't physically possible.

(Anyway even if possible it would be the Star Trek transporter problem .. it's a copy of you not *you*. It's not immortality just a nonbiological form of reproduction, a sort of mental cloning.)

All I can do is live in hope, lol. I really wouldn't be worried what version of me lives on if it became a possibility.

On the other hand, look at the technological advances we have made in the last 100 years! I sometimes think of how I would try to explain the present to a Rip Van Winkeled Leonardo Da Vinci. That is after getting him to bathe, get a haircut and dressing him in cargo pants, nikes and a hawaiian shirt. Silly, I know.

[Vultur]

What I would give to be able to be a fly on the wall for the next 200 years. Neuralink needs to get its Psyche-Upload-To-The-Cloud service launched.

I am very skeptical that uploading minds is even theoretically possible, even with arbitrarily advanced technology. The brain doesn't store information in the same way an electronic computer does; how would you get all the information out without destroying the 3D structure that is key to storing it? I think you'd need something like Star Trek scanners, which probably aren't physically possible.

(Anyway even if possible it would be the Star Trek transporter problem .. it's a copy of you not *you*. It's not immortality just a nonbiological form of reproduction, a sort of mental cloning.)

All I can do is live in hope, lol. I really wouldn't be worried what version of me lives on if it became a possibility.

Eh, I guess this is a philosophical divide. From my perspective a copy of me wouldn't be "me" in any meaningful sense, there would be no continuity of identity. (Ship of Theseus, etc )

But this is off topic.

Back on topic ... I actually don't think this is a "wait 200 years" thing. With the end of Moore's law, I think these sorts of computer based things will either work out in the next 20 years, or never. After that the primary direction of advancement probably won't be computer/IT at all.

[meekGee]

What I would give to be able to be a fly on the wall for the next 200 years. Neuralink needs to get its Psyche-Upload-To-The-Cloud service launched.

I am very skeptical that uploading minds is even theoretically possible, even with arbitrarily advanced technology. The brain doesn't store information in the same way an electronic computer does; how would you get all the information out without destroying the 3D structure that is key to storing it? I think you'd need something like Star Trek scanners, which probably aren't physically possible.

(Anyway even if possible it would be the Star Trek transporter problem .. it's a copy of you not *you*. It's not immortality just a nonbiological form of reproduction, a sort of mental cloning.)

Just like you are just a copy of yesterday's you.  The real yesterday's you died overnight, as happens every night. Yup, that conversation from early childhood - how to define continuity of consciousness... 

Good thing current plans are not dependent on such pseudo-technology.

[Twark_Main]

Summarizing, there seem to be two main arguments in favor of space-based AI compute:

    1. Fewer PV and batteries needed vs terrestrial solar. Because AI requires so much power 24/7, this is actually a big deal.

    2. It doesn't deplete the (large but still finite) anthropogenic waste heat rejection capacity of the surface of the Earth. This is the "if you tried to beam the Sun's power to Earth the planet would melt" reason.


#1 is the short-term reason, and it's the reason why Musk says space-based AI will be there cheapest option in 2-3 years.


#2 is the long-term reason, and it doesn't effect current economics because we don't have a Joule Tax yet. This would be like a coal company planning around a Carbon Tax in 1890. It's too early.


And yes, before someone says it, I'm aware that PV-powered chips don't change the Earth's radiant power balance, but it does still "leach" exergy (aka "useful work") from the biosphere and agriculture and other industrial processes. There is a finite (but large) limit on the amount of available exergy that can be siphoned off from the Earth's total exergy budget before those other systems start being starved of useful work, due to unavoidable physics (thermodynamic) constraints.



_{Hopefully we all understand thermodynamics enough that we can skip the back-and-forth about entropy vs exergy vs energy.}

[meekGee]

Summarizing, there seem to be two main arguments in favor of space-based AI compute:

    1. Fewer PV and batteries needed vs terrestrial solar. Because AI requires so much power 24/7, this is actually a big deal.

    2. It doesn't deplete the (large but still finite) anthropogenic waste heat rejection capacity of the surface of the Earth. This is the "if you tried to beam the Sun's power to Earth the planet would melt" reason.


#1 is the short-term reason, and it's the reason why Musk says space-based AI will be there cheapest option in 2-3 years.


#2 is the long-term reason, and it doesn't effect current economics because we don't have a Joule Tax yet. This would be like a coal company planning around a Carbon Tax in 1890. It's too early.


And yes, before someone says it, I'm aware that PV-powered chips don't change the Earth's radiant power balance, but it does still "leach" exergy (aka "useful work") from the biosphere and agriculture and other industrial processes. There is a finite (but large) limit on the amount of available exergy that can be siphoned off from the Earth's total exergy budget before those other systems start being starved of useful work, due to unavoidable physics (thermodynamic) constraints.



_{Hopefully we all understand thermodynamics enough that we can skip the back-and-forth about entropy vs exergy vs energy.}

#2 is true in theory but is sooooo far away that other mechanisms will kick in first.

You're talking about directly influencing the energy balance of the planet, not just messing with greenhouse gasses.

Starship: 1-100 GWatt/yr (reasonable estimate)
Lunar: 0.1-10 TWatt/yr (very handwavy)
Earth energy uptake 120,000 TWatt

So...  If we install 10 TWatt/yr over 10 millennia, we'll have ourselves a competition.

Until then, it's mostly reason #1 and some other related ones.

[Twark_Main]

Summarizing, there seem to be two main arguments in favor of space-based AI compute:

    1. Fewer PV and batteries needed vs terrestrial solar. Because AI requires so much power 24/7, this is actually a big deal.

    2. It doesn't deplete the (large but still finite) anthropogenic waste heat rejection capacity of the surface of the Earth. This is the "if you tried to beam the Sun's power to Earth the planet would melt" reason.


#1 is the short-term reason, and it's the reason why Musk says space-based AI will be there cheapest option in 2-3 years.


#2 is the long-term reason, and it doesn't effect current economics because we don't have a Joule Tax yet. This would be like a coal company planning around a Carbon Tax in 1890. It's too early.


And yes, before someone says it, I'm aware that PV-powered chips don't change the Earth's radiant power balance, but it does still "leach" exergy (aka "useful work") from the biosphere and agriculture and other industrial processes. There is a finite (but large) limit on the amount of available exergy that can be siphoned off from the Earth's total exergy budget before those other systems start being starved of useful work, due to unavoidable physics (thermodynamic) constraints.



_{Hopefully we all understand thermodynamics enough that we can skip the back-and-forth about entropy vs exergy vs energy.}

#2 is true in theory but is sooooo far away that other mechanisms will kick in first.

You're talking about directly influencing the energy balance of the planet, not just messing with greenhouse gasses.

Starship: 1-100 GWatt/yr (reasonable estimate)
Lunar: 0.1-10 TWatt/yr (very handwavy)
Earth energy uptake 120,000 TWatt

...

Until then, it's mostly reason #1 and some other related ones.

Indeed! I thought I made that clear (in fact, pointing out this distinction was my main reason for posting), but it's always good to re-iterate and re-phrase the point.


"#2... would be like a coal company planning around a Carbon Tax in 1890. It's too early."

"(large but still finite) anthropogenic... heat capacity"


We may wish that fossil fuel companies had the foresight to address global warming from the start. Well Tesla is doing exactly what we might wish, but the popular reaction is instead to ridicule the fix as "uneconomical."

Fortunately, reason #1 is enough to make space-based AI economical even today. But you gotta give points for forward thinking... 

So...  If we install 10 TWatt/yr over 10 millennia, we'll have ourselves a competition.

Growth is exponential, more like 2% CAGR.


Thanks. A rare opportunity to use "exponential" in its real math meaning instead of "very big." 

[meekGee]

Summarizing, there seem to be two main arguments in favor of space-based AI compute:

    1. Fewer PV and batteries needed vs terrestrial solar. Because AI requires so much power 24/7, this is actually a big deal.

    2. It doesn't deplete the (large but still finite) anthropogenic waste heat rejection capacity of the surface of the Earth. This is the "if you tried to beam the Sun's power to Earth the planet would melt" reason.


#1 is the short-term reason, and it's the reason why Musk says space-based AI will be there cheapest option in 2-3 years.


#2 is the long-term reason, and it doesn't effect current economics because we don't have a Joule Tax yet. This would be like a coal company planning around a Carbon Tax in 1890. It's too early.


And yes, before someone says it, I'm aware that PV-powered chips don't change the Earth's radiant power balance, but it does still "leach" exergy (aka "useful work") from the biosphere and agriculture and other industrial processes. There is a finite (but large) limit on the amount of available exergy that can be siphoned off from the Earth's total exergy budget before those other systems start being starved of useful work, due to unavoidable physics (thermodynamic) constraints.



_{Hopefully we all understand thermodynamics enough that we can skip the back-and-forth about entropy vs exergy vs energy.}

#2 is true in theory but is sooooo far away that other mechanisms will kick in first.

You're talking about directly influencing the energy balance of the planet, not just messing with greenhouse gasses.

Starship: 1-100 GWatt/yr (reasonable estimate)
Lunar: 0.1-10 TWatt/yr (very handwavy)
Earth energy uptake 120,000 TWatt

...

Until then, it's mostly reason #1 and some other related ones.

Indeed! I thought I made that clear (in fact, pointing out this distinction was my main reason for posting), but it's always good to re-iterate and re-phrase the point.


"#2... would be like a coal company planning around a Carbon Tax in 1890. It's too early."

"(large but still finite) anthropogenic... heat capacity"


We may wish that fossil fuel companies had the foresight to address global warming from the start. Well Tesla is doing exactly what we might wish, but the popular reaction is instead to ridicule the fix as "uneconomical."

Fortunately, reason #1 is enough to make space-based AI economical even today. But you gotta give points for forward thinking... 

So...  If we install 10 TWatt/yr over 10 millennia, we'll have ourselves a competition.

Growth is exponential, more like 2% CAGR.


Thanks. A rare opportunity to use "exponential" in its real math meaning instead of "very big." 

Awright!

---

Meanwhile double checking:

Start with 1 GWatt/yr and grow by 10% every year, getting to 100,000 TWatt is a 170 years.

So yup, about the same time as from the industrial revolution to now.

Though, chat argues that worldwide energy use, since 1800, only grew by a factor of 30, not 100,000,000...

(I pushed back a bit but he's defending that 30x pretty well.)

[mikegi]

Your brain and mine require something like 20 to 25W of power. I don't think that the current models of AI capture what's really going on (by many orders-of-magnitude, even correcting for the magnitude of analog-to-digital conversion). I don't know the correct algorithm, but I do know that this isn't it. The current algorithms *might* be equivalent to the retina, but that's about it.