But if that would be a real market case, then Elon's solution is actually the optimal one, and very similar to the possible approach which more knowledgeable space analysts said for years. Basically it is super pointless for humans now to have big manned space stations for any economical purposes. We should build them of course, but they will be a net loss for centuries. If any space research or space manufacturing would be required by the market, then one-off space missions are the only rational way to do it cheaply. Something like the platform showcased here, or something bigger, like a converted Dragon/Cygnus/etc. ship into a mini lab. Commercial Space Stations is a bunk program, specifically in the word "Commercial".
leave it in space then; massive star ships shouldnt need to come down to planets anyway
LOL.
Hydrolysis can be used to make Hydrogen, so as long you have enough ice. You fill it up with fuel needed to go up again, and send in down with the gravity well with the payload.
This isn't more absurd, but may offer a potentially much better payoff than microgravitated materials research
Evolve bony vertebrates to thrive in microgravity. Use that learning to genetically enhance humans- America. Much harder since the elections. Not appealing.
- Europe. Doesn't have enough electricity with anti-AI, anti-big tech and high tax governments. Not appealing.
- Latin America. Expensive, likely to have your assets appropriated by socialists. Not appealing.
- Africa. No.
- Middle East. Your GPU cluster might get blown up. No.
- China. To some extent yes. Plenty of power and space, pro-AI government, but many companies don't want to send their IP to a country famous for copying it and they're still under GPU sanctions.
- Japan, Korea, etc. Maybe. They're not really famous for big datacenters though.
- Australia, Canada. Possible, but you're exposed to Europe-style political risks.
There just aren't many places on Earth where it's realistic to build and operate AI datacenters once political issues and grid constraints are taken into account.
Did they though? From what I’ve read the places where they build get lots of money and infrastructure (not only DC-related) from them.
https://www.wired.com/story/amazon-says-it-is-going-to-stop-...
https://www.wpr.org/news/microsoft-to-stop-using-ndas-with-l...
https://www.npr.org/2026/08/27/nx-s1-5879528/data-center-nda...
I mostly agree with you, and obviously there is a lot of intelligence in the valley, but there doesn't seem to be much wisdom.
Because the backlash isn't actually about the data centers, it's about AI. 80% of the complaints about data centers are nonsense and the remaining 20% only apply to the ones where they're doing something that isn't actually required, like burning fossil fuels to power them.
Which you can tell because the demands look like "stop building data centers" rather than "to build one you have to add 110% as many watt hours in new non-fossil fuel generation to the grid as you use".
Maybe they didn't predict the magnitude and vitriol of the backlash, but NIMBY resistance to all infrastructure projects has only been growing over the years. For example, wrt solar and wind power. It was obvious data centers were going to see at least as much resistance because you need to build the data centers and more power generation.
Do data centers in space make less sense than build outs in more accommodating countries? Not necessarily. It's not an either/or thing, and it all depends on the concrete bottom line numbers. In space you don't have to worry about managing political winds at all; you control everything. More accomodating countries tend to be poorer with worse infrastructure, so there are more unknown risks. And you don't need to develop or manage a foreign workforce, which is why 100% of Silicon Valley jobs didn't instantly jump to India or Poland.
I agree there is a profound lack of wisdom in tech, and they seem to be proud of it… like they believe they’ve discovered a workaround.
> I agree there is a profound lack of wisdom in tech, and they seem to be proud of it… like they believe they’ve discovered a workaround.
Even in discussions here I'm constantly put off by comments that are this hand-rubbing talking to oneself about how the "AI haters" are all stupid, or how they can be forced to use or consume things they don't want against their will. It kind of baits all the sociopaths out into the open, because if their deus ex machina is about to show up, why bother with the tedious masking?
> I don’t think you have to be dumb to fail to acknowledge reality.
No, but you can be emotionally stunted to force you to be dumb. Call it narcissism or whatever, but no sense of others or self = constant need to rewrite the world around you, and fearing logic and honesty like a vampire would fear sunlight. A dumb person can learn, people who are forced to act dumb cannot. They are still, effectively, very dumb. They will go to their graves dumber than an illiterate or mentally handicapped person would have. Not knowing a bunch of things that are true is meh... but thinking you know a trillion things that are all just the scab on a wound that happened in childhood, none of them true, and one dumber than the other, now that is astonishingly dumb.
https://www.govtech.com/question-of-the-day/how-many-anti-da...
Whether it’s A16Z tweeting about how the value of all white collar wages is their market or Anthropic basically saying entry level jobs are done.
In seriousness this is an example of the previous point. Rather than engaging with the issue the tech/DC industry seems to think they can wordplay their way out of it by quibbling about irrelevant technical details and people are primed to get extremely angry at grinfucking type behaviour nowadays.
If you have the technical chops to set up a data center in space — with no connections to the electrical grid or municipal water, you can use your technology to build your data center with the same specs right here on earth in 90% of places that you think are issuing "bans". You'll also get much better cooling for free via convection.
A remote town with a population of 43 will be happy for a data center to employ 2 locals, but data center operators aren't doing that today because they want tax breaks tiny towns can't give, infrastructure tiny towns don't have, nevermind a well-armed police force to deter would-be GPU thieves. Space is a red herring.
Remote towns with populations of 43 are exactly the sorts of places that are pushing back on having datacenters in their areas.
And space is the ultimate physical protection. In reality a bigger issue than GPU thefts are protests or leftist groups trying to firebomb you. See the problems the Tesla factory in Germany has had with leftist violence, like setting its electricity pylons on fire (because electric cars "consume natural resources and are not sustainable", lol). The left has become very anti-AI and so the risk of attacks on facilities, especially if built outside the USA, is quite high which in turn means you have to overbuild physical security like high fences.
Whether they can be worth the money it cost to put them in orbit is another story.
Apples to oranges, long-lived satellites have modest power needs. Doing a back-of-the napkin math on required size of solar array and radiative elements for a single-rack's worth of energy is a sobering exercise for those who are a overly exuberant on the topic. The all-in power estimate for a single GB300 chip is 2.1kW
Edit: ISS is about 90KW steady state, and the collector to radiator area ratio is about 5:1. Dunno about the mass ratio.
My point is not that space data centers are viable or sensible. Only that the need for cooling is limited. If a terrestrial datacenter was powered only by its rooftop solar panels, cooling it would be trivial. Space is harder, but there just isn’t that much heat to dump.
A simple rock in orbit radiates exactly as much heat as hits it from the sun, on average.
You are correct that the first law of thermodynamics dictates the energy, but anyone who has parked a car in the sun will tell you, the temperature where the absorption/dissipation equilibrium is achieved is often incompatible with the survival of the cargo.
edit to your edit:
> Edit: ISS is about 90KW steady state, and the collector to radiator area ratio is about 5:1. Dunno about the mass ratio.
Cool; that's about half the wattage of one modern full-height Nvidia AI rack (172kW) - you need two ISSes to service just 1 rack; extrapolate that to a data center's worth of racks and see if the math is mathing. The ISS is about the size of a(n American) football field.
You won't immediately freeze if you enter space without a suit. You'll die almost immediately, but it will be because your blood will begin to boil because of the lack of pressure kinda like an extreme form of divers sickness - it's just not the cold that'll get you first.
You know space rocks where we live don’t melt because the Earth and moon don’t melt.
Space datacenters are just SDI2, a convenient bullshit to be able to funnel more money to campaign donors.
As long as there’s space on earth, Space Datacenters can never be economical. It’s rally that simple.
We’re not talking about launching little satellites here, we’re talking about replacing giant buildings.
The radiators are not a big deal. The solar panels required are much bigger. You can just run the electronics really hot to improve the efficiency of the radiation. Nobody is going to get close to those chips so the usual safety requirements don't apply.
You don't service it. Over time the chips degrade and become less useful. Eventually you deorbit it and it burns up. It's a sealed unit, not meant to be replaced.
The chip temperature isn’t limited by safety issues (unless self destruction of the chip counts as safety to you). Certainly not to protect the people around it.
Data centers in brownfields and urban environments see near-zero resistance. There are data centers in skyscrapers in the center of most major U.S. cities, and nobody cares.
Stop plowing down the forests, slurping up the aquifers, and digging up the prairies, and people will stop complaining.
Put a data center next to BoAT, and nobody cares.
Realistically you can only hear it if you're within 30-50 feet of the data center, unless there's a separate diesel/natural gas generator.
Not visited one in the middle of the day. Lived next to one 24/7.
A recent law also gave data centers the last priority for grid access. The idea was to prioritize productive industry over less productive uses of electricity. Even something with as little added value as steel production tends to employ ~10 people (directly or as subcontractors) and produce ~€10 million/year of exports for every MW of average power consumption.
a. put them a few miles outside urban centers
b. put them in space
Which is a cheaper, better option?
Also pretty sure there will always be empty land somewhere on earth where no one cares. It will always be easier to build and maintain even in a desert than in space.
1. Everything rusts.
2. There is no permanent power source, unlike in sun synchronous orbit where you have 24/7 daylight.
3. The US has no ship building industry of any size.
4. Storms can destroy things.
There is no empty space on Earth unclaimed by governments. Given the results of the American midterms essentially all land is currently controlled by left wing governments, and the left is very anti-AI. Only China really wants to build AI at scale but they don't have enough chips yet. That might change, but even then not many companies want to upload all their IP to China.
So you can forget building more or less any large industrial projects on Earth for the foreseeable future. It may not matter. There appears to be a GPU glut starting anyway, so who knows if anyone will even need new datacenters for a while?
From Wikipedia, at 300km the atmosphere is around 80% atomic oxygen https://commons.wikimedia.org/wiki/File:Msis_atmospheric_com...
A new dimension to optimize for, that space is going to incentivize, is highest stable operating temperature.
Higher temperature operation sheds heat energy via radiation much faster.
Some post-silicon substrate research is aimed at wafers that can operate at extremely high temperatures.
I imagine given how often materials research pulls a rabbit out of a hat, there is a lot of head room for temperature.
Isn't this what heat pumps are for? Then you can run the radiators at higher temperatures without needing to run the electronics at those temperatures.
There’s probably an optimum that involves a heat pump, but I don’t think it’s vastly better than not using a heat pump, and it’s certainly heavier and more complex.
Power and heat radiation are the solved problems at scale in LEO. 50+% of solar energy is attenuated by the atmosphere, so panels can be huge and thin and flimsy in zero-G and get twice as much power as on the ground. Space-Based solar power plants are being actively explored by the US, UK, Chna, and Japan. The real hitch is how to send the power to the ground via microwave without also needing a receiver 20+ km across that lightly crisps birds flying above it.
If the thing that needs the power is in orbit already, all you're doing is sending data, something Starlink already has figured out.
The impractical part is shielding, maintenance, and a modular design that can be remotely maintained from the ground. When you're already running a high-cadence launch schedule, those costs and issues narrow, and you can swap out a component within days.
Again, it makes sense on paper. It might not be more practical than a data center on the ground, but there would be trade-offs.
What’s the practical solution to heat radiation that doesn’t require running your chips at infeasible temperatures?
Even if all the technical issues can be worked out, which I doubt, the incentives are there to pump as many as possible up there. I don’t think we want to live with that raining down on us all.