AI needs 100× more compute by 2030. Earth is out of room. Space looks like the answer until you run the physics, the economics, and the carbon math nobody wants to have.
Moving AI to space feels more like an 'accountability detour.' We’re letting a few companies own the 'visionary future' while the rest of us are left to deal with the brutal physics of orbital pollution and resource extraction. Another case to ask who is actually responsible for the mess we're planning to make in the stars. Thanks for writing this, Farida!
Accountability detour” is the perfect phrase for this. It links directly to Oliver Williamson’s transaction cost economics, which maps how firms structure deals to offload costs they can’t price onto suppliers, the public, or the environment. Great article, Farida!
I was asked to write about the economics of space datacenters. As I dug deeper to find reliable sources and make an informed assessment, I ended up having fascinating conversations with experts in tech and physics. That exploration brought me to a central question: is it truly worth it? Are we financing a visionary dream or a ONE MAN Dream
And worse, China often seems to come out ahead in the end outperforming. Elon Musk is famous for his hype, turning science fiction into perceived reality for man but from an economic standpoint, his ventures are often far from profitable.
Thank you Farida. I had a similar moment of misplaced confidence years ago when I assumed moving servers to a colder climate would solve a cooling bottleneck, and it turned out the issue was never ambient temperature, it was convection, which cold air still handles, but vacuum absolutely does not. The heat has nowhere to go. That lesson stuck with me but I never extended it to orbital thinking until just now.
The framing of "exporting infrastructure crises" rather than solving them is the part I'll be sitting with.
I had a similar moment , for the beam and lights , how it can be secured, and that led me to go learn more about it, and I have to admit, that I didn’t cover all the complexity of it, didn’t want to sound alarmist , while the ambition is wrappped by a good faith , it is handled by greed and speed , the 2 traits that nature doesn’t like
Thank you, Farida. This is a brilliant breakdown. The only thing that makes me think about this as a potential solution is the same doubts I have with regards to terraforming of the planets. Surely it would be better if we could work out the problems on our own first...
The terraforming parallel is the most honest comparison anyone has made, thank you for that. The uncomfortable truth is that "fix problems at home first" is always the right answer and almost never the one humanity chooses. The moon race didn't wait for poverty to be solved. The internet didn't wait for literacy to be universal.
The pattern isn't new. Neither is the bill, which is always paid by people who weren't in the room when the decision was made.
this is why I ended the article with a question to address that point
Thank you Farida for this, much needed breakdown. I was just reading about how many data centers are going to build in the USA. 🤦🏻♀️ Too many! And the talks about building data centers on the moon?! To me is absolutely foolish.
It is ambitious and challenging, but it could be achievable if we approach it with reason and logic rather than speed and greed. At the end of the day, the final word and the upper hand belongs to Nature.
What concerns me most is the race for it. I didn’t even touch on climate change, debris, or radiation. I intentionally kept the discussion technical so people could stay focused without being overwhelmed by physics and the many other factors that will play a major role in the process.
Such a cool article. Of everything you ran the numbers on, which constraint do you think is most likely to be the actual dealbreaker in the next decade?
The part that stuck with me: we're basically saying "Earth's power grid can't handle it, so let's build one on the Moon" instead of, you know, fixing the power grid. It's the tech equivalent of moving to a new apartment because you dont want to do dishes.
That said, I'm curious whether the author thinks theres *any* near-term case for off-planet compute (like genuinely latency-sensitive workloads), or if the whole premise is just expensive avoidance theater. Because I suspect the real answer is boring: we'll do both, hyperscalers will greenwash the Moon thing, and meanwhile someone in Waterloo will quietly solve 40% of this with better chip design.
I don’t get to have such conversations with the folks, as wonderful as they are, around here that spark such ideas as you. I have to rely on yourself and others like you, Farida, on social media. So, thank you.
To me it’s interesting that we have been exposed for some 60 years now to futurist thinking but of a much more benevolent and hopeful kind. There will be space stations orbiting Earth, endless power will be found, satellites and planets will be colonized, even terra-formed for our habitation. But nearly all of that has been viewed as positive with a unified humanity, calm demeanor and clarity of shared purpose. Certainly there have been some dystopian thinking where the basest of human depravity have been exposed in frightening detail displaying a life in the future not unlike life as it is now ruled by demigods, gangsters and mind controllers.
It seems to me that the benevolent era of space exploration and useful emerging technologies in real life with the landing on the moon, with the Space Station still in operation, are being supplanted by the very real threat in real life of a dystopian view when the names Musk, Thiel, Ellison are brought up. Again it’s the selfish ‘me’ versus the selfless ‘we’ humanity has struggled with for eons.
But I have no doubt, though it looks improbable today, that technologies will do many unbelievable things. From Kitty Hawk to X-1 to Apollo, from ENIAC to pad and phone, all in such a short time.
Question is will we become Federation or Klingon. 🤪
Nature will give us plenty of room if we play nice, so to speak. If we don’t play nice and try to force things, Nature will slap back hard. I think we intuitively know this, it’s when we deviate from the path towards our human vices, greed, avarice, etc, that we fall into trouble.
Moving AI to space feels more like an 'accountability detour.' We’re letting a few companies own the 'visionary future' while the rest of us are left to deal with the brutal physics of orbital pollution and resource extraction. Another case to ask who is actually responsible for the mess we're planning to make in the stars. Thanks for writing this, Farida!
Thank you for the discussion that led to this article
we will have to wait to see what the nature will say. nature has the upper hand.
Accountability detour” is the perfect phrase for this. It links directly to Oliver Williamson’s transaction cost economics, which maps how firms structure deals to offload costs they can’t price onto suppliers, the public, or the environment. Great article, Farida!
Thank you so much
Love this breakdown! SatDevOps is definitely a different beast, thanks for laying out the engineering challenges so clearly! 🙌
I was asked to write about the economics of space datacenters. As I dug deeper to find reliable sources and make an informed assessment, I ended up having fascinating conversations with experts in tech and physics. That exploration brought me to a central question: is it truly worth it? Are we financing a visionary dream or a ONE MAN Dream
Could be just a side effect of all the easy money slushing into AI right now.
Like how low interest rates funded business models and business growth that never would have worked otherwise.
And worse, China often seems to come out ahead in the end outperforming. Elon Musk is famous for his hype, turning science fiction into perceived reality for man but from an economic standpoint, his ventures are often far from profitable.
Thank you Farida. I had a similar moment of misplaced confidence years ago when I assumed moving servers to a colder climate would solve a cooling bottleneck, and it turned out the issue was never ambient temperature, it was convection, which cold air still handles, but vacuum absolutely does not. The heat has nowhere to go. That lesson stuck with me but I never extended it to orbital thinking until just now.
The framing of "exporting infrastructure crises" rather than solving them is the part I'll be sitting with.
I had a similar moment , for the beam and lights , how it can be secured, and that led me to go learn more about it, and I have to admit, that I didn’t cover all the complexity of it, didn’t want to sound alarmist , while the ambition is wrappped by a good faith , it is handled by greed and speed , the 2 traits that nature doesn’t like
Thank you, Farida. This is a brilliant breakdown. The only thing that makes me think about this as a potential solution is the same doubts I have with regards to terraforming of the planets. Surely it would be better if we could work out the problems on our own first...
The terraforming parallel is the most honest comparison anyone has made, thank you for that. The uncomfortable truth is that "fix problems at home first" is always the right answer and almost never the one humanity chooses. The moon race didn't wait for poverty to be solved. The internet didn't wait for literacy to be universal.
The pattern isn't new. Neither is the bill, which is always paid by people who weren't in the room when the decision was made.
this is why I ended the article with a question to address that point
"Everyone has a plan until they get punched in the mouth" - Mike Tyson.
At some point nature is going to fight back. And we won't like it.
I love this quote.
I agree that Nature has the upper hand and final word
Thank you Farida for this, much needed breakdown. I was just reading about how many data centers are going to build in the USA. 🤦🏻♀️ Too many! And the talks about building data centers on the moon?! To me is absolutely foolish.
It is ambitious and challenging, but it could be achievable if we approach it with reason and logic rather than speed and greed. At the end of the day, the final word and the upper hand belongs to Nature.
What concerns me most is the race for it. I didn’t even touch on climate change, debris, or radiation. I intentionally kept the discussion technical so people could stay focused without being overwhelmed by physics and the many other factors that will play a major role in the process.
I’m glad it resonated with you.
Excellent insight!
Thank you @Joel
Big ambition means nothing if the math never works.
agreed, but when it is promoted and presented as the golden ticket.. then math becomes a secondary factor
Such a cool article. Of everything you ran the numbers on, which constraint do you think is most likely to be the actual dealbreaker in the next decade?
the laser beam security, who will have the accountability if a Data center disappear in a vacuum due to a nature hazards ?
The part that stuck with me: we're basically saying "Earth's power grid can't handle it, so let's build one on the Moon" instead of, you know, fixing the power grid. It's the tech equivalent of moving to a new apartment because you dont want to do dishes.
That said, I'm curious whether the author thinks theres *any* near-term case for off-planet compute (like genuinely latency-sensitive workloads), or if the whole premise is just expensive avoidance theater. Because I suspect the real answer is boring: we'll do both, hyperscalers will greenwash the Moon thing, and meanwhile someone in Waterloo will quietly solve 40% of this with better chip design.
I don’t get to have such conversations with the folks, as wonderful as they are, around here that spark such ideas as you. I have to rely on yourself and others like you, Farida, on social media. So, thank you.
To me it’s interesting that we have been exposed for some 60 years now to futurist thinking but of a much more benevolent and hopeful kind. There will be space stations orbiting Earth, endless power will be found, satellites and planets will be colonized, even terra-formed for our habitation. But nearly all of that has been viewed as positive with a unified humanity, calm demeanor and clarity of shared purpose. Certainly there have been some dystopian thinking where the basest of human depravity have been exposed in frightening detail displaying a life in the future not unlike life as it is now ruled by demigods, gangsters and mind controllers.
It seems to me that the benevolent era of space exploration and useful emerging technologies in real life with the landing on the moon, with the Space Station still in operation, are being supplanted by the very real threat in real life of a dystopian view when the names Musk, Thiel, Ellison are brought up. Again it’s the selfish ‘me’ versus the selfless ‘we’ humanity has struggled with for eons.
But I have no doubt, though it looks improbable today, that technologies will do many unbelievable things. From Kitty Hawk to X-1 to Apollo, from ENIAC to pad and phone, all in such a short time.
Question is will we become Federation or Klingon. 🤪
thank you so much for your kind words, i have to credit Aaron for challenging to investigate it.
well, ambition is validated , science can achieve it, however, Nature will have the final word, and that enough to give us hope.
Nature will give us plenty of room if we play nice, so to speak. If we don’t play nice and try to force things, Nature will slap back hard. I think we intuitively know this, it’s when we deviate from the path towards our human vices, greed, avarice, etc, that we fall into trouble.