> This feels like the result of a competition to design the worst possible user interface. To about 5% of people it might be an accessibility feature, to everyone else it's worse, and people with beards, marks, or dark skinned faces are going to find it a disaster.
You say that, but I have seen in the wild a scroll gesture to increase or decrease the value of a telephone number.
Wasn't even capped at zero, so I could scroll to a negative (phone) number.
Indeed. I may have to remember this for the next time someone says America's freedom of speech means people don't go to jail just for posting stuff on the internet.
The difference is she was a mayor of an American city. She was an elected official in direct contact with CCP officials, and while the examples provided are more or less trivial, I would hope that she would be investigated. A US official has access to certain private information and it's not far-fetched to conceive she could be disseminating private information about, at the very least, local politics
The problem with the industrialisation analogy is that it ought to have already happened.
I'd already been coding from textbooks for a decade by the time my formal education included it, but that formal education began with VisualBasic.
Of all the apps I've been paid to work on, the only ones which you couldn't have done just fine in something like VB were the games, because they were performance limited.
Everything else, well, things like REST APIs and JSON are implementation details that don't fundamentally matter. The UI and UX team should have been able to plug in the relevant values for how to talk to the database like they set a Figma design's colour codes, without needing to care if those things meant a database was on-device or remote. Reactive UI and its bindings are *exactly* the kind of thing best done graphically rather than in code, even if you still need a database administrator for the database to work well.
The actual real hard stuff? Last time I was paid to actually make trade-offs to figure out optimisations I needed to make was 2012 or so. To the extent that one of my recent agentic coding experiments was to give myself such constraints once again (turns out you can find the travel time from any start location to all other locations in Berlin in 60ms in a web browser if you try).
The real work has become the politics of fashion, not of good code but of which framework and design pattern is in vogue.
I believe the proponents think it will make all the content better, not worse, but the current state of affairs is indeed definitely worse.
But I essentially agree with you, the destruction of all white collar and creative jobs is something the proponents hand-wave too easily with vague mention of UBI as a thought-terminating cliché rather than a real policy proposal.
(Myself, I'm in a strange position with doppelgängers, because I simultaneously want real human connection and keep getting disappointed with many of the real humans).
> (Myself, I'm in a strange position with doppelgängers, because I simultaneously want real human connection and keep getting disappointed with many of the real humans).
I don't think you're in a strange position at all. You are just being honest about what is the undercurrent of much of modern technological development. AI is simple the apex of a trend to replace human interaction with a simulacrum, which without your awareness, reduces your own tolerance for the frustrations of real human interaction, thus making you more dependent on the simulacrum.
> I believe the proponents think it will make all the content better, not worse
I believe the proponents think it will make content more profitable, especially in the sense of concentrating the wealth that can be extracted from content creation. Related, there’s a desperate attempt to redefine code as “content”, at least implicitly, so that it can fall under that umbrella as well.
A fantastic way to break your economy. Even the extra half trillion/year Trump wants to spend on the military is increasing investor concern about buying US treasury bills.
I've seen it suggested that the IPO they're targeting (plus the similarly-timed ones from OpenAI and Anthropic) may be so high as to break the financial system. Investors need to have a trillion dollars available to invest if the stuff being added to the market is going to be worth a trillion dollars.
(Phrased that way because while I hear they're targeting 1.5T valuation that doesn't mean they'll be selling 1.5T of shares).
> Current mix of electric generation is not going to get 37 times cleaner in 10 years. That's the ratio we're talking about right now.
Doesn't need to: because data centres are already power-constrained, the correct question is marginal new supply in each case.
If you build 25 GW of new data centres (which seems to be roughly the scale being proposed here), the options are (1) build more power plants or (2) have brownouts and/or rolling blackouts.
What are those new power plants going to be? Even though Trump hates renewables, they're now the cheapest new power source, which was already necessary to be even considering putting 25 GW of PV onto satellites in the first place.
This makes it a question of what's cheaper: over-provisioning and storage, or launch costs?
> 1. Demand for AI compute will continue to grow for the next 10 years. 2. The cost of orbital datacenters will approach the cost of terrestrial datacenters (in $ per token terms).
> If you don't buy either of these two premises (and I agree that neither is guaranteed) then you don't have to worry. No one is going to waste money on orbital datacenters if they aren't profitable.
I'm glad you said the final quoted paragraph; while it has always been difficult to tell which websites are accurate vs. slop, and moreso today with fully automated slop, I see claims the current rate is doubling between 15 and 3 months. Even at the slowest of these, 15 months, this gets 2^(120/15) = 256x growth in 10 years, which would raise it significantly above current total global electrical demand: https://www.iea.org/reports/energy-and-ai/energy-demand-from...
> A 100MW datacenter will emit 1.5 million tons of CO2 over a 5 year lifetime (given average USA energy sources). In contrast, 10 Starship launches (~1,000 tons to orbit) will emit no more than 40,000 tons of CO2. Almost all other environmental effects will be proportional.
Not so: given that existing power capacity is already a constraining factor for new ground-based data centres, the alternative to be considered is not the CO2 emissions of existing average USA energy sources, but the CO2 emissions of new energy sources. This may be many things, but given that renewables are now the cheapest new energy, if politics stops renewables in the US it just means the data centres (especially at this kind of scale) aren't going to be in the US, but it doesn't say that they'll be in space specifically.
From what I've seen (Google paper, IIRC), it only makes sense economically if Musk's stretch goals for price/kg to orbit for Starship become true.
Technically it's fine, just take something like Starlink and use most of the power for compute rather than for comms.
But financially, it depends on price to orbit being extremely low; not just lower than Falcon, but as low as Musk's best public claims about what may be coming at some point.
Land is expensive, water is scarce, people don’t want sound pollution anywhere near them.
Building a datacenter in the neighborhood is already unpopular enough that companies do tricks to prevent public from knowing what is being built and by whom in advance.
Sending a small box with a panel to space may be a solution if a: the inside of the box is expensive and the cost to launch is cheap.
You amortize the box over 2-5 years and burn it in the atmosphere afterwards.
If the math is mathing, multiply by a million and voila, you have a datacenter in space where each rack is flying separately.
With a regular compute it may not be profitable but with GPUs connected to each other by optical links? I think it may be possible.
In the Sahara there is plenty of space and plenty of solar. Any heat you can radiate away in space, you can radiate away on Earth. Or, more simply, dig into the ground and pump heat into the cool earth.
In theory it "should" be much easier to build on earth, but in some ways it's just different challenges. On earth you're forced to deal with those pesky government things and in the Sahara not a lot of them are exactly reliable or good-faith actors. Then there's night time. So out of the gate, you're dealing with needing massive power storage for the night time.
So you invest $5b into a solar farm and data center outside of Tunis and 5 years after you finish construction a popular uprising topples the government and now you're dealing with new management? Nah, nobody is going to do that. And who's going to work there? How are you going to get data out of there? You're going to end up using satellite comms anyway. It's not 1953 any more and (thankfully) nobody is in a position to "Operation Ajax" your popular uprising when that happens too. I mean, maybe, but yeah, I would not do it.
Even in relatively stable places like the US or the EU, let's say you bought some random parcel of land in the New Mexico or West Texas desert region. Or even in Southern Spain or something. Even if you get the land cheap, with relatively easy fiber access (doubtful, but whatever), you're still beholden to the communities there. You think they have spare water to cool your facility? How are the schools for the kids of the engineers working there? You think that people are going to be head-over heels in love with Amarillo or Extremadura? The land is cheap because people don't want to live in these places, so picky people are going to steer clear. And at the end of the day, you're still going to have to get everything permitted, approved, stamped 80 times, and the project will grind on for months.
No, space is an end run around dealing with bureaucracy and politics. It's space. There's basically nobody to tell you "no" up there. You can park the satellites in a sun synchronous halo that lives about on the terminator, and just pull in power constantly and radiate directly away from the sun in the other direction. It's going to be expensive, it's going to be technically challenging but we will do it. Also, think about the California high-speed rail stuff. If you try to build on earth you're going to be permitted and social-media'd to death anywhere on earth you decide to build one of these. For better or worse people hate Elon. I mean, I understand it, he's kind of insufferable and his dalliance with politics was a bit of a disaster (seriously, USAID cuts are killing people), but he's certainly no moron and I do not think he's entirely un-selfaware. He knows that people aren't going to let him build these wherever he wants. You're going to have to ask for permission thousands of times, there's going to be social media campaigns to stop him, literally any screw up (his fault or not) is going to be loudly shouted to everyone. If he decides he wants to expand his facility, that's more permits, more restrictions, more permissions.
So they'll go to the place where they do not need to ask anyone. Initially that was red states or at least relatively "non-hostile" states like where Tennessee, but even there people will squawk about it. I don't mean to say "squawk" to dismiss those folks, well, maybe I do, but I just think it's a bit silly in the context of us burning gazillions of gallons to bomb the Iranians. Nobody is going to do a damn thing about the climate or anything right now, and stopping data centers from getting built feels like stepping over dollars to pick up pennies, but I digress.
Anyway, space has none of those problems. Indeed, the problems are almost all technical. The technicians and engineers can live in California, or work remotely from anywhere really, and you won't have to deal with increasingly well funded and clever NIMBYs. The real challenge is going to be finding optimal launch sites for this stuff. Hilariously, my neck of the woods up here in Alaska is uniquely suited to launch into inclinations that would allow for constant sunlight. It's what, 98 degrees inclination for an SSO? So you can launch launch north out of Poker Flat and south out of Cape Chiniak. Though we don't have the infrastructure up here to support that out of Poker Flat yet. And nobody will squawk too loudly about it up here. I think those lunatics trying to slingshot satellites into space are trying to launch out of Adak too, so, hypothetically, that's an option as well other than the logistics of getting vehicles up here.
Anyway, this has turned into a bit of a book report, but these companies are not optimizing for cost savings right now, they're optimizing to avoid people telling them no.
>> No, space is an end run around dealing with bureaucracy and politics. It's space. There's basically nobody to tell you "no" up there.
>> So they'll go to the place where they do not need to ask anyone.
>> Anyway, space has none of those problems. Indeed, the problems are almost all technical.
This is pretty naive. What happens when one of the other sovereign nation destroys your space assets or holds them hostage. There is also no defense in space.
You think Grand Forks ND or Tempe Arizona is going to say, “we’re going to shoot down your datacenters?”
Of course not. The only people to stop you is like 6 nation states that have the capability to tell you no, you know? Maybe less? And most of them all need your launch capabilities?
Cmon. Who is going to tell them no? The US government? And jeopardize NRO satellite launch abilities or whatever? No, the Feds won’t stand in the way.
Destroying a satelite is much easier than launching one, even with existing systems. Worse, given rate of improvements, I think we're going to get ground-to-orbit anti-satelite lasers before 10% of this constellation gets launched.
And at least one of the nations with the existing military capacity to make a "no" stick is currently considering criminal charges against Musk personally, while another has a long history of assassination including of their own oligarchs.
I do not think this is true? I mean, the point seems valid, but how many countries have ASAT capabilities? Then how many countries actually want to use that? Then even if we could get ground-to-orbit ASAT lasers (obligatory "pew pew"), then what countries will push that particular button?
I'm sorry, at least in this exact moment I don't see it. Even if france wants to bring charges against Musk, are they going to start downing his satellites? That seems like a leap. This kind of further illustrates my point. In space, you're literally kind of "above" the fray...
Almost entirely depends on how much it costs to deal with the ground having night, and if this is more or less than the cost of putting it somewhere that doesn't have night.
Both are already things that can be done in principle, the question is just how expensive the solutions are.
For scale: if these million satellites were 25kW each, that's 25 GW total; Tesla supplies about 150 GWh of batteries each year between cars and PowerWall units, so provided they didn't need replacing more than every four years this would be enough to supply a data centre that size for 24 hours, so you'd just need to put this all somewhere without much cloud cover.
Solar capacity factor is 10-20%. So your state of the art chips are utilized 10-20%. That just makes no sense. Adding batteries help, but does not solve it entirely.
787 is vastly more expensive than Starship, but Starship uses 7x more fuel to get to orbit than a typical 787 flight. _IF_ Starship can achieve same re-usability as 787, then cost to orbit will be like $7.5/kg.
While true, this is insufficient to make the new claim credible. If the proposed satellites only weighed 100kg and remain on orbit for 3 years, to keep a million up requires:
They've been approved for 44 Starship launches from Kennedy Space Center in Florida, and are aiming for 160 total launches in 2026. They've recently purchased a giant tract of land in Louisana to build a third starport. 222/year is looking doable.
At this point, 160 Starship launches in 2026 would be close to every weekday.
They already have three launch sites for Falcon and can't do 200.
(Also see edit, my first post relied on Apple's autocomplete for maths and it used a short ton, plus point about these numbers corresponding to a mere 100 kg per satellite).
The 160 launches figure includes falcons. Seems like Starship fuels and flight tests faster than Falcon though. And if they manage to reuse second stages, then that eliminates a significant manufacturing bottleneck.
If you're counting Falcons, you are making my point for me: even with those, on three launch sites, they still can't get close to the minimum for an extremely small, to the point of being unreasonable, target satellite mass.
Further, until they actually do solve upper stage reuse, it is an "if" which can kill the economics of the vehicle itself, let alone reach the eventual potential cost reductions necessary for space based data centres to be worthwhile.
I don't see any reason a non-renewable Starship upper stage would kill the economics of the vehicle. No one else has a renewable upper stage yet, so there's no competition in that space until someone else does. Stoke have an interesting design but it hasn't flown yet and is only about the size of Falcon.
If they do manage to reuse the upper stage, then they should have no problem exceeding falcon launch cadence. Starship is much easier to build than Falcon. Welding is simpler and less expensive than the carbon composites used on Falcon upper stages.
The competition isn't other launch providers, it's not going to space at all.
According to Google, the price threshold to make space make more sensible than building on the ground is $200/kg: https://arxiv.org/pdf/2511.19468
Without full reusability, the estimated cost for Starship to LEO is kinda hard to find (necessarily, given the design isn't yet finalised), Wikipedia says $100m/launch in expendable mode, and the SpaceX website* says 250 metric tonnes in expendable mode, which is $100e6/250 metric tonnes = $400/kg.
You say that, but I have seen in the wild a scroll gesture to increase or decrease the value of a telephone number.
Wasn't even capped at zero, so I could scroll to a negative (phone) number.
reply