As far as a race goes there's only one other runner currently and they are 1/10th as far along (6 vs 60 satelites) and without a plan to launch as cheaply as SpaceX. Spacex will be the first to have a system and unless it is too expensive or has other operational issues they are going to be hard to catch.
SpaceX has a huge price advantage on launches so a competitor would have to be able to significantly undercut them on the cost of the hardware just to become competetive.
Starlink is promising well under a million per satellite, and OneWeb is reportedly already at a million. Satellite costs might end up lower than launch costs, and even if they're higher that doesn't seem like they'll dominate.
That may be true for heavy ones that are only ever built once but these are light and mass produced. I cannot find the source but I'm pretty sure elon musk said that the rocket costs more to build than all of the satellites on a launch.
The expert quoted in the article believes that first mover advantage won't buy SpaceX much. She believes convenience, price and performance are going to be a lot more important in choosing winners.
Although you’re downvoted here, I believe you have a point. This expert doesn’t seem to factor in the likelihood that providing better quality service will be boosted by first mover advantage given the nature of the system. A linked constellation system is still novel and will certainly require trial and error to hone it. SpaceX has a major edge here in being first and having related expertise in a vertically integrated company. Unless they totally botch it they could readily remain a generation ahead of other players for a decade or more just by leveraging extra launch ability. E.g. they can iterate quicker and more often.
Given this constellation and phased array antennas are novel combination an expert’s opinion is worth about as much as most here on HN. Also, it’s annoying how the article questions finding a market, as the competition in semi-rural areas is Comcast, charter, Verizon etc if lucky. That’s a pretty low bar IMHO.
Elon recently stated the phased array antennas for Starlink will be the most advanced phased array antennas in the world. That's saying something as the AESA radars in the newer jets use phased array radars + antennas.
Once those start being manufactured at scale, maybe we'll have AESA radars available for hobbyist use? Say, aren't Tesla cars supposed to be using radar?
No mention of the legal problems. SpaceX is pitching to become a worldwide last-mile intenet provider. Each country, the US included, will have issue with that. They will each set regulatory burdens, censorship limitations, that will make SpaceX's life difficult.
I have yet to see any real plans for downlink locations, the places where the satellite network will integrate with the fiber backbone that is 'the internet'. The system cannot be funneled through a handful of locations in the US. For the mega constellation concept to work you need hundreds of local downlinks so that connections don't have to hop across too many sats before exiting the network.
Yeah, this doesn't seem like a huge problem. It's no different than trying to figure out taxes when Amazon is spread across the globe; the fact that the entity "Amazon" is spread out doesn't mean that the seller, distributor, and purchaser of any given product has concrete legal locations. With the downlinks needing to be on the ground for performance and hookup reasons, SpaceX would have to go out of their way to break local laws; they're going to automatically pick up the ambient network environment unless they do something special and very deliberate.
Ya, but look into the regs for even a developed country like Canada. When it comes to downlinking satellite data they are all stuck in a pre-megaconstellation regulatory regime. There are often laws stating that the host country have veto power over all data going up/down to the sat, that the sat not handle data that the host country cannot control. They do not envision the concept of multiple satellites communicating between themselves.
Given that regulations were once written, I got to think that regulations can change. If there's a technology that really works, and people really want/need it, and it solves the expensive problem of connecting rural communities, I don't see regulators just being like ¯\_(ツ)_/¯ 'Yeah, that'd be cool, but law's the law! We already printed these books!'
And if other countries have it and it prevents yours from 'staying competitive', important people pay attention.
And if that civic approach doesn't work, some rich farmer tells his senator buddy on a hunting trip that he wants Starlink to watch Netflix in the woods, and the senator tells the regulator to make it happen.
I think a lot of discussion I have seen online has overlooked the importance of the phased array antennas, which are too large to make Starlink practical for someone using a phone or laptop on the go. However, it would be very convenient for people needing Internet service at home.
I have also heard people (not in this thread) talk about Starlink as a way to bypass censorship in places like China, but this is almost certainly not possible. What authoritarian regime would allow the sale of the Starlink antennas in their country and the local downlinks needed to have good latency?
This is a fascinating project, but I hope we don't end up treating it as a panacea to all issues of information freedom.
There are no technical means needed to bypass internet censorship in China. People are using the free internet there perfectly, using obfs4. Problem is that there is no way to express their opinion freely on the internet because:
- Expression of opinion almost always lets you being identified even if your IP address can't be seen.
- Most people won't support your opinion - because they are paid trolls (a large fraction of app posts, because a comparatively small - 'only' hundreds of thousands - trolls post in insane quantities) and because many online spaces are shaped by trolls and by people who are too afraid of expressing their opinion freely. Internet is simply not seen as a place appropriate for sensitive discussions in China and this will not change if people will be able to access all blocked resources, or reliably hide their locations.
> There are no technical means needed to bypass internet censorship in China.
Individuals bypassing technical restrictions is very different from having a commercial operation. Without the mandated censorship and monitoring, Starlink's entire business is completely illegal in China, so the entire thing (from setup to service to support) would have to be run underground, they'd have to accept payment in Bitcoin, no SpaceX executive could ever travel to China, and Tesla would probably have to get out too.
It's not gonna happen. There is no chance, none whatsoever, that Starlink ever does business in China unless it goes through the same filters as everyone else, even though from a purely technical perspective it theoretically could.
> Most people won't support your opinion - because they are paid trolls (a large fraction of app posts, because a comparatively small - 'only' hundreds of thousands - trolls post in insane quantities) and because many online spaces are shaped by trolls and by people who are too afraid of expressing their opinion freely.
I wonder if that situation is exclusively true in China, or if similar situations exist where it is covert, rather than overt.
I think people need to realize spacex has a primary goal of developng rockets.
I see all this as analogous to the problem of beginning pilots: the cost of getting hours in a plane is expensive.
Offering <x> where x requires the pilot to fly a plane will get the pilot discount hours in the air. x could be flight lessons, flying packages around, aerial photo services, etc.
Indeed. If Starlink merely breaks even, it'll massively increase the demand for launch services which allows them to justify the development of large, fully reusable rockets like Starship. From SpaceX's perspective, besides the straight up revenue they provide, the main point of megaconstellations like Starlink is fodder for reusable rockets.
...and having multiple megaconstellations competing thus isn't necessarily a bad thing even from SpaceX's perspective. More mass to launch to orbit.
...that's one thing about reusable rockets that is key to understand. There's no fundamental physics that says reusable rockets can't be done. Shuttle proved upper stages can be reused and SpaceX showed first stages can be reused. The main question is demand: reusable rockets make sense IFF there's more than sufficient demand to pay back the R&D and the keep the factory busy.
To over-simplify things a bit:
Rocket factories for even expendable rockets seem to struggle to be affordable unless you're building at least 10 airframes per year, so if you want to get 10-100 reuses out of your rocket, you'd best be launching at least 100-1000 times per year...
These articles ignore a vast number of global web users who are living in or around major cities and towns just to have connectivity, who will likely be early adopters seeking the freedom to live in more affordable rural locations. The potential size of this group is hard to judge, so it seems there's some reluctance to even speculate on a figure.
For reference GEO is something like 250ms and in LEO you can get sub 50ms.
I did the math a year ago and so numbers may be a little off due to memory, but they should be in the ball park. Also these would be optimal numbers, so realistic would be worse.
The existing Internet satelite are incredible expensive and not only the latency is really bad, but the bandwidth is also very limited, you can not consider it as a broadband ISP. Starlink will be a true alternative to the fiber optic ISPs, and for world-wide will be even faster thanks to the direct lasser link among the constellation.
The current state of satellite is the dial-up of high speed internet. The latency is horrible. You can't video conference or play online games with it. SpaceX's satellites are a lot lower, so they have better latency.
I could see it even driving migration out of cities. One of the big reasons for not already doing so is the terrible connections for remote workers further out.
Worth noting here that SpaceX is currently raising another round: pi hundred million dollars at a valuation of $35 billion. And it is possible to participate as an individual accredited investor on https://fundnel.com
There's a lot of rural places that barely have 3G. The problem is if only 20 people benefit from a tower being stood up, they won't do it. This is the advantage of satellite internet, connecting the sparsely populated. Towers are the future of cities and outlying areas, low-orbit sat-internet is the future of rural areas.
Perhaps a viable business model could be hybrid solution where mobile carriers partner to create LTE and 5G towers with Starlink uplinks. This would make it easier for carriers to deploy towers in remote locations. It’s my understanding that the primary challenge for deploying towers in underserved areas is lack of existing infrastructure for uplink. Call me crazy, but if you could solar power the ground station you could plop them anywhere on earth.
Put the transceivers on the roofs of cars. Then make the cars rolling supercomputers -- oh, already done! Then start selling a part of the compute power of the cars as a cloud with no search warrantable address. All of the virtual servers will be ephemeral and will migrate such that they're only ever running on moving cars.
This is actually believed to be the first market SpaceX is going to go after. It lets them focus more on the tech and make sure it works before trying to scale things like the production of the phased array antennas to the point an average consumer can afford it or even things like user signups and subscriptions and all the legal "fun" that will come with trying to be a last-mile providers globally.
Some of their largest markets are going to be in countries where people cannot afford to have their own phased array antennas anyways. So yeah, especially early this will probably be their market.
If the constellation costs $10 billion to launch as estimated and has a operating cost of $1 billion/year and a lifetime of 5 years, Spacex will need only $3 billion/year to break even. If they sell subscriptions at $50/month they will only need 5 million subscribers to break even. They only have to be better than current offerings in poorly served markets to get that. I think it will be a cash cow.
I live in Venezuela and our internet service is crumbling and censored (Dont fucking vote for communists or populist socialists)
I work online so a stable internet connection is a must. Right now we're surviving with boutique wireless ISPs (not cell, think point to point networks used in rural areas but deployed in the city because you have post-apocalyptic levels of collapse) They buy their uplink from the monopoly government ISP but pay handsomely for priority.
Communism and socialism do not have anything to do with censorship.
Edit: I was responding to parent's first sentence, "I live in Venezuela and our internet service is crumbling and censored (Dont fucking vote for communists or populist socialists)". That is like living in Russia, China, Saudi Arabia, etc. (look here if you want more examples: https://commons.wikimedia.org/wiki/File:Internet_Censorship_... ), and then blaming Capitalism for the censorhip of the Internet.
From what I have read it sounded like the US government has been working to undermine Venezuela economically for some time. Socialism was working too well, it brought huge swathes of the population out of poverty. A working less-capitalist model being dangerous to the global status quo, had to go.
Im really impressed at the number of downvotes, despite the lack of subsequent replies. Here's another source:
"The first UN rapporteur to visit Venezuela for 21 years has told The Independent the US sanctions on the country are illegal and could amount to “crimes against humanity” under international law.
Former special rapporteur Alfred de Zayas, who finished his term at the UN in March, has criticized the US for engaging in “economic warfare” against Venezuela which he said is hurting the economy and killing Venezuelans."
You're probably being downvoted without replies because your post of such low quality, it is not worth replying to.
"Socialism was working too well, it brought huge swathes of the population out of poverty. A working less-capitalist model being dangerous to the global status quo, had to go."
This is ideological drivel so far removed from reality, it's basically trolling.
This comment breaks the site guidelines. You have a long history of doing that. We've warned you more than once. Would you please fix this? I don't want to ban you, but if you don't stop doing this, we're going to be forced to.
History remembers the winners - so even though I actually agree with what you're saying about US government economic warfare, the end result was the Venezuelan government turned to shit. It's become oppressive, their policies have failed and now their society is crumbling and people are suffering.
If Venezuela's form of socialism was working so well, then it should have sustained itself no matter what the US did. Instead it was supported by oil money, and as soon as the oil dollars were cut off it collapsed because there was nothing revolutionary or innovative about it. I'll be voting for Andrew Yang this primary because I like the idea of UBI and a social safety net and how he's planning to implement his policies. I think that form of 21st century socialism might work - the form that Venezuela chose just led to cronyism and oppression.
Did you see the part about competition? It can end up like airlines - never ending price war. Surely good for consumers, but a measly, possibly negative, total return on investment for the whole sector.
SpaceX has a gargantuan advantage when it comes to the cost to get stuff to space, though. And it doesn't seem like anybody will catch up with them any time soon.
And then you'll vote (purchase) based on which government you trust the most for whom the company will be under the umbrella of - US, China, India, etc.
Or it could end up like AT&T making $20 billion dollars a year in operating income. The high initial capital costs to launch enough satellites to provide decent coverage should prevent there from being too much competition.
Going from 60ms to 40ms for a transatlantic hop isn't quite that significant IMO. An HFT organization will simply co-locate their compute servers in the same continent as the exchange. Actually, they will colocate their servers in the same geographic neighborhood, if not within the very same data center. Source: was HFT systems dev for 10 years.
Wellll, that's not actually how it works. Capital expenditures are only a fraction of total expenses, which is only a fraction of annual cash flow. For example, in a previous life I worked at a VoIP company. At that time, our technical infrastructure was largely paid for. We had a couple of million subscribers and were struggling financially. The lion's share of cash flow went to Customer Service, Advertising, and Debt financing and repayment. (Disclaimer--I haven't kept up with the company and have no idea how they're doing now.)
It's very difficult for an ISP to make much money at $30/mo per customer, unless you go the Google route and only provide automated customer service. That's why established ISP such as Verizon are losing interest and selling their operations to bottom-feeding companies like CenturyLink. All of the money is in Mobile.
I can't see SpaceX itself actually providing the retail experience. It's entirely foreign to how the company functions. I'd see SpaceX either contracting to provide technical services (like Level3), or wholesaling the product to sort-of MVNOs (Mobile Virtual Network Operators). If wholesale, it could be a captive sister company (Musk does have the retail experience) (by analogy, like Mobile PCS which is owned by T-Mobile), or to independent operators (analogy being Virgin Mobile).
I don't have any deep insight into the market, but some interesting possibilities would be a hungry cell carrier like T-Mobile branching out into ISP service, hoping to provide more attractive land/mobile bundles.
Here's a radical idea. Do you think SpaceX launching 1,000+ satellites for internet is part of a larger experiment for creating an earth shield as a last ditch effort to fend off climate change? How much surface area would these satellites need to cover to have a discernible impact on the amount of sunlight reaching earth?
It appears the potential market in the US is based on Federal broadband coverage figures. From the story:
"In 2017, more than 26% of people who lived in rural
areas in the U.S. were not covered by terrestrial
broadband internet service, according to a May report
from the Federal Communications Commission."
FCC numbers are pure fiction. It is well known that providers are GREATLY exaggerating their coverage and available capacity to rural and semi rural areas. Much of the coverage that actually exists is also garbage; fixed wireless suffering absurd contention and unreliable, low performance cable system or DSL systems.
There are well-heeled communities adjacent to large urban areas that can get nothing worthwhile; they're inflicting substantial new property taxes on themselves to fund FTTH build-outs. I suspect the actual market for rural broadband is many times larger than whatever is assumed in this story. I also know for fact that there is a cohort of people that will go further out into the hinterlands the instant they can solve the bandwidth problem at a reasonable cost. I'm one of them.
> FCC numbers are pure fiction. It is well known that providers are GREATLY exaggerating their coverage and available capacity to rural and semi rural areas.
Thank you.
At risk of piling on a "+1" comment, I will add a little bit of context.
My parents live about 35 miles from a "midwest urban core", they are served by Frontier. Supposedly they can get "1.5Mbps DSL" which qualifies as broadband. On a good day they can get 64k. This has degraded from about 768k ~10 years ago, to ~64k today. Absolute no maintenance other than "electrons kinda flowish" has been done on these lines.
But, my parents and their neighbors of ~150 households totally count for "broadband" stats.
The fraud going on here is absolutely mind boggingly in scope. The "lie big if you're going to lie" thing comes into mind here. Billions evaporated due to nothing other than absolutely in-your-face fraud.
That's slower than ISDN :( To stream Netflix at 10 frames per second, they'd have to watch a show in something like 28 px by 21 px. That'd be like playing Minecraft on a 1970s LED calculator.
Indeed. They have since canceled as of about 4-5 months ago. Now they use a T-Mobile re-seller (via walmart, iirc) that gives them 25GB/mo for less than their "broadband DSL" cost. So far they have been happy - watching shows at 480p.
But you won’t be able to see more than a few hundred satellites at the same time because they’re so low. Whereas you can see at least 1000 stars in the sky from any location.
The latency will be top-notch, but only if they strictly limit total traffic. So, streaming will probably cost way too much for normal people, unless they can contrive to cache copies of the absolutely most popular stuff on all the birds. Then, everything big but not cached will be super expensive, throttled, or actually forbidden.
Of course they will charge the publishers for caching, by spot auction, and the publishers will use a good chunk of their allotment to cache ads.
Honestly this could be true, however I dont think musk would risk most of his wealth on non-viable tech. The dude does a lot of really dumb stuff with his money, but in terms of investments and new products, his largest tech investments have a remarkable track record.
The lowest-latency long-distance communication in the world, by several milliseconds, will be worth far more, to just the right people, than any number of streamed sitcoms and whodunits. I do not doubt Musk will make a fortune on it. I just don't think it will be of much immediate use to you, me, or Disney.
But first he has to launch it, and, before that, raise money to launch it. Investors' tiny minds need to believe in massive disruption, just long enough. They will get paid, just not the way they might guess.
98 comments
[ 4.5 ms ] story [ 171 ms ] threadStarlink is promising well under a million per satellite, and OneWeb is reportedly already at a million. Satellite costs might end up lower than launch costs, and even if they're higher that doesn't seem like they'll dominate.
Given this constellation and phased array antennas are novel combination an expert’s opinion is worth about as much as most here on HN. Also, it’s annoying how the article questions finding a market, as the competition in semi-rural areas is Comcast, charter, Verizon etc if lucky. That’s a pretty low bar IMHO.
I hope we get some good competition in this area!
I have yet to see any real plans for downlink locations, the places where the satellite network will integrate with the fiber backbone that is 'the internet'. The system cannot be funneled through a handful of locations in the US. For the mega constellation concept to work you need hundreds of local downlinks so that connections don't have to hop across too many sats before exiting the network.
And if other countries have it and it prevents yours from 'staying competitive', important people pay attention.
And if that civic approach doesn't work, some rich farmer tells his senator buddy on a hunting trip that he wants Starlink to watch Netflix in the woods, and the senator tells the regulator to make it happen.
I have also heard people (not in this thread) talk about Starlink as a way to bypass censorship in places like China, but this is almost certainly not possible. What authoritarian regime would allow the sale of the Starlink antennas in their country and the local downlinks needed to have good latency?
This is a fascinating project, but I hope we don't end up treating it as a panacea to all issues of information freedom.
- Expression of opinion almost always lets you being identified even if your IP address can't be seen.
- Most people won't support your opinion - because they are paid trolls (a large fraction of app posts, because a comparatively small - 'only' hundreds of thousands - trolls post in insane quantities) and because many online spaces are shaped by trolls and by people who are too afraid of expressing their opinion freely. Internet is simply not seen as a place appropriate for sensitive discussions in China and this will not change if people will be able to access all blocked resources, or reliably hide their locations.
Individuals bypassing technical restrictions is very different from having a commercial operation. Without the mandated censorship and monitoring, Starlink's entire business is completely illegal in China, so the entire thing (from setup to service to support) would have to be run underground, they'd have to accept payment in Bitcoin, no SpaceX executive could ever travel to China, and Tesla would probably have to get out too.
It's not gonna happen. There is no chance, none whatsoever, that Starlink ever does business in China unless it goes through the same filters as everyone else, even though from a purely technical perspective it theoretically could.
This is not entirely true. It's been getting tougher and tougher to get past TGF, even with obfuscation.
I wonder if that situation is exclusively true in China, or if similar situations exist where it is covert, rather than overt.
I see all this as analogous to the problem of beginning pilots: the cost of getting hours in a plane is expensive.
Offering <x> where x requires the pilot to fly a plane will get the pilot discount hours in the air. x could be flight lessons, flying packages around, aerial photo services, etc.
...and having multiple megaconstellations competing thus isn't necessarily a bad thing even from SpaceX's perspective. More mass to launch to orbit.
...that's one thing about reusable rockets that is key to understand. There's no fundamental physics that says reusable rockets can't be done. Shuttle proved upper stages can be reused and SpaceX showed first stages can be reused. The main question is demand: reusable rockets make sense IFF there's more than sufficient demand to pay back the R&D and the keep the factory busy.
To over-simplify things a bit: Rocket factories for even expendable rockets seem to struggle to be affordable unless you're building at least 10 airframes per year, so if you want to get 10-100 reuses out of your rocket, you'd best be launching at least 100-1000 times per year...
(where eggs = rockets and chickens = starlink)
:)
I did the math a year ago and so numbers may be a little off due to memory, but they should be in the ball park. Also these would be optimal numbers, so realistic would be worse.
Any disruption in the current sat-internet market is welcome.
Some of their largest markets are going to be in countries where people cannot afford to have their own phased array antennas anyways. So yeah, especially early this will probably be their market.
I work online so a stable internet connection is a must. Right now we're surviving with boutique wireless ISPs (not cell, think point to point networks used in rural areas but deployed in the city because you have post-apocalyptic levels of collapse) They buy their uplink from the monopoly government ISP but pay handsomely for priority.
I pay $30 a month for a 2mb connection.
I would GLADLY pay for Starlink.
Edit: I was responding to parent's first sentence, "I live in Venezuela and our internet service is crumbling and censored (Dont fucking vote for communists or populist socialists)". That is like living in Russia, China, Saudi Arabia, etc. (look here if you want more examples: https://commons.wikimedia.org/wiki/File:Internet_Censorship_... ), and then blaming Capitalism for the censorhip of the Internet.
https://news.ycombinator.com/newsguidelines.html
I'm sorry and I hope it gets better. Here is one article, as a source, its long though. Source: https://nacla.org/news/2018/05/18/united-states%E2%80%99-han...
"The first UN rapporteur to visit Venezuela for 21 years has told The Independent the US sanctions on the country are illegal and could amount to “crimes against humanity” under international law.
Former special rapporteur Alfred de Zayas, who finished his term at the UN in March, has criticized the US for engaging in “economic warfare” against Venezuela which he said is hurting the economy and killing Venezuelans."
Per the Indpendent: https://www.independent.co.uk/news/world/americas/venezuela-...
"Socialism was working too well, it brought huge swathes of the population out of poverty. A working less-capitalist model being dangerous to the global status quo, had to go."
This is ideological drivel so far removed from reality, it's basically trolling.
Source: https://www.pbs.org/wgbh/americanexperience/features/reagan-...
If you'd please review https://news.ycombinator.com/newsguidelines.html and stick to the rules from now on, that would be good.
If Venezuela's form of socialism was working so well, then it should have sustained itself no matter what the US did. Instead it was supported by oil money, and as soon as the oil dollars were cut off it collapsed because there was nothing revolutionary or innovative about it. I'll be voting for Andrew Yang this primary because I like the idea of UBI and a social safety net and how he's planning to implement his policies. I think that form of 21st century socialism might work - the form that Venezuela chose just led to cronyism and oppression.
Good domestic policy makes you impervious to interference by foreign powers with 40x your GDP?
I don’t get it, what’s the logic there?
It's very difficult for an ISP to make much money at $30/mo per customer, unless you go the Google route and only provide automated customer service. That's why established ISP such as Verizon are losing interest and selling their operations to bottom-feeding companies like CenturyLink. All of the money is in Mobile.
I can't see SpaceX itself actually providing the retail experience. It's entirely foreign to how the company functions. I'd see SpaceX either contracting to provide technical services (like Level3), or wholesaling the product to sort-of MVNOs (Mobile Virtual Network Operators). If wholesale, it could be a captive sister company (Musk does have the retail experience) (by analogy, like Mobile PCS which is owned by T-Mobile), or to independent operators (analogy being Virgin Mobile).
I don't have any deep insight into the market, but some interesting possibilities would be a hungry cell carrier like T-Mobile branching out into ISP service, hoping to provide more attractive land/mobile bundles.
There are well-heeled communities adjacent to large urban areas that can get nothing worthwhile; they're inflicting substantial new property taxes on themselves to fund FTTH build-outs. I suspect the actual market for rural broadband is many times larger than whatever is assumed in this story. I also know for fact that there is a cohort of people that will go further out into the hinterlands the instant they can solve the bandwidth problem at a reasonable cost. I'm one of them.
Thank you.
At risk of piling on a "+1" comment, I will add a little bit of context.
My parents live about 35 miles from a "midwest urban core", they are served by Frontier. Supposedly they can get "1.5Mbps DSL" which qualifies as broadband. On a good day they can get 64k. This has degraded from about 768k ~10 years ago, to ~64k today. Absolute no maintenance other than "electrons kinda flowish" has been done on these lines.
But, my parents and their neighbors of ~150 households totally count for "broadband" stats.
The fraud going on here is absolutely mind boggingly in scope. The "lie big if you're going to lie" thing comes into mind here. Billions evaporated due to nothing other than absolutely in-your-face fraud.
https://www.theguardian.com/technology/2019/may/28/spacex-sa...
The latency will be top-notch, but only if they strictly limit total traffic. So, streaming will probably cost way too much for normal people, unless they can contrive to cache copies of the absolutely most popular stuff on all the birds. Then, everything big but not cached will be super expensive, throttled, or actually forbidden.
Of course they will charge the publishers for caching, by spot auction, and the publishers will use a good chunk of their allotment to cache ads.
But first he has to launch it, and, before that, raise money to launch it. Investors' tiny minds need to believe in massive disruption, just long enough. They will get paid, just not the way they might guess.