Musk can save the world

…or make us freer from oppressors

tldr: Musk already flies the hardware to give the whole world uncensored internet, but Direct to Cell physics caps it to a trickle, so the real move is open specs plus cheap ground gear plus a portable pay-key – and he will not do it, because nobody with that power cares about freedom.

Imagine a world where the internet is free from everyone without any kind of governmental interference at all. Where no single fucking state can oppress their own citizens anyhow by violating their right to informational exchange.

We could have that, if tomorrow Musk decided: fuck you all, I’m gonna carry the freedom into the world now.

(sorry for potential technical mistakes, this is more like an idea)

The Direct to Cell dream and the ugly math

Recently I found the news about how Direct to Cell now allows Starlink satellites to connect right to your phone (what it actually is: 650+ satellites with 25-square-meter phased arrays, talking to normal phones on T-Mobile 1900 MHz spectrum as distant cell towers, messaging first, data since October 2025). And I got immediately excited. If you extrapolate the launches and all that, that means that such tech could cover the whole world and connect to any phone beneath it.

Imagine the dream. You can connect to the internet anywhere, direct to a provider who does not filter, does not censor you, does not give a fuck about your local government and what they think, and just provides you free, whether you are in Russia, China, Europe or whatever.

If he had the willingness and balls to actually do it.

But the actual sad part. I was going around calculating how much and how feasible it is and the physics here are really brutal.

Direct to Cell even with the next generation satellites would be no way populous enough to actually spawn uplink in any amount in the next 10 years, nor would it really be feasible to be useful as something that could become half of your city traffic. And it has so many limitations indoors and outdoors, also kinda crap. No way you could use it reliably. In the best case you would get too much interference and slowness, and of course the adversary is gonna fight you too.

The problem is what? Direct to Cell technology is just unfeasible because the phone, the transmitter and the receiver, are absolutely forced. Your phone outputs milliwatts, the satellite sits 550 km above you, and no firmware update fixes that gap.

The pipe is spectrum and it is finite. Starlink Direct to Cell runs on roughly 5 MHz, maybe 65 MHz after EchoStar. One US carrier alone holds hundreds: T-Mobile about 375 MHz, AT&T 314 MHz, Verizon 279 MHz. Shannon caps it: Capacity = Bandwidth x log2(1 + SNR). You cannot manufacture Hertz. Adding satellites is like adding faucets to the same pipe. Total flow stays capped, and satellites over the same spot start jamming each other unless beams stay angularly separated.

Beam geometry kills cities. A tower covers 3 to 20 square miles. One satellite beam covers 100 to 600. In Paris that beam holds millions splitting maybe 20 Gbps into a trickle. Musk himself said it is only effective where there is no existing cellular service, 1 Mbps or less in a 12- to 20-mile-wide beam. Analysts put D2D capacity density 2,000x to 1,300,000x lower than mid-band 5G.

And hopping to free frequencies does not save it. The signal arrives too weak to go through a building or even a car windshield. Different frequency does not amplify it. LEO moves at 7 km per second, so Doppler hunting eats more than you gain, and uplink from a phone capped by battery and regulation is even worse. That requires years of development, years of satellite deployment improvement, which is something that the whole world will be fighting against if he decides to actually provide us with free internet.

Just to cover the rural US area you would need 40,000 of those satellites to get any kind of reliable and speedy reception. So it is way too out of reach.

Until we have dedicated antennas in phones, and yet it still would suffer so much from the problem of being unable to work indoors, it could be practically very limited. You would still want your router to sit outside and have better power.

We already have a network that works

When I thought, well, okay, we have so much problem with phone antennas, there is no way we can design anything around that. I thought Wi-Fi routers are also ubiquitous, but again, they also just physically do not have enough built-in power to be utilized. What you need is something where you can modulate power manually anyway.

And then I thought: why don’t you just use existing technology? It already works, worldwide, coverage is good. But the problem is that you need a custom router to use that. And this is crap, it is proprietary and everything is closed. But what if it wasn’t?

We already have the network, and it works fine. The first generation dish had motorized targeting, it physically chased satellites across the sky. And recently we got Starlink Mini, which does not have to be motorized, targeted, and is just a big chunk of metal, practically. It just sits in one place, and you just point it once and call it a day.

Look at the spec that actually matters: Orientation: Software Assisted Manual Orienting. It does not auto-track. You just point it in the right direction and the phased array handles the rest electronically within its 110-degree cone.

Because they proved the Mini antenna can work fixed like that, a makeshift antenna would work poorly, but doable. Your home-built antenna just needs to point roughly at the sky, and theoretically everyone can make their own to work with those satellites without having to have a dedicated router.

What the poor man’s kit actually is

Basically, everyone should be able to create an antenna, plus use a microcomputer as a software layer.

The full stack is three physical things:

[Antenna] -> [RF Frontend] -> [Compute Board] -> [Your device]

The antenna is the easy part. Ku-band at 12 GHz has about 2.5 cm wavelength, so your dish only needs to be 30-50 cm to get usable gain. That is a salad bowl situation, not a NASA dish. Wire on wood is garbage though, you need flat copper patches and they need to be dimensionally precise, because at 12 GHz, 1 mm of error actually matters. But copper-clad PCB board is sold in literally every electronics market on earth. You print the pattern, iron transfer it, acid etch it. Genuinely buildable anywhere by anyone with basic skills. Or a wire mesh dish on a PVC frame with a tin can feed horn. Under 30 dollars if you scavenge properly.

The hardest part is the actual modulator. The RF frontend. That is the hardest part to source, because those components are rare and not as accessible as you think.

Why can’t any computer just do it? Your Pi GPIO outputs maybe 200 MHz max in practice. Satellite uplink is 14,000 MHz. CPU only understands digital bits, RF is analog waves oscillating 12 billion times per second. Your CPU physically cannot sample that fast. Period. Physics. You need one analog translator. Nothing else.

The receive side is already solved globally. Satellite TV LNBs exist everywhere. Every country that has ever had satellite TV, including Iran, Russia, everywhere, has millions of these sitting on rooftops right now. Already Ku-band. Already 5-8 dollars. Already a commodity nobody can crack down on because banning them means banning satellite TV, which governments actually want to keep.

The transmit side is the real bastard. Upconverter for sending the signal up at 14 GHz, you cannot salvage that from a TV dish.

But the uplink does not have to be great. On the transmit side free internet counts even if it is slow. Still beating satellites with terrible uplink is still better than nothing, and you can crank the power here reliably, optimally 20 W. Uplink data for browsing is tiny. You are basically just sending URLs and requests. GET google.com is maybe 200 bytes. That is it.

So ideally: stupid DIY antenna plus one dedicated Ku-band component plus phone. Phone handles the crypto key, protocol stack, all the brain stuff. To be honest, here we move way up from just makeshift territory. You cannot salvage the transmit side from TV parts, you need the dedicated Ku part, and that means the expertise required goes up. But if the manual on how to do it all is published, motivated people would be able to craft something like that anyway. Phone with USB-C, module, wire to dish. That is the whole hardware stack.

The catch: that Ku-band module does not exist as a 5-dollar commodity yet. The reason Starlink dish costs what it costs is partly that chip. Unless someone manufactures specifically for this open ecosystem, your normie is still stuck. Does Musk also release a dead-simple reference chip design that Chinese factories clone within 3 months? Because that is how this actually becomes accessible. Not DIY RF from trash, but commodity cloned module plus foil antenna plus phone app.

And remember: the satellites are frozen in time. Whatever RF hardware is up there stays up there. A software update can change auth, protocol layer, who gets access, encryption keys. It cannot change frequencies, modulation, physical RF specs. So you are stuck working with what is already flying.

As the market develops for such components, governments would start cracking down on everyone trying to source them. So it is a temporary solution, unless the satellites get better adjusted, if possible, to use other ranges too.

One little problem remains: the satellite uplink has to be synchronized. Your bursts have to line up with everyone else’s, and that part is on you – you compute your own timing and your own Doppler from GPS and the broadcast ephemeris, because at 550 km there’s no base station close enough to correct you in time.

And your device is worse than the reference. A cheap dish is cruder, points sloppier, and needs more power to hold the same link. That alone isn’t fatal – Starlink already hands you a time slot, a frequency range, and tells you to turn down when you’re too loud. The real problem is what you send without meaning to: sidelobes spraying power into the next beam, emissions leaking outside your slot. That’s the part that gets you located, and it’s the exact thing the licensing regime exists to police.

But the density problem can’t be patched by a smarter algorithm. Adding terminals doesn’t add hertz. One beam over a city is a single 50-km cell – more devices inside it just split the same pool, with less signal for each. The ceiling is spectrum and beam count, and both are frozen in orbit. Which is why the community answer was never "everyone builds their own terminal". It’s one good gateway per area, pointed properly, with everyone else on a local mesh.

Just open it: one app, one key, anywhere

Next to the practical, here is how it should work. Musk just opens every specification for antenna, protocol and all that. So it is accessible as long as you just open the app, get the encryption key, pay, top up the balance with crypto or something.

If he opened the channel to be used by us, anyone can just connect from anywhere in the world using makeshift antennas and the only requirement is that you open the app, you pay for it and get your encryption key to use the service.

That key importantly should be usable anywhere. Think of a signed token, like crypto meets JWT. You pay anywhere, you receive a cryptographically signed key. The key contains bandwidth tier, expiration. Every packet you send includes your token signature. The satellite verifies it on the fly, drops your packets if invalid. No central ISP needed. The satellite is the verifier. Totally stateless, totally portable. Pay in Bangkok, use in Brazil.

That key can come from anywhere. You can pay in one country, you can pay with crypto or whatever. So citizens in oppressed countries can still utilize it by having it paid by someone else or donated or just sourced collectively for the neighborhood. Everyone can just use that without additional problems of actually paying for it.

So it becomes a collective open protocol which anyone can connect to and what you actually pay is for the bandwidth. Anyone can use the antenna they made themselves and build it in any free shop, really accessible without much knowledge.

How they will fight it, and why it still wins

The problems everyone will face when governments try to jump in. The protocol would need to be updated to be usable against interference attempts. It has to adapt, change frequencies on hop and potentially use side channels if possible, just to be more resistant.

Someone might complain that blah blah blah, it is gonna be military use and all that and your adversaries everywhere would be using that too. And yet, free for all means that you do not discriminate.

Yes, a couple of militaries are gonna use that, or state regimes are gonna also use that, but that is nothing compared to the actual empowerment of millions of citizens who are living under such regimes, who would finally be able to know the truth, to coordinate and eventually crush the same regimes in question.

And besides, we already have techniques for limiting. For example, stripping uplink on flying rockets, when it just cannot connect when your router is physically moving. And second, when the military is gonna use that, they would naturally cluster anyway. And you find out, you can clearly map out: haha, this is a military base on the map, from our intelligence we know what you fuckers are doing. Let us just shut you down in this particular area.

However, still, in cities you cannot just do that because people would start creating mesh networks. The coverage in cities would naturally be bad, so people would start moving receivers outside, up to the city outskirts, creating a receiver outside of the city with a better sky view, which transfers back to the city via mesh networks.

Why only Musk can do this

Why only Musk can do this? Consider that if he tries to do that, then he has to forego any kind of land business or any kind of presence in every other country. Because they will hate him for it. And they will hate and sanction and try to attack indirectly with any assets. So you need a company who just does not care or have to have physical presence anywhere else, except domestically. That condition we already have. Current terminals are not shipped much over physical stores and he already half confidently can fuck off every other country. However, the actual space business is important here. That means that the space business is leverage when you practically have a monopoly on cheap satellite launches. Therefore everyone has to rely on you. Therefore also, anyone who ever wants to put satellites in space has to accept the condition that he will sell you the space on a rocket, but also that means that citizens of your country also receive free internet regardless of what you think. That clears a lot of half-woke spaces like Europe or Asian countries, while the true adversaries like China or Russia already have their own space launching capacity and just would not even care.

Would he do this? Of course not

Would he do this? Of course not. Nobody even cares about freedom anymore. They only want more control for themselves and would never produce anything good and pure for our world now.

Yes, you can complain that it is still a commercial service and they have a limited amount of power, blah. But yet, if they really cared, we would already be crowdfunding to kickstart this project, to help build better rockets that can spawn more satellites and more and more to cover the whole world with better coverage in just a couple of years.

It could potentially be a nationally funded project from a nation that once cared about liberty. Not anymore.

If he tried anything like that, they would eat him alive domestically. Regardless of who is in power, no state wants people to start living in a new reality where nobody can snoop on your internet. Once people get used to that, you cannot put it back.

Yet one man can still one day just say fuck you all and just turn it on to be free for everyone.

Today, Falcon 9 can carry 30 satellites per launch and launches are limited. But space is being commoditized every day. And what today costs trillions, in 3 years from now, could be done with a budget of 10-20 billion. Not too big a price for global freedom. Maybe there is a glimmer of hope that in 2030 we would one day go to sleep and wake up with somebody quietly enabling the switch, dropping all the specs, all the apps and waking up in a better world.