SpaceX Is Building The Thing That Could End It

[00:00:00]We live in a world where one rogue satellite clipping another at 17,000 m an hour is a very real possibility. And the reality is that this scenario would be absolutely catastrophic because it would lead to an unstoppable cascade that would destroy everything in orbit.

[00:00:16]Destroy almost all of our communication infrastructure, GPS, defense, and it would make it essentially impossible for us to leave Earth. It's so catastrophic, in fact, that even SpaceX called this out as a real warning when it filed to become a public company. By the end of this video, you'll understand the likelihood of something like this happening. And I hate to break it to you, the likelihood is rapidly accelerating as time passes. And all it takes is one satellite stops answering commands. But guess what? Satellites going dead and unresponsive is a known documented problem in this industry.

[00:00:51]Now, why do I care so much about one dead satellite out of tens of thousands?

[00:00:56]It's speed. When two objects cross paths up in space, the closing speed runs 7 to 8 km/s, which is equivalent to New York to Miami, in about 4 minutes. At that speed, a piece of debris that's 1 centimeter across, which is the size of your thumbnail, hits with roughly the energy of an exploding hand grenade.

[00:01:14]Isn't that insane? The first collision throws off hundreds of new fragments.

[00:01:19]Every one of them now it's its own freaking grenade crossing the orbit of the tens of thousands of satellites up there that are currently packed into the same shell that's about 550 km up. So satellites start dodging, but the dodging burns fuel. And the fragments too small to track don't show up on anybody's screen until something else breaks apart. Tens of fragments become hundreds. Hundreds become thousands. and the shell carrying most of the world's satellite internet starts eating itself.

[00:01:49]And SpaceX knows this is a real risk.

[00:01:51]When they went public in June of 2026, which is the largest IPO in history, they had to file something called an S1 with the SEC. That's the legal document where a company selling stock to the public, has to list every risk that could seriously hurt the business.

[00:02:05]Written by lawyers whose entire job is making sure no investor can ever say, "See, you never told us." And deep in that document, SpaceX's own filing warns that a collision of debris generating event quote could result in a cascading series of collisions that could render certain orbital altitudes unusable and that this would quote materially and adversely affect the business. So SpaceX's own lawyers already settled the question of whether this is possible.

[00:02:32]But the question I actually care about is the odds. How likely is this really?

[00:02:38]Because what's writing on that number is massive. For example, Starlink is the internet connection for somewhere north of 10 million customers today in ranches, islands, ships, war zones, places that never had a real connection before. SpaceX and Blue Origin, Jeff Bezos, the space company, are both racing to put actual data centers in orbit. And that entire buildout assumes orbits stay usable. And since June of 2026, SpaceX is a roughly $2 trillion public company with plans to launch up to 1 million satellites in orbit. Now, back in 2019, there were roughly 2,000 working satellites around the entire planet. All the countries and companies combined. Now, 7 years later, SpaceX alone flies more than five times that number. Nobody has ever built anything in space that fast. The pace that SpaceX has built rockets and launched satellites is unlike anything that has ever happened before. And the only thing that's remotely close to it is the insane boom in AI we are all experiencing today. Not just in the digital world with chat bots and agents and chatt and claude and Gemini, but also in the home in the form of physical AI powered robots. For example, I have a Madic robot vacuum in our house that we use every single day. And it is the most reliable robot vacuum we've ever owned.

[00:03:55]And I mean that. It runs a five camera vision system powered by AI, which is the same pure vision approach Tesla uses with FSD, full self-driving. And just like a Tesla, it gets software updates that make it smarter over time. The app is easily one of the most intuitive apps I've ever used. And our vacuum, Fernando, that's what we've named him.

[00:04:13]He's great with the kids, with the pets, and my wife, who has some of the highest cleaning standards I've ever seen in a human being, does not hate this thing.

[00:04:21]If you knew my wife, that would be a 17,000 star review on Amazon. If you want to check it out, use my referral link in the description or the comment section below. And thank you so much to Madic for sponsoring this video. Now, getting back to space, there are 16,000 active satellites around this planet right now with every country and every company combined. About 10,000 of them are Starlink. So, one company owns roughly 66% of every working machine flying over your head. That's 2thirds of everything in space. But the filings for what comes next is the most important part. SpaceX has proposals on the table for space-based computing constellations that scale up to 1 million satellites. A million. What does that even mean? It means data centers in orbit. Instead of building giant AI warehouses on the ground, the kind that eat entire power grids that everyone's freaking pissed off about, you put the computers in space where the sun hits your solar panels around the clock, and you never fight anybody for electricity. For scale, a million satellites is roughly 60 times everything currently flying right now, proposed by just one company.

[00:05:29]And SpaceX is not alone in this. Blue Origin is chasing that same orbital data center prize with something called Project Sunrise with 52,000 satellites running data centers in space. Then you've got Rocket Labs racing into the same exact economy. This is a gold rush.

[00:05:45]Now, where does all this hardware actually go? Basically the same place.

[00:05:50]Orbits in space work like floors of a building. A satellite parked at a certain altitude stays on that floor, circling the planet over and over. And everyone is fighting over the same low floors. The shells between roughly 500 and 1,000 km up. Why? Because low means your internet signal makes a shorter round trip, which means faster speeds.

[00:06:12]And low is cheaper to reach with a rocket. Which means that the most valuable hardware humanity has ever launched is piling into a slice of sky a few hundred km thick. Everyone wants the same floor, the exact same floors. The reality is the sky we have in 2026 is a rounding error next to the sky these companies have already asked permission to build. That number is nuts. But somebody already did the math on this back in 1978. Two NASA scientists, Don Kesler and Burton Corp, by the way Kesler is literally the guy the whole field is named after. He ran NASA's orbital debris program, published a paper that year. And the mechanism in that paper is brutally simple. Objects in orbit sometimes hit each other. When they hit, they shatter into hundreds of thousands of fragments. And every single fragment is now a new object in orbit, moving just as fast as the thing it came from. More objects mean more collisions.

[00:07:13]More collisions means more fragments.

[00:07:16]And the only cleanup crew in the entire system is the atmosphere. There's actually a faint wisp of air even hundreds of kilometers up and over years it drags that stuff down until it burns up on re-entry. So you are in a race. Is the sky making debris faster than the atmosphere can flush it or slower?

[00:07:36]Kesler's warning was that there is a density threshold below it. Orbit cleans itself and mistakes get very easily forgiven. Above it, the collisions start feeding themselves. You could stop launching rockets forever and the debris count would still grow on its own for decades, maybe centuries. That's the Kesler effect or Kesler syndrome. It's an actual 1978 NASA paper 35 years before the movie Gravity turned it into a plot. And I'm telling you, the paper is scarier than the freaking movie.

[00:08:07]Okay, so how close are we really to that threshold? We can track about 40,000 objects right now. Those are roughly softballs sized and up. The models say the real count above 10 centimeters is closer to 54,000. Between 1 and 10 cm is about 1.2 million and down in the millimeter class, which are paint flex slivers of old solar panels, the estimate is somewhere around 140 million. At 8 km per second, every one of those is a freaking bullet. And everything under about 10 cm is effectively invisible. We can't track it, which means nobody can dodge it.

[00:08:45]Over a million bullets, nobody can see what the Now, real quick, about 70% of you that watch this channel often aren't subscribed. And if you're enjoying this video, clicking the subscribe button is the easiest way to support the channel.

[00:08:58]It's completely free and it goes a really long way. Thank you so much. Now, the worst number that I found. There's a metric that space tracking researchers publish called the crash clock and it asks one question. If every satellite up there stopped dodging with no maneuvers, everybody's just flying in orbit, how long until the first collision between two tracked objects? In 2018, the answer was 164 days, which is roughly 5 1/2 months of margin. By June of 25, it was down to 5 days. By May of 26, 2 1/2 days. And at that level, the math says there's roughly a 30% chance the first collision happens inside 24 hours. one in three odds that if the dodging stops, something gets hit within a freaking day. So why hasn't it happened? Because the dodging never stops. For example, SpaceX's Starling satellites performed about 300,000 collision avoidance maneuvers in 2025 alone. That's up roughly 50% in a single year. Somewhere over your head right now, a satellite is firing its thruster so it doesn't hit something else. And on average, that's happening every couple of minutes around the clock, fully automated without anyone even realizing.

[00:10:13]You think of space and you think this vast space, right? Nobody needs to move around. They move around a lot. That margin of safety is already gone at these altitudes. And what's holding the whole thing together is software. Tens of thousands of machines moving out of each other's way with no pause button.

[00:10:32]All of that hardware packed into those same crowded floors stays safe only as long as every machine keeps making its moves on time every time. So the obvious question is what could actually stop the dodging? And I hate the answer. Start with the accidents. One day a satellite can be a healthy machine, super happy, taking and executing commands like any other happy satellite. Then it goes quiet. Then it breaks apart. tens of tracked pieces where one controllable satellite used to be drifting through the busiest floors in orbit and debris doesn't have software, right? The whole safety system assumes both objects in close pass can move. The moment one of them can't, it's over. That's one satellite, but what takes out hundreds at once? The sun. In February of 2022, SpaceX launched 49 Starling satellites.

[00:11:25]A geomagnetic storm heated the upper atmosphere right as they deployed, dragged, spiked, and roughly 38 of them got dragged back down and burned up within days. That's one solar storm.

[00:11:38]Most of the launch completely gone. Then in May 2024, the strongest solar storm in about two decades hit, and something much sneakier happened. The atmosphere puffed up so much that the orbit predictions everyone uses for collision screening degraded for days, right?

[00:11:54]While thousands of satellites were maneuvering at once to compensate, satellites dodging each other depends on ground systems predicting where everything is going to be during a storm like that. They can't. A satellite cannot dodge what nobody can see coming.

[00:12:09]And the sun doesn't give a crap if we've parked billions of dollars of hardware in its freaking blast radius. And then there's the very boring reality.

[00:12:18]Satellites just die. Some percentage of any fleet fails either because of battery, radio, flight computer, whatever. And a small percentage of tens of thousands of machines is a lot of stuff that no one can control. But what if it's no accident at all? What if you can hack a satellite? The way I see it, a satellite is a computer with a thruster just strapped to it. Take control of one and you own a machine that can burn its engine wherever it wants inside the densest shells humans have ever built. Imagine an uncontrolled burn through a freaking packed floor.

[00:12:50]That's about as powerful as a nuclear bomb. And then there are actual missiles. In 2007, China fired one at its own dead weather satellite, Fengune 1C, at roughly 860 km up. That single test created over 3,000 trackable fragments. And almost two decades later, most of them are still circling. Because at that altitude, the air is too thin to drag anything down for decades, sometimes centuries. That's one test.

[00:13:18]Russia ran its own version in November of 2021. They fired a missile into a dead Soviet satellite called Cosmos 1408. Roughly 1,500 trackable fragments, and the cloud was scary enough that the astronauts on the International Space Station climbed into their return capsules in case they had to evacuate.

[00:13:37]The United States looked at all of this and in 2022 declared a ban on these kinetic anti-satellite tests. It was meant as a unilateral ban, but guess what? It binds exactly one country, the one that announced it, the US. China never signed anything. Neither did Russia. So all of these things can, and I hate to break it to you, will happen at some point. And if you want that kind of problem gone, someone has to physically go up and fix it. And who is supposed to do that? How can you even do that when potentially millions of bullets are circling the Earth? But there is a plan. There's actually three of them. Plan one is rules. Back in 2022, the FCC, the agency that licenses these constellations in the US, tightened its disposal rule. A dead satellite now has to come down within 5 years of ending its mission. The old guideline was 25 years. Europe is pushing even harder. ESA wants to be net zero on space debris by 2030 with missions that leave nothing behind.

[00:14:35]Unfortunately, that does nothing for the junk that's already up there, but we'll get there. Plan two is software.

[00:14:41]Starling satellites dodge on their own autonomously. The satellite sees a close approach coming and moves itself out of the way, and they're actually very cautious about it. SpaceX moves a satellite when the collision odds get anywhere near a few in 10 million. While the old industry habit was to start sweating somewhere around 1 in 10,000.

[00:14:59]That is crazy. And it's a big part of why those 300,000 dodges in 2025 state dodges. But it has the exact same blind spot that I keep running into in this whole freaking story. It only steers satellites that are alive and listening with dead ones, hacked ones, shrapnel.

[00:15:15]It can't touch any of them. So then we get to plan three. You go up and clean up the garbage. ESA is funding a mission called Clear Space 1, whose entire job is to fly up, capture one piece of ESA's own leftover junk, and drag it down to burn. A Japanese company called Astro Scale already flew a mission called Adas J that crept up on a dead rocket stage and inspected it from close range. And Rocket Lab is winning launch contracts in this little cleanup economy. Now, missions like these remove roughly one object each. They cost serious money.

[00:15:48]You're talking nine figures for clear spaces once single piece of junk. And all the while the industry is putting up thousands of new satellites per year. So you've got one down but thousands up every year. At that pace, you're never going to clean anything up. I'll be following this development for a long time for cleanup. And if you want to nerd out on this stuff with other like-minded people, head over to farza.fm fam to join my community with weekly Q&As's with me, exclusive content, a private Discord, and so much more. Link below. Now, it's super interesting to think about SpaceX's deliberate low parking spot. At roughly 550 km, there's still a whisper of atmosphere. There's barely anything, but it's there, and it drags. At that Starlink at that altitude doesn't need rules or grabbers or a rescue mission.

[00:16:37]Physics does the cleanup for free. drag bleeds off the speed, the orbit decays, and the thing falls and bursts up in about 5 to 6 years. It's literally a self-cleaning orbit. Now, what if the Kesler effect actually does happen? Who has the most to lose if this goes wrong?

[00:16:55]SpaceX by a mile. Starlink is the cash cow of that business, the business that's throwing off billions a year.

[00:17:02]It's bankrolling Starship. And then on the other side is Blue Origin with 52,000 satellites with Project Sunrise to build its orbiting data centered hardware. And they want to move it into that same exact shell. Not to mention SpaceX's 1 million plus satellite plan for orbital space compute. It would be an absolute disaster. So now the percentages, what's the likelihood that something like this actually happens?

[00:17:30]And the skeptics against something like this ever happening is actually quite strong. NASA has a long-term debris model called legend. And when researchers run it forward under current mitigation practices, it projects roughly linear debris growth through the year 2100. That means that they do not expect any sudden change in debris. That self-cleaning 550 km shell is real physics working in our favor. It also helps that most operators really are getting their dead hardware out of the sky on schedule. And the serious people in this field, the ones who spend entire careers on the remodels, will tell you that a lot of things have to fail at once before you get a true runaway. If they're right, the number is basically zero, less than 1%. and they might be right, but the reality is that the amount of stuff we're going to be putting in space, namely because of SpaceX, which is without a doubt a worldclass company, is going to increase dramatically. Starship, which is SpaceX's newest rocket, will make rocketry fully reusable, meaning that you can fly the same booster to space and back multiple times per day for weeks at a time. This will vastly increase humanity's ability to put stuff up on space. And SpaceX won't be the only company that's putting satellites on these rockets. It can be any company.

[00:18:48]And if it becomes cheaper and easier to put stuff in space, that means more and more companies will build satellites to put in space. So what's the likelihood even just one of those companies becomes a little careless over time as this technology gets easier and easier to access and use. Think about air travel.

[00:19:08]Airplanes are about as safe as anything in this world. Some statistics say it's more likely that you'll get hit by lightning in your lifetime than dying in a plane crash. But they still happen.

[00:19:19]Why is that? Because we have so many planes flying on any given day. And the chances of one of them running into a catastrophic error in any given year is essentially certain. The same thing applies here. But here's the deal.

[00:19:31]Humanity has always embarked on extremely dangerous things before. Look at the arc of history. Those adventures have led to the incredible life we all live in today. And even though the Kesler effect is a very real and dangerous possibility, the ark of history says that over time we will overcome it even through crazy failures and end up much better off on the other side. That is what I'm banking on today.

[00:19:58]I hope. Thank you so much for watching.

[00:19:59]I hope this video was informative and helpful. And we'll see you on the next one. Take it easy, everybody. Bye bye.

[00:20:04]Bye.