Idea and concept: Daniel Tibor Fuchs
Text developed and refined with the support of artificial intelligence
Final review: Daniel Tibor Fuchs

On Saturday, November 14, SpaceX did something no company and no government had ever done. Ship 42, the upper stage of Starship, spent 22 and a half hours in space, circled the Earth 14 times, released 26 Starlink V3 satellites and then flew back to Texas, where it was caught in midair by the arms of the launch tower, the famous Mechazilla. It was 3:20 pm local time.
In the hours that followed, X filled up with posts announcing a new world order. One of them, from the account @Kekius_Sage, said SpaceX had permanently altered the balance of global economics, politics and warfare, and that every finance minister and defense ministry on Earth had woken up to a new reality. Another post from the same account bet that within three years every major power will be forced to rewrite its defense doctrine.
I think the direction is right and the calendar is wrong. It is worth separating the two, because people who decide on enthusiasm tend to get the trend right and the timing wrong.
What actually happened
The catch of the ship was clean. It came through reentry at more than 17,000 mph, lit three engines, flipped to vertical and descended slowly until it settled between the arms of the tower. Three of the satellites it had deployed carried cameras and photographed the heat shield while still in orbit, so the team could know whether it was safe to bring the ship home.
But the flight as a whole was not perfect, and the most excited posts skip that part. The plan was to recover both stages on the same mission. The Super Heavy booster gave up on the tower and came down in the Gulf of Mexico because, according to Musk, the engines slightly underperformed and too little propellant was left for the maneuver.
There is another important detail. This was only the second orbital flight of Starship, and no ship has flown twice so far. The catch proves the ship comes home in one piece. It still has to prove it can take off again within days, without spending months in the workshop, and the whole cost equation depends on that. Orbital refueling is also still missing, and without it there is no Moon and no Mars.
Why it matters anyway
Until Saturday, every orbital rocket in the world threw away its upper stage, Falcon 9 included. Imagine an airline that recovers the wings and discards the cabin after every flight. When both stages come back to the tower, the cost of a launch starts to look like fuel plus maintenance, and that changes what is worth putting in orbit.
The planned numbers show the size of the ambition. Starlink has about 11,000 satellites today. The plan for the V3 generation talks about 100,000, and the Starmind project, for artificial intelligence computing in orbit, talks about one million. None of those plans add up with expendable rockets. With a reusable ship, they at least become debatable.
Where enthusiasm runs over the facts
The first point in those posts is about orbital data centers. It is true that SpaceX signed a contract with the German company Singulus worth a little over 60 million euros to manufacture solar cells in Texas. But the first production line arrives in the first half of 2027, and the 100 GW per year Musk mentioned is the size of the equipment order, not installed capacity. The first Starmind satellites are expected in 2028 at best. Cooling chips in a vacuum, where there is no air to carry the heat away, remains an engineering problem with no cheap solution.
The second point talks about a network that cannot be censored and that would bypass operators and regulators. Here I disagree more. Satellites still need spectrum, and spectrum is licensed country by country. In Brazil, Starlink operates under an authorization from Anatel, the telecom regulator, and in 2024, when the Supreme Court ordered X blocked, the company resisted for a few days and then complied. Whoever has customers, antennas and revenue in a country has a lot to lose there. What really changes is bargaining power, and that change is already big enough.
The third point is the military one. The program exists and it is called Rocket Cargo. It was created by the US Air Force in 2021, gave SpaceX a 102 million dollar contract and aims to deliver up to 100 tons anywhere on the planet in under one hour. It is still a research program. SpaceX itself has said it is not building Starship for that. And there is a difficulty none of the posts mention: a rocket descending over another country looks exactly like a ballistic missile on radar.
The second Starbase, in Louisiana, follows the same logic. The announced investment is 100 billion dollars, but construction starts in 2027 and the first launch is planned for 2029. The 10,000 launches per year are a target, and the company itself admits that American air traffic control cannot handle anything like that today.
What I take from this
None of this diminishes what we saw on Saturday. It only puts the dates in the right place. For those who want to turn the news into decisions, I leave five suggestions.
Watch four milestones before believing in the revolution. The same ship flying a second time, both stages caught on the same flight, orbital refueling and, the hardest of them, reuse without remanufacturing, the way an airplane lands, refuels and takes off again. Musk himself acknowledged on Sunday, after inspecting Ship 42, that it could fly again with a few weeks of refurbishment and that many small fixes are still needed to reach immediate reusability, the kind that requires only scheduled maintenance. Until those milestones arrive, low cost remains a promise.
If you work in telecom, treat direct satellite to phone connectivity as both a competitor and a partner. It arrives first where towers do not reach, and that is where coverage and IoT contracts will be fought over. But keep one caveat in mind. There are still no statistics on how devices perform indoors when talking directly to a satellite, the way we have had for years with 4G, 5G and other terrestrial networks. Without that data, nobody can say how much of the promised coverage holds inside a warehouse, a building or a vehicle.
If you are a regulator, discuss spectrum and data sovereignty now. Once the capacity is in orbit, the negotiation gets harder.
If you think about defense, look at satellite replenishment before you look at cargo in 45 minutes. Putting an entire constellation back in orbit within days is the advantage that arrives first.
Be suspicious of the dates, not of the direction. Musk is usually late. He rarely fails to deliver.
Space did not become the center of world power on November 14, 2026. But I bet that when that day comes, we will remember the image of that ship hanging from the tower as the moment it became clear where things were heading.

