Former SpaceX executive: the rocket was never the point — what Musk wants is all of space
SpaceX never thought of itself as a rocket company; the rocket is just the means. Falcon 1 was retired the moment it worked, Falcon 9 pushed cost per kilogram to orbit from $20,000 down to $3,000, and Starship is aiming below $100.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Landing the rocket is the real starting point
On December 21, 2015, SpaceX successfully put a rocket into orbit and recovered it for the first time. Hong Lide says that day mattered enormously to everyone at SpaceX, and that booster is now at the LA headquarters. His analogy: if a train or a plane were thrown away after a single trip, the whole thing could never be quantified. Only after the recovery worked did the entire reuse path begin. At the time many space experts in the industry thought what SpaceX was doing was unreasonable and made no economic sense, but a startup pulled it off.
— Hong LideStarlink exists to raise money for Mars
SpaceX had two purposes in building Starlink. First, going to Mars takes a great deal of money, and the nearest thing between Earth and space that can make money is a communications network. Second, from first principles, in the vacuum of space information travels twice as fast as through the most advanced fiber on Earth — a natural advantage. So Starlink was launched in 2015, exactly when rocket recovery was done and the company had cleared its first step. Hong Lide stresses that everything at SpaceX starts from first-principles thinking; it is not building a communications network for the sake of building a communications network.
— Hong LideUS data centers are stuck on power and permits
Building infrastructure in the US is extraordinarily hard. First, you have to secure the site, get approvals, bring power over, and confirm everything with everyone nearby — it takes a huge amount of time and energy. Second, the US is short on power: across the three big grids on the East Coast, West Coast and Texas, most facilities are over thirty years old, and the two newest nuclear plants were built in the 1990s. Even before AI, electrification demand had already created a clear grid shortfall, and adding AI makes it that much harder. Space's advantage is that expansion is limited only by how much you can launch, no permit is required, whoever gets there first gets the slot, and solar conversion efficiency in space is higher than on Earth.
— Hong LideMusk's companies may eventually merge into one X
Hong Lide tells a story: in 2014 a colleague said the boss had bought the X.com domain. His guess is that one day everything Musk owns may sit under a holding company integrated the way Alphabet is. He lists them: X has merged with SpaceX, Tesla is doing robots, autonomy and Physical AI, Neuralink is doing brain-computer interfaces, Boring Company digs tunnels. He speculates that the ultimate delivery of humans and AI may be the brain-computer interface, and that in the high-radiation environment of outer space people may need to live in mountains or underground, so the tunneling company has a place too. He stresses this is pure speculation, but ‘I don't think these companies are together by accident.’
— Hong LideLaunch cost per kilogram fell from $20,000 to $3,000
Before SpaceX, putting a kilogram of anything into space cost roughly $10,000 to $20,000. By recovering Falcon 9, SpaceX brought the cost down to $3,000 — about a tenth. Starship is meant to get below $100. Hong Lide says that when you go from $20,000 to under $100, a hundredth or two-hundredth of the old price, many things once dismissed as fantasy in space become projects you can actually pencil out. That, in his view, is SpaceX's real turning point: a massive cut in transport cost alongside a massive increase in the capacity and weight of the vehicle.
— Hong LideSpaceX's production team came from Mini Cooper
Hong Lide had the reporter guess: to build a rocket factory, which existing company in the world fits best? Boeing is out; Tyson, a food company, is one option; Mini Cooper, a car company, is another. The answer is Mini Cooper. Mini Cooper has the highest SKU count of any car line — roof, mirrors and interior color combinations are enormously varied — and its production line genuinely achieved high volume with high SKU variation control. The first team SpaceX brought in, and then expanded across the whole production and development process, was a group from Mini Cooper. Early Falcon 9s were never two alike; every launch collected data for the next vehicle and iterated immediately.
— Hong LideMusk only asked: have you seen a Coke can being made?
The team developed its own ultra-high-pressure cylinder, at 90 percent lower cost than the industry, with output that could keep up with Falcon 9, and proudly took it to show Musk. Musk's first sentence was: have you ever seen a Coke can being produced. Hong Lide says a Coke can, from first principles, is also a pressure vessel — so why can it be made a thousand a minute at a cost of a few cents. From a strict engineering standpoint, you cannot argue with him. It is not an insult; it is constantly pushing you to a level you did not know you could reach. The team was still optimizing after walking out of the meeting room.
— Hong LideOne 'impossible' and Musk accepted the resignation on the spot
Hong Lide witnessed it once: a fairly senior engineer reported that progress was behind expectations, and Musk began to push back, asking why not do it this way, why not think of it that way. In the end the man said flat out that what you want to build is impossible, it cannot be done. Musk, very calmly, said: fine, then I accept your resignation. Hong Lide says the air froze for the next thirty seconds. Musk did not yell, did not slam the table, did not throw a chair — he does not yell. The outside world thinks he is a tyrant, but that is not what Hong Lide saw.
— Hong LideNASA's own rocket costs five times more than Starship
Hong Lide uses the Moon rocket as an example: NASA's own SLS started in 2011, took nearly twenty-some years, cost more than five times what SpaceX did, with $25 billion still counting, and produced a rocket far inferior to Starship, in both payload and recovery. It is a very typical showcase product of the old-space era of the American space program. Once NASA realized this, it shifted from government-led before 2000 to government-assisted after 2000, and the first showcase of that was the resupply contract for the space station given to SpaceX. The International Space Station was established in the early 2000s, cost nearly $100 billion, is due to retire in 2030, and NASA has decided not to build a new-generation station, letting private companies take the orders.
— Hong LideIn their own words · checked verbatim
Before SpaceX, putting a kilogram of anything into space cost about ten or twenty thousand dollars — one kilogram, ten or twenty thousand dollars. Then SpaceX showed up and said this is absurd, OK, we are absolutely going to bring this down through Falcon 9, through recovery. And SpaceX really did it. SpaceX took that cost, OK, straight through recovering Falcon 9, and brought it to $3,000.
在SpaceX之前 每发一公斤东西上太空 差不多是一两万美金的事情 一公斤 就一两万美金 SpaceX出现 就是说这是一个 非常非常扯的事情 OK 我们一定要通过 Fain 19号 通过回收 我们要把这个东西 大大降幅 那SpaceX的确做到 这个东西 SpaceX 把这个成本 OK 直接透过了回收 Fain 19号 涨到了3000元
Hong Lide1:15:44
If you told Elon, today I added thirty percent to this project, OK, that kind of efficiency — Elon would tell you to get out.
你跟英朗说 我今天给这个项目 增加了百分之三十 OK 这样的效率 英朗会让你滚出去
Hong Lide1:36:56
Elon's first sentence was to ask: have you ever seen a Coke can being produced? That was his only answer.
伊朗第一句话来问 你 有没有看过可乐罐的生产 是他唯一的一个回答
Hong Lide1:43:59
He just said it, very calmly: fine, then I accept your resignation.
他直接来一句 就很淡定的 说好 那我接受你的辞呈
Hong Lide1:51:01
Our competitor is the speed of our own progress. OK, that is the exact quote. But we closely follow the development of China's space program. Exact quote.
我们的竞争对手 是我们自己的进展的速度 OK 这一句话是原话 但是 我们密切关注中国航天的发展 原话
Hong Lide2:22:20
Because I think SpaceX is at least a trillion — at least a trillion-something company. Here on Earth, that is, my application — at least, at least, I said it many times, at least.
因为我认为SpaceX 至少是一个万亿 至少 是一个万亿几的公司 在地球上面的这个 就是我的应用 至少 至少 说了好多遍 至少
Hong Lide2:40:31
Why did no one build Starship before? Why did no one build a reusable rocket before?
为什么之前没有人做新建 为什么没有之前没有人做回收火箭
Hong Lide2:50:34
Figures
| Date of first successful rocket recovery | December 21, 2015 | 5:06 |
| SpaceX launch cost per kilogram (after recovery) | $3,000 | 1:15:44 |
| Launch cost per kilogram before SpaceX | $10,000 to $20,000 | 1:15:44 |
| Starship target launch cost per kilogram | below $100 | 1:15:44 |
| Falcon 9 payload weight multiple over Falcon 1 | 25x | 1:12:43 |
| NASA SLS spending | $25 billion | 2:27:22 |
| Total number of spaceports worldwide | 77 | 2:20:20 |
| Number of 10-cm debris objects in low orbit | over 30,000 | 2:58:42 |
| SpaceX headcount | 20,000 | 2:40:31 |
Glossary
- reusable rocket
- A rocket that can return after launch and be used again; the core of SpaceX's cost reduction.
- SKU
- The number of product models or configurations; the more SKUs, the more complex production management becomes.
- first principles
- Reasoning from the most basic facts of physics rather than copying existing practice.
- Class Plus
- A cost-plus contract where the government pays cost plus profit, commonly used by traditional aerospace firms.
- Elon Mafia
- The group of young entrepreneurs who came out of SpaceX or Tesla and are building hard tech in the US.
How to listen
Founders and investors watching commercial space, hard-tech investment and Musk's management style, especially anyone who wants the inside development history of SpaceX and its cost curve.
The opening self-introductions and Hong Lide's early years at Epson and in food startups can be fast-forwarded.