Zhang Xiaojun | 张小珺Jùn｜商业访谈录 Summary

Guest: Lewis Hong (洪力德), Co-Founder and GP of Aerys Fund; former Chief Manufacturing Engineer at SpaceX (2012–2019) Host: Zhang Xiaojun (张小珺)

1. Key Themes

The 2015 Reusable Rocket Landing Was Space's True Inflection Point — Not Today's IPO

Lewis argues that December 21, 2015 — when SpaceX successfully launched and recovered a Falcon 9 — was the real paradigm shift for the entire space industry, not the current IPO moment. The IPO is merely bringing that earlier revolution to public consciousness.

"December 21, 2015 — every SpaceX person remembers this date very clearly. We successfully sent a rocket into orbit and successfully recovered it. That rocket is now in LA. Friends visiting LA can actually go to SpaceX headquarters and see this historically significant rocket." 00:05:14

"In our industry, you could say this was even bigger than GPT-3. Because think about it — if today any form of transportation, after going from point A to point B, you just throw it away — whether it's a train or a plane — could that industry ever scale? It's impossible. But when you can recover it, that's the starting point for everything." 00:05:52

SpaceX Was Planning Its Next Generation While Still Struggling With the Current One

By 2012–2013, when SpaceX could barely launch once a year, the Mars group had already been formed and Starship was already being conceived. This multi-generational planning is a defining organizational trait.

"When our first rocket had just barely succeeded in launching — the whole company could barely launch once a year — the company had already established a Mars team. The company had already decided that this rocket couldn't reach Mars... So at that time, in 2012, what is now called Starship was already born." 00:07:38

"Everything you see at SpaceX now — Starship, Starlink, space data centers, going to the moon — people think Musk came up with these randomly one night while bored on X. But these possible paths were all discussed more than a decade ago at SpaceX." 00:08:21

Falcon 1's Deliberate Retirement Was a Masterclass in Mission-Driven Decision Making

SpaceX retired Falcon 1 immediately after its first success in 2008 — a decision that seems counterintuitive but reveals the company's core philosophy. The rocket was a means, not an end.

"SpaceX never positioned Falcon 1 as the company's core purpose. SpaceX never positioned itself as a rocket company. It never treated the rocket business as the company's purpose. The rocket is something that must be completed to reach its destination, but the rocket is only a process." 01:11:13

"After Falcon 1 succeeded, SpaceX immediately retired it. Why? Because SpaceX's goal requires something much more powerful. Even the US government at that time told Musk — just forget about rockets, come help with defense or other sectors. SpaceX flatly refused: that is not our mission." 01:18:43

The Real Rocket Manufacturing Revolution: Mini Cooper, Not Boeing

SpaceX's breakthrough in high-volume rocket production came not from aerospace expertise but from automotive manufacturing — specifically Mini Cooper's high-SKU, high-volume production system.

"SpaceX, the first team they brought in to expand the entire SpaceX production and development process was actually a team from Mini Cooper. Mini Cooper has a characteristic — in its vehicle line, it has the highest SKU count of any car line. The SKU variation is the highest of all car lines. That high-volume but high-variation production control system was the most compatible with SpaceX's needs." 01:24:03

"In SpaceX's early days, no two Falcon 9s were identical. Every Falcon 9 launch collected data — right, wrong, good, bad — and immediately incorporated everything into the next Falcon 9. Immediately improved, immediately iterated, immediately launched. So across hundreds of early flights, no two Falcon 9s were the same." 01:26:06

Musk's Management Style: Not a Tyrant — A Flat, Calm, Relentless Standard-Raiser

The popular image of Musk as a screaming tyrant is wrong. The reality is a deeply introverted, emotionally flat man who fires people quietly and uses first-principles Socratic challenges rather than rage.

"If you told Elon you improved efficiency by 30%, Elon would tell you to get out. You haven't been thinking at all." 01:37:09

"Elon doesn't yell. He doesn't flip tables. He just very calmly said: 'Good, I accept your resignation.' Think about what the next 30 seconds in that room felt like." 01:51:59

"The 2012–2013 Musk was an extremely introverted person. So introverted that even in his own company, doing an annual address, you could feel Musk was extremely uncomfortable. You in the audience felt more uncomfortable than he did on stage." 01:02:22

Space's First-Principles Case for the Next AWS

Lewis makes a systematic, physics-based argument for why space communications and computing are not hype — data transmits roughly twice as fast in the vacuum of space versus the best fiber on Earth, and space offers unlimited solar energy and zero permitting constraints.

"Information transmission from point A to point B in the vacuum of space is still about twice as fast as the most advanced fiber optics on Earth — 5G, 6G, whatever G. That is a natural advantage created by the space environment itself." 00:13:08

"In space, the only thing controlling your speed of expansion is how much you can launch. In space, you don't need to get permits, you don't need various approvals. Whoever gets there first, whoever claims that position first — that's it." 00:17:37

"Space has limitless solar energy. Solar conversion in space is proven to be more than ten-plus percent higher than on Earth. Limitless energy, limitless space, and no restrictions on development." 00:18:35

SpaceX Is the East India Company of the Space Age

Lewis's most provocative valuation thesis: SpaceX is not a rocket company or even a tech company — it is the dominant commercial power of a new geographic domain, analogous to the East India Company during the Age of Sail.

"SpaceX is the East India Company of the Age of Exploration — the company that crossed from Europe to the Americas, with enormous financial and influence power across all domains. And it's only just getting started." 02:22:02

"SpaceX is worth at least a trillion dollars — at least. I've said 'at least' many times. And that's just from Earth-based applications. If I don't need to sell, I will never sell." 02:41:44

Musk's Companies Are Not Separate — They Are One Master Plan

Lewis believes all of Musk's companies — SpaceX, Tesla, X/xAI, Neuralink, Boring Company — form a coherent, undisclosed master plan, like Alphabet but with a deeper civilizational thesis.

"In 2014, a colleague told me: the boss bought X.com today. Not one of his companies has an X in it — but SpaceX, X, xAI... I don't think these companies are coincidental. Perhaps one day all of Musk's things will, like Alphabet, have a holding company and all merge together into one X." 00:19:49

"Tesla does physical AI — robots, autonomous driving. Neuralink does brain-computer interfaces. Perhaps that is the ultimate interface between humans and AI. Boring Company — in high-radiation outer space environments, humans will need to live underground. Perhaps all of Musk's companies are serving one single grand plan." 00:21:35

America's Hardware Renaissance Is Being Led by the SpaceX and Tesla Mafia

Lewis's core investment thesis: a generation of young engineers trained under Musk now know how to build advanced hardware in America. This is historically unprecedented and investable.

"SpaceX and Tesla created a cohort of young entrepreneurs who know how to build advanced hardware in America. Everyone thought manufacturing advanced hardware in America was basically impossible. These entrepreneurs — in their 20s and 30s, very young, with both brains and resources — their next projects are what we call the SpaceX Mafia, and they are what our fund focuses on." 02:42:47

2. Contrarian Perspectives

Musk's Self-Published Biography Is Inferior to the Third-Party Account

Against the obvious intuition that more access and more resources produce a better book, Lewis argues the authorized Musk biography missed what actually matters.

"I personally think the Elon Musk autobiography had ten times more resources and time invested. But I don't think it performed as well as Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future. He didn't focus on what should be the most important essence of Musk's companies — which is what we discussed today — but instead on things he was personally curious about, like politics and the X acquisition, which I don't think are the most important parts of Musk." 02:54:02

The Hardest Period for SpaceX Was Not the Near-Bankruptcy of 2008 — It Was 2015–2016

Most people assume the Falcon 1 near-death experience was SpaceX's darkest hour. Lewis was there and disagrees.

"Everyone thinks SpaceX's most difficult moment was the early Falcon 1 days up to 2008. No. The most difficult moment was exactly in 2015–2016. We had already experienced multiple successes. We were the industry leader. And then we failed. When you fail, you can't ask anyone — because no one is ahead of you. We were the most advanced, and nobody knew what went wrong." 02:00:05

Expertise in the Field You're Disrupting Is a Liability, Not an Asset

SpaceX deliberately avoided hiring traditional aerospace professionals as its core team — people with decades in the industry were actually harder to integrate than fresh graduates or people from unrelated fields.

"It's not that they weren't capable. It's that this company's culture and the traditional rocket company culture are basically two completely different companies... The people SpaceX was really looking for were from outside the industry — bringing fresh thinking into the industry. So SpaceX actually recruited from all walks of life." 01:07:57

A 30% Efficiency Gain Will Get You Fired at SpaceX

Standard corporate performance metrics are actively counterproductive at SpaceX. The minimum acceptable unit of improvement is an order of magnitude.

"In many other companies, if you tell your boss you saved 10%, 20%, 30% efficiency on a project — great, promotion, bonus. You tell Elon you increased efficiency by 30% — Elon will tell you to get out. 'You haven't been thinking at all.' The standard at SpaceX: if you haven't achieved an order-of-magnitude jump — 10x or 100x — you haven't been doing your job." 01:37:09

Grok Is the Most Underestimated AI Lab Going Into 2026

While the market views xAI/Grok as a laggard, Lewis argues the 2025 infrastructure buildout — 100,000 GPUs — is about to surface results that will surprise people.

"In 2026 predictions, I believe this year's Grok is something everyone should watch closely, because I think a lot of the investments and buildout Grok did last year will emerge this year. They built a 100,000-GPU super data center so fast — where did all that compute go? Let's see if something comes out in 2026." 00:26:43

3. Companies Identified

SpaceX

Aerospace manufacturer and space transportation company founded by Elon Musk. The central subject of the episode; Lewis worked there 2012–2019, managing 7 departments and 3,000+ rocket components. Cited as the East India Company of the space age, worth at least $1 trillion, having reduced cost-per-kg to orbit from $10,000–$20,000 to $3,000 (Falcon 9) with a Starship target below $100.

"SpaceX is the East India Company of the Age of Exploration. Right now it's only just getting started." 02:22:02

Aerys Fund

Early-stage VC fund based in Los Angeles, co-founded by Lewis Hong. Focuses on American hard tech — space, energy, Physical AI. Distinguishes itself by integrating space and Earth-based investments as one system.

"We are an early-stage US VC based in LA. We specialize in investing in American hard-tech projects — whether in space, energy, or what is now called Physical AI." 00:01:29

Rocket Lab

Commercial launch company; its Electron rocket is cited as evidence that Falcon 1's payload class (under 1 metric ton to orbit) has a real commercial market — the very market SpaceX chose to leave.

"Rocket Lab's Electron rocket is now launching perhaps 75 times a year. Its payload capacity is less than one-third of Falcon 1. So Falcon 1, from the beginning, actually already had a very obvious, monetizable use case — SpaceX just chose not to stay there." 01:18:13

Blue Origin

Jeff Bezos's space company; cited as proof that commercial space tourism has lower physical barriers than people assume — older passengers, people from non-aerospace backgrounds, have already flown.

"Blue Origin has already proven that the physical requirements for going to space aren't that high — in terms of age or different professional backgrounds." 00:58:16

Tesla

Electric vehicle and AI company; described as SpaceX's "sister company" under Musk. The Tesla Model 3 experience triggered Lewis's career pivot from SpaceX to the EV/battery sector.

"From a mechanical engineering perspective, I believed the EV platform would not be a gimmick or a Tesla-exclusive platform. In every dimension it had greater potential than the combustion car. At that moment I decided: after SpaceX, I want to join an EV company." 02:09:26

xAI / Grok

Musk's AI lab, recently merged with SpaceX. Lewis argues Grok's unique advantage is Physical AI and world models, given Musk's ecosystem data advantage. He predicts 2026 will be a breakout year.

"Grok, having built a 100,000-GPU super data center so quickly — where did all that compute go? Maybe in 2026 we'll see what comes out." 00:27:10

OpenAI

Cited as the consumer-facing general AI incumbent. Sam Altman's dismissal of Musk's space compute thesis is attributed to competitive jealousy.

"I think he's just envious. OpenAI also... I won't say too much trash talking." 02:37:00

Anthropic (Claude)

Cited as the AI lab most focused on a B2B / coding tools monetization path from day one.

"Claude very clearly from the beginning wanted to be a tool usable in coding — a B2B approach from the start." 00:24:57

Google / Gemini

Cited as leveraging Search dominance to anchor its AI position in text and search.

"Gemini has Google Search, so it wants to dominate in search and text-based applications." 00:24:28

Epson

Japanese electronics conglomerate; Lewis's first employer. He became the youngest executive in the global Epson group in his second year, managing a 70+ person US ink cartridge manufacturing facility.

"At that time I was in my early twenties managing the whole factory — about 70-plus people. Within the entire conservative Japanese Epson global organization, this was unimaginable." 00:36:44

Amazon / Amazon Japan

Lewis turned down an offer to help launch Amazon Japan in 2012 to join SpaceX instead. His former Epson boss was overseeing Amazon's Japan launch.

"My former boss said: we are just starting Amazon in Japan. This is an opportunity you will not encounter a second time in your life. Come quickly, come join me to build Amazon here." 01:47:10

Boeing / ULA / Lockheed Martin

Traditional aerospace incumbents. At SpaceX's scale in 2012, Boeing/ULA employed 10x as many people doing equivalent work. Cited as examples of cost-plus contract failure.

"Compared to Boeing, Lockheed Martin, ULA — organizations doing the same competitive launch work — they had ten times as many people doing the same thing. SpaceX was a one-to-ten ratio." 01:04:52

NASA

US space agency. Described as the pivotal institutional enabler that shifted from government-led to government-supported (private sector-led) space programs, starting with the CLD commercial resupply contract to SpaceX.

"Before 2000, America was a government-led space program. After 2000, it became a government-assisted space program. The first landmark example was SpaceX receiving the first contract to resupply the International Space Station." 02:28:46

Neuralink

Musk's brain-computer interface company. Lewis speculates it represents the ultimate human-AI interface in Musk's master plan.

"Perhaps the most direct way for humans to interface with AI is a brain-computer interface — a direct connection. Imagine always having a computer's supplementary intelligence updating you in real time." 00:21:19

Boring Company

Musk's tunneling company. Lewis suggests it serves the Mars habitat use case — underground structures to shield humans from radiation on high-radiation planets.

"In high-radiation outer space environments, humans will definitely have to live underground or in mountains to avoid radiation problems. Perhaps all of Musk's companies are serving one grand plan." 00:21:35

4. People Identified

Lewis Hong (洪力德)

Co-Founder and GP, Aerys Fund; former SpaceX Chief Manufacturing Engineer. Stanford mechanical engineering graduate (class of 2003, one of only two students admitted from mainland China that year). Became youngest global executive at Epson in his second year of work. Joined SpaceX in 2012 as employee ~1,000; left in 2019 managing 7 departments covering 3,000+ rocket components. Subsequently joined an EV battery startup that caught the EV wave, exited in 2020, then co-founded Aerys Fund. Cited throughout as the primary source of operational SpaceX insight.

"When I left SpaceX I was managing 7 departments — more than 3,000 rocket components." 00:01:29

Elon Musk

CEO of SpaceX, Tesla, xAI; owner of X. Lewis provides an unusually grounded portrait: a deeply introverted man in 2012–2013 who became the world's greatest storyteller through deliberate practice; emotionally flat (not a tyrant); fires people quietly; insists on 10x–100x improvements; never holds status meetings; appears only when there is a critical problem to solve.

"Musk's attitude: his time is only allocated to the company's hardest problems. All his companies — wherever the hardest, most unsolvable problem is, where everyone needs to work together — that's when Musk appears." 02:51:49

Buzz Aldrin

Second man to walk on the Moon (Apollo 11). Cited as an example of the aerospace establishment that initially dismissed SpaceX.

"Buzz said SpaceX would ultimately screw things up because they didn't know what they were doing — a bunch of Silicon Valley cowboys thinking they could innovate in space. Space is something that does not tolerate any errors." 02:24:44

Jeff Bezos

Amazon founder; cited in the context of Blue Origin proving space tourism is accessible beyond billionaires.

"Going to space shouldn't require being Jeff Bezos to do it." 00:57:33

Sam Altman

CEO of OpenAI. Mentioned for dismissing Musk's space compute thesis as "laughable."

"I think he's just envious." 02:37:00

5. Operating Insights

The "Accepted Your Resignation" Firing Technique Eliminates Ambiguity and Sets Standards Without Drama

Musk's firing method — calm, immediate, zero emotional escalation — accomplishes something conventional management cannot: it makes the standard crystal clear to everyone in the room without creating martyrs or confusion.

"Elon very calmly said: 'Good, I accept your resignation.' Think about what the next 30 seconds in that room felt like — everyone sitting there as senior managers watching this." 01:51:59

The lesson for operators: the moment someone declares something impossible in a high-stakes environment, their departure should be immediate and undramatic. Anger and drama dilute the signal.

Musk Only Holds Three Types of Meetings — All Crisis-Driven

Musk does not do status updates or relationship maintenance meetings. His calendar only contains problems: late deliveries, failed builds, and over-budget items. Silence from him means you are performing.

"Elon only meets with you about three things: first, your thing is late. Second, your thing can't be built. Third, your thing needs more money. Those are the only three meetings with Elon — never a casual conversation like this one. If Elon is not talking to you, you're doing fine — just keep going." 01:38:32

Cross-Industry Hiring as a Manufacturing Superpower

The Mini Cooper hiring insight is directly actionable: when building a high-mix, high-volume production system, the relevant expertise is not in the target industry but in whatever industry has already solved high-SKU variation at scale.

"SpaceX, the first team they brought in to expand production was a Mini Cooper team. Mini Cooper has the highest SKU count of any car line — and they had achieved high-volume production while managing high variation. That was the most compatible system with SpaceX's needs." 01:24:03

The "Coke Can" Standard: Always Reference a Commodity Benchmark to Drive Cost Reduction

When Lewis's team achieved a 90% cost reduction on a hyperpressure vessel — an unprecedented feat — Musk's only response was to ask if they'd studied Coca-Cola can production. The lesson: always identify the commodity analog for your high-spec component and use it as a north star, even if the gap seems absurd.

"Elon's first and only response was: 'Have you looked at Coca-Cola can production?' A Coke can is also a pressure vessel. From first principles it is the same thing. Why can a Coke can be produced at thousands per minute for pennies? You still have room to improve." 01:44:52

The Staffing Principle: Hire for Learning Speed and Adaptability, Not Domain Experience

SpaceX's edge came from preferring people who could absorb new domains rapidly over people with deep but domain-specific experience. The question is not "what do you know?" but "how fast can you grow?"

"SpaceX values how much a person can absorb in a short time, how much they can grow, how much they can adapt — more than decades of prior experience. Why? Because experience gives you a higher starting point, but SpaceX never defined itself as any other kind of company. If you told Elon you improved 30%, he'd tell you to leave." 01:36:40

6. Overlooked Insights

Space Debris Is an Underappreciated Near-Term Threat to the Entire Low-Earth-Orbit Economy

Lewis drops a single sentence about low-orbit debris that implies a critical infrastructure risk — and an investable problem — that received no follow-up in the conversation.

"You guess — right now in low orbit, how many objects roughly ten centimeters in size are floating around? Over 30,000. Moving at hypersonic speeds. So in low orbit, this is a problem that needs to be solved and overcome." 02:58:45

This is not a science fiction problem. With SpaceX targeting hundreds of annual launches, Blue Origin, and 77 national space agencies all adding hardware to low orbit, the collision risk and debris cascade (Kessler Syndrome) becomes an existential constraint on the entire commercial space economy. There is no major funded company in Lewis's portfolio discussion focused on active debris removal — which suggests a genuine white space.

SpaceX's Fastest Path to a Trillion-Dollar Valuation Is Semiconductor Manufacturing in Space, Not Rockets

Lewis briefly mentions investing in a company trying to manufacture next-generation silicon wafers in microgravity — using the natural space environment to achieve what TSMC spends tens of billions to simulate on Earth. This was stated and immediately passed over, but it is potentially the highest-value single application in the entire space economy.

"We're looking at a company hoping to use the space environment to develop next-generation wafers. The wafer's purity and crystallinity determine downstream chip performance. The environment you're spending tens of billions to create in a fab on Earth — isn't it very similar to what space already naturally provides? You're essentially spending enormous resources to simulate a space environment on Earth. Why not just do it in space?" 02:39:21

The implication: a single microgravity wafer facility, if it can produce chips of meaningfully higher quality than terrestrial fabs, would be worth more than all current launch revenue combined — because the semiconductor industry is a $600B+ annual market and wafer purity is the binding constraint on next-generation chip performance. This is the "overlooked" payload that could make the entire launch infrastructure economically justified even without Starlink.