Sam Parr | My First Million Summary

My First Million – Sam Parr & Shaan Puri

1. Key Themes

SpaceX Is Not a Rocket Company — It's Three Companies Stapled Together

The hosts frame SpaceX not as a single aerospace business but as a bundle of fundamentally different businesses that happen to share an owner. The launch business enables the satellite business, which funds the AI ambitions.

"You're not buying SpaceX. You're buying three different companies that are stapled together. You have the space launch business, the internet connectivity business, and the AI business." — Sam Parr 00:29:08

Starlink Is the Quietly World-Class Cash Engine Nobody Talks About Enough

Starlink has all the hallmarks of a generational SaaS business — recurring revenue, ~40% EBITDA margins, no real competitors, an extreme cost advantage — and it's only four years old.

"The Starlink business has like all of the wonderful characteristics of a business. There's no competitors. They have an extreme cost advantage. It's recurring revenue. It's extremely high margin. And it's a product that everybody on Earth essentially needs is the Internet." — Sam Parr 00:09:57

Orbital Data Centers: SpaceX's Trillion-Dollar Next Act

The core thesis of SpaceX's forward value is that compute delivered from space — powered by the sun, cooled by the void — will be structurally cheaper than any terrestrial data center. Elon's regulatory frustration with land-based buildouts is the strategic trigger.

"It is easier to figure out how to launch the heaviest rocket ever and take and build a data center in space than it is to get like Alameda County to approve of a data center in your backyard." — Sam Parr 00:15:08

"Basically if you look at when you have a satellite in space, it's powered by the sun. You don't need any cooling because space has like space is basically freezing cold and has like a radiative cooling mechanism where as the chips get hot from being used, there's like a natural physical mechanism to cool them down... streaming AI tokens from space is going to be far cheaper and more effective than any land buildout of data centers." — Sam Parr 00:20:49

SpaceX as the Saudi Arabia of Compute

Sam frames the next 20 years not around oil as the world's scarce energy resource, but compute — and positions SpaceX as the potential lowest-cost, dominant-scale provider of that compute, analogous to Saudi Arabia's oil position.

"Saudi Arabia made trillions of dollars because they owned the oil. They were the largest and lowest cost producer of oil and energy and the world ran on energy. Well, if you look at the next 20 years, it seems like the world is going to run on compute... Then he's Saudi Arabia in space. And he's the only one positioned to do that currently." — Sam Parr 00:33:42

Elon's Superpower: No Market Risk, Only Execution Risk

A recurring and important thread throughout the episode is that Elon consistently bets on things where demand is obvious and the only question is whether he can build it — and he has an extraordinary track record of answering that question in the affirmative.

"The only question is, can I make it? And the answer for him has always been yes." — Sam Parr 00:21:18

Colossus Becoming an Accidental Profit Center Via Rentals to Google and Anthropic

Elon built the world's largest GPU cluster, Colossus, to power Grok — which hasn't matched ChatGPT or Anthropic's scale. Rather than letting the asset sit idle, he rented it out to his direct AI competitors. Two deals alone are reportedly worth a combined $20B+.

"He just turned into Airbnb. He started renting out Colossus to Anthropic and to Google. And so just in the last two months, they announced two deals that I think are combined worth like $20 billion or more. A billion dollars a month. So $12 billion a year to use his data center. And that's Google. Google themselves was the best at building data centers. And they need to rent from Colossus." — Sam Parr 00:27:09

The Elon Premium Makes Traditional Valuation Frameworks Break Down

At 100x revenue and $1.75 trillion market cap, the standard tools of analysis don't apply. The hosts argue you have to either reject the valuation or accept a new framework — the "price to Elon ratio."

"It's not about the price to earnings ratios. It's not about price to sales. It's about the price to Elon ratio. And if it's an Elon company, you're going to have 10 times the price you would probably otherwise have in a company." — Sam Parr 00:06:50

Simplicity Beats Sophistication in Investing — The Gigafund Lesson

A quietly profound investing insight: Luke Nosek left Founders Fund and built Gigafund around one single, seemingly unsophisticated idea — back everything Elon does. The most powerful move in hindsight was the one that looked the least clever.

"The optimal way for him to invest was simply to just back every Elon company. And that's it. Take all the money and put it into the Elon companies. And at the time, that seemed like kind of crazy because Elon was doing pretty crazy companies. And in addition, it also seemed not sophisticated, which I think is the underrated part of this." — Sam Parr 00:52:48

Direct-to-Cell Could Disrupt the Entire $2 Trillion Telecom Market

Beyond home internet, Starlink's satellite-direct-to-phone capability removes the need for terrestrial towers entirely — potentially the first genuinely differentiated product in a market defined by sameness.

"Whether they launch their own T-Mobile or it's this little add-on to every cell phone plan on earth where it's like for an extra three bucks a month, five bucks a month, ten bucks a month, you just have guaranteed coverage everywhere... The market for internet services across cell phones and home internet is like $2 trillion." — Sam Parr 00:12:39

2. Contrarian Perspectives

Adjusted EBITDA in a Capital-Intensive Business Is Fiction

While most analysts and presentations highlight adjusted EBITDA as a key metric, both hosts argue this is particularly misleading for SpaceX, which spent ~$20 billion in CapEx in the last 12 months. Adding back depreciation on physical rockets and infrastructure is not a legitimate adjustment.

"Buffett and Munger famously hated EBITDA as a thing. I think they just called it bullshit earnings... this is not a software company. And then the adjustments, you're right. Oh, except for we adjust back in for stock-based compensation. It's like, well, that's how you pay the people to do the work. So why do you keep adding that back in?" — Sam Parr 00:41:09

The Biggest Risk Is Not Technical Failure — It's That Elon Dies

Counterintuitively, the hosts argue the engineering risks (Starship, orbital data centers) are smaller than a single biological risk — Elon's continued existence. No succession plan or second founder exists at the same capability level.

"I, if I'm a shareholder, I'm like, you have, you must have a dietitian. You must have a sleep coach... The biggest risk is that he dies." — Shaan Puri 00:04:43

Betting Against Elon's Technical Ability Is the Most Historically Unprofitable Bet You Can Make

At a time when many rational analysts scoff at Starship timelines and orbital data centers, the hosts argue the record shows that being a skeptic of Elon's engineering ambitions has been financially catastrophic, even when you've been temporarily right.

"Betting against Elon's technical ability has proven to be like the most unprofitable bet you could make. Even if you're right for a year or two, you're eventually wrong." — Sam Parr 00:18:29

Twitter Dragging the Bundle Down Is Actually Fine Because It Got Rolled Into Something Bigger

Conventional wisdom holds that Elon's Twitter acquisition was a disaster. The contrarian read is that the bundle of assets makes the bad Twitter trade largely irrelevant — like putting a broken footrest in a king-size super bed.

"The only downside when you look at the analysis of this thing is that Twitter revenue is like half of what it was when he bought it. So that one hasn't really worked out, but he rolled it together with all this other great stuff so that the investors did well, even though that business didn't do so hot." — Sam Parr 00:06:01

Failing Forward Is a Legitimate, Underrated Strategic Competency

Rather than seeing Grok's failure to match OpenAI as a waste, or the Colossus buildout without users as overreach, the hosts argue that Elon's ability to continuously repurpose underperforming assets is itself a durable competitive advantage.

"I think one of the great hallmarks of entrepreneurship is basically failing forward... He used his Grok asset where he's trying to build a ChatGPT Anthropic competitor, even though it's way, way behind... Are we fucked? No. What if we did it this way? And then finding a way to survive and advance." — Sam Parr 00:25:58

3. Companies Identified

SpaceX

Vertically integrated space company and parent entity going public at $1.75 trillion. Controls 85% of mass to orbit, owns Starlink, Colossus, and is developing orbital data centers and Starship. $18B revenue, $6.6B adjusted EBITDA, losing $2.5B on a GAAP basis.

"You're not buying SpaceX. You're buying three different companies that are stapled together. You have the space launch business, the internet connectivity business, and the AI business." — Sam Parr 00:29:08

Starlink

SpaceX's satellite internet division. 10 million paying subscribers, $11B/year revenue, ~40% EBITDA margins, and only ~4 years old. The genuine cash engine of the SpaceX bundle.

"The Starlink business has like all of the wonderful characteristics of a business. There's no competitors. They have an extreme cost advantage. It's recurring revenue. It's extremely high margin." — Sam Parr 00:09:57

Cursor

AI coding tool. SpaceX has an option to acquire Cursor for $60B. Currently at $3–4B in revenue. Mentioned as a move to gain ground in the AI race, providing Cursor with compute and models.

"They bought Cursor for $60 billion or have an option to buy Cursor for $60 billion. And Cursor, I think, is at $3 or $4 billion in revenue." — Sam Parr 00:30:06

Colossus (XAI Data Center)

World's largest GPU cluster, built by Elon's XAI division. Now being rented to Google and Anthropic for a combined ~$20B+ in deals, at roughly $1B/month from each. A failed internal asset turned accidental profit center.

"He built the largest cluster of GPUs, bigger than Google, bigger than Facebook, bigger than anyone. The problem is he doesn't have enough users... He just turned into Airbnb." — Sam Parr 00:26:35

Anthropic

Leading AI company (Claude). Renting compute from Colossus at reportedly over $1B/month.

"Google themselves was the best at building data centers. And they need to rent from Colossus. And then same thing with Anthropic, where Anthropic is paying, you know, over a billion dollars also a month to rent out." — Sam Parr 00:27:19

Gigafund

Venture fund founded by Luke Nosek (ex-Founders Fund) and Steve. Strategy: exclusively back Elon Musk's companies. Listed as a significant stakeholder in the SpaceX S-1.

"It was called Gigafund and it was just fund Elon... the optimal way for him to invest was simply to just back every Elon company. And that's it." — Sam Parr 00:52:23

Founders Fund

Peter Thiel's venture fund. Gigafund spun out of Founders Fund when Luke Nosek departed.

"Gigafund spun out of Founders Fund. Luke was at Founders Fund, and then he left to start Gigafund." — Sam Parr 00:52:48

Valor Equity Partners

Antonio Gracias's fund. Antonio is SpaceX's second-largest individual shareholder at ~7% ownership. Will net approximately $90B from the IPO.

"The second biggest shareholder of SpaceX is Antonio, who owns about 7% of the company... he's going to make like $90 billion in this IPO." — Sam Parr 00:49:27

Tesla

Elon's electric car company. Referenced as the template for Elon's "machine that builds the machine" philosophy — designing and producing at speed that no one else matches.

"He's really good at building the machine that builds the machine. So he's basically the, what he's the best at is building factories at this point." — Sam Parr 00:26:24

X (formerly Twitter)

Social media platform acquired by Elon. Ad revenue now $1.8B, down from $4.5B when purchased. Total revenue ~$2.8B including subscriptions. Has underperformed but provided data for Grok.

"Twitter's ad business is 40% the size of old Twitter. So X advertising is at $1.8 billion now. It's down $100 million from last year." — Sam Parr 00:25:01

Grok (XAI)

Elon's AI model/product. Has 100 million users vs. ChatGPT's 1 billion. Significantly behind competitors but provided the strategic rationale for building Colossus.

"Grok has 100 million users, whereas ChatGPT has a billion. So 10 times smaller." — Sam Parr 00:27:04

Ontario Teachers' Pension Fund

Canadian pension fund managing assets for ~300,000 teachers. Invested in SpaceX in 2019 and stands to make ~$12B from the IPO — approximately $33,000 per teacher.

"Shout out to the Ontario Canada Teachers Pension Fund, which in 2019 decided, you know what? We're going to sling an investment into SpaceX... and they're going to make $12 billion this year." — Sam Parr 00:43:06

TerraFab

Elon's planned chip manufacturing facility, described as potentially the largest in the world. Strategic rationale: getting ahead of the chip bottleneck constraining AI development.

"TerraFab is basically his way of getting ahead of what he believes is like the chip bottleneck in the world, which is there needs to be more chip production, ideally in the United States." — Sam Parr 00:04:55

Mercury

Business and personal banking product. Sponsored segment — Sam Parr uses it across ~7–8 businesses.

"I use Mercury for all of my businesses. I think I have like maybe seven or eight businesses. We use Mercury as our business banking across all of them." — Sam Parr 00:51:00

Future Ventures

Steve Jurvetson's current venture fund after leaving DFJ. Jurvetson was one of the earliest investors in both Tesla and SpaceX, and never sold a share.

"He started DFJ... now he's got his own fund, Future Ventures... he was one of the earliest investors at Tesla, earliest investors at SpaceX. Never sold a single share." — Sam Parr 00:55:41

4. People Identified

Elon Musk

Founder and CEO of SpaceX, XAI; owner of X. Owns 42% of SpaceX with 85% voting control after 20+ years of capital-intensive fundraising. Will become the world's first trillionaire upon IPO. Has a compensation package that would pay him ~$750B in hypothetical grants if SpaceX reaches $7.5T market cap and establishes a 1 million-person Mars colony.

"He owns 42% of the company... After 20 years, and you would think rockets are like the most capital intensive business that he's had to raise for. For him to still own 42%... he has 85% of the voting control. Unbelievable." — Sam Parr 00:30:43

Antonio Gracias

Founder of Valor Equity Partners. SpaceX's second-largest individual shareholder at ~7%. Became Elon's manufacturing and operations advisor early in both Tesla and SpaceX, helping identify production bottlenecks. Once personally loaned Elon money during Tesla's darkest days with no equity ask.

"He became his study buddy for ramping up production. And Antonio was always there to help... There's a story about how when Tesla was like on the brink, he loaned Elon like a million dollars or something like that. Personally, not for any equity, just like you need it. Here you go. And so he's going to make like $90 billion in this IPO." — Sam Parr 00:50:08

Luke Nosek

Co-founder of Gigafund, formerly of Founders Fund and PayPal founding team. Made what appeared to be an unsophisticated but devastatingly correct decision: concentrate everything on Elon's companies and do nothing else. Listed as a major stakeholder in the SpaceX S-1.

"When Luke was at Founders Fund, he just kind of like sat back and realized that the optimal way for him to invest was simply to just back every Elon company. And that's it... it also seemed not sophisticated, which I think is the underrated part of this." — Sam Parr 00:52:48

Steve Jurvetson

Founder of Future Ventures, formerly DFJ. True futurist; one of the earliest investors in both Tesla and SpaceX. Never sold a share. Documented early rocket launches and Bitcoin mining rigs on Flickr with detailed written commentary — described as a valuable, underappreciated knowledge base.

"If you go, I think a fun, interesting read is go Google Steve Jurvetson and go look at his Flickr... he was one of the earliest investors at Tesla, earliest investors at SpaceX. Never sold a single share, which is the other remarkable thing is to hold through all the ups." — Sam Parr 00:55:56

Sam Bankman-Fried

Convicted founder of FTX. Highlighted as a cautionary counterexample — his SpaceX, Anthropic, Cursor, Solana, and Robinhood investments (made using misappropriated customer funds) would be worth $114B today, which would have made him look like the greatest investor of his generation.

"His overall portfolio out of what he his slush fund that he was inappropriately using illegally using customer funds to go invest. It'd be 114 billion dollars today. He'd be seen as one of the greatest investors of all time." — Sam Parr 00:38:50

Chamath Palihapitiya

Former Facebook executive turned investor. Cited as a cautionary tale about not holding your best asset — had he simply held his Facebook stock, he would have outperformed everything he's done since, including his SPAC era and Warriors ownership.

"Chamath, after Facebook, started his own fund and then started doing SPACs and bought the Warriors. And he's done this and that and this and that. All he had to do was just hold his Facebook stock from when he was an employee, he would have done better than everything he's done since then." — Sam Parr 00:54:13

Bill Gates

Microsoft founder. Cited alongside Chamath as a canonical example of activity being the enemy of returns — Gates selling Microsoft stock is described as one of the most expensive decisions in investment history.

"Famously, Bill Gates selling his Microsoft stock. Bill Gates would be the wealthiest man in the world had he just not sold his Microsoft stock." — Sam Parr 00:54:06

5. Operating Insights

The "Charlie Munger Prompt" as a Due Diligence Framework for Any Business

Sam ran his SpaceX analysis through Claude using a specific framing: "Explain it as if you're Charlie Munger trying to understand the business." The prompts that emerged — Is it in my circle of competence? How do I lose money here? Show me the incentives. What's the moat? — are a replicable template any operator or investor can use with AI to stress-test a business.

"I told it to explain it, to be like I'm Charlie Munger trying to understand the business. And it's like, okay, Charlie Munger would ask, is it in my circle of competence? How do I lose money here? Show me the incentives. What's the moat? Price is what you pay. The value is what you get." — Sam Parr 00:29:08

Stagger Major Announcements Around Your S-1 Filing to Sustain Momentum and Control Narrative

SpaceX amended their S-1 quietly — after initial filing — to include Google's $1B/month Colossus deal, effectively creating a second wave of investor excitement and press coverage. Whether intentional or not, it's a playbook for founders going public: file, then add a landmark customer announcement mid-process.

"I have to think the 9D chess move there was that they knew and they staggered the announcement... it was almost like a quiet amending of the document that says, hey, Google just signed up to spend a billion dollars a month with us to become a customer." — Sam Parr 00:28:19

Replace Absolutes With Probabilistic Language to Make Better Decisions

Shaan's decade-long lesson from Silicon Valley: stop saying "that will never work" and replace it with "it almost never works, but sometimes it does." This small linguistic shift dramatically changes how you evaluate opportunities and avoids the systematic blind spots that come from binary thinking.

"I used to say things like that will never happen or that cannot happen or that's impossible. And then I started changing this to where I see something that thinks unlikely, I almost always try to say it almost never works, but sometimes it does. Or like it mostly ends this way instead of saying that will fail and it will end this way." — Shaan Puri 00:59:48

6. Overlooked Insights

Cursor's Acquisition at $60B Is Possibly the Most Revealing Strategic Signal in the Entire S-1

The hosts mention Cursor's acquisition option almost as a footnote, but this deserves far more attention. Cursor is at $3–4B in revenue and SpaceX is paying $60B — roughly a 15-20x revenue multiple — not because Cursor is cheap, but because this is the fastest path to compute-hungry AI coding users that Grok desperately needs. The subtext: Elon knows Grok can't win on model quality alone. He needs distribution. And Cursor, which faces existential pressure from Codex and Claude Code (Anthropic's coding product), needed compute and model access it couldn't build alone. This is essentially a mutual survival merger masked as an acquisition — and the $60B price tag signals just how desperate the AI coding arms race has become.

"They bought Cursor for $60 billion or have an option to buy Cursor for $60 billion. And Cursor, I think, is at $3 or $4 billion in revenue. And so like, Cursor needed to figure out a way to have compute and their own models and a way to survive against Codex and Claude Code. And so it's this sort of game theory where the bottom players like in Survivor will form an alliance in order to flip the game on its head." — Sam Parr 00:30:06

The SpaceX Blue-Collar Employee Millionaire Story Is a Preview of Demographic Wealth Reshuffling That Will Hit Real Estate Markets

Mentioned only briefly, but over 4,000 SpaceX employees — including cafeteria workers — are becoming millionaires from stock options in this IPO. Sam's anecdote of a San Francisco home selling in five minutes for all-cash, without the buyer even returning to view it, is an early data point. As Anthropic, OpenAI, and SpaceX all go public within a short window, a wave of newly liquid wealth will hit coastal housing markets simultaneously — and potentially spread inward as recipients invest, relocate, or buy second homes. This is an investable real estate thesis, not just a fun fact.

"Over 4,000 new millionaires, including some cafeteria workers whose compensation packages included employee stock options." — Shaan Puri 00:37:13

"My sister just sold her house... He walked in for five minutes without his wife and made an all cash offer on the house at like what they were going to list it at. And that was it. It was done. He literally never came back even and bought the house. I've never seen anything like it." — Sam Parr 00:38:02