Summary

SpaceX’s proposed IPO would raise $75 billion at a $1.75 trillion valuation, making it the world’s eighth-largest public company at the price Jason cited. The expected June offering could establish the external mark Chamath believes is needed before an eventual Tesla combination, which he rates “99.999%” likely.
A Tesla-SpaceX merger would consolidate an increasingly unified industrial stack spanning xAI, robotics, chips, factories, launch, and satellite communications. Chamath argues that public marks for both companies would reduce governance disputes over Elon Musk’s time while letting cross-disciplinary learning “compound and compound and compound.”
Friedberg sees the moon—not asteroid mining—as the likelier next industrial frontier because robots could mine, refine, and manufacture there before shipping finished materials to Earth. His speculative design uses 500 square meters of solar panels to power a four-kilometer mass driver launching one package every 10–15 minutes, with 15 centimeters of moon rock serving as its re-entry shield.
The broader space trade is infrastructure built on top of SpaceX: orbital last-mile delivery, debris collection, power generation, stations, mining, and communications. Chamath’s analogy is that Falcon delivers containers to “Long Beach,” leaving room for “the FedEx of space” and every other terrestrial service to be rebuilt above Earth.
The IPO window may close quickly once SpaceX, OpenAI, and Anthropic absorb investors’ finite capital. Chamath’s pricing paradox is stark: “If AGI is real, the durability of most companies is slim to none”; if it is not, AI companies’ multibillion-dollar fundraising needs deeper scrutiny—“both cannot be right.”
Public price discovery could be painful because many natural buyers already own these companies privately while long-held shares create pent-up selling pressure. Chamath still calls OpenAI and Anthropic deserving trillion-dollar businesses, but expects their arrival to pull money from listed software and force tech multiples toward non-tech levels: “It’s going to be nasty.”
The Iran war’s immediate costs sit beside larger shocks to energy, sovereign capital, fertilizer, and food. Jason cited $70 billion spent by day 34 and 50,000 US troops deployed; Friedberg warned that 35% of global nitrogen fertilizer normally passes through the Strait of Hormuz, while urea had jumped from roughly $350 to above $700 per ton.
Friedberg gives Bitcoin’s ecosystem five to seven years, by his estimate, to complete a quantum-resistance migration before it becomes an obvious “honeypot.” He said theoretical improvements had cut one cited factoring workload from 28 million quantum operations to 500,000; post-quantum standards exist, but wallets, nodes, transaction flows, and internet-wide communications still face a heavy implementation lift.

Deep dive

1. SpaceX is attempting an IPO at nation-state scale

Jason’s opening figures put SpaceX’s confidential April 1 filing at a $1.75 trillion target, with a proposed $75 billion raise and expected June debut. At that valuation it would rank eighth globally, just behind TSMC and Saudi Aramco; Tesla, valued around $1.37 trillion, ranked tenth.
The reported operating picture was Starlink supplying 50%–80% of revenue and approaching $20 billion annually, while launch generated about $5 billion in 2024. Jason separately cited Reuters estimates of $15 billion–$16 billion in 2025 total revenue and $8 billion in profit.
SpaceX had also acquired xAI for $250 billion, in Jason’s account, including X/Twitter and the model company. A theoretical Tesla-SpaceX combination would be worth roughly $3.1 trillion at the cited prices, placing it fourth globally and inspiring his imagined ticker: “ELON.”

2. Public marks could clear the path to a Tesla combination

Chamath put the merger probability at “99.999%.” His mechanism begins with the IPO creating a “validated external mark-to-market valuation” for SpaceX, matching the daily market price already available for Tesla and reducing legal and governance friction around combining the companies.
The legal backdrop is what his advisers called a tax on American capital-market creativity: shareholder suits form classes, pursue directors-and-officers insurance, and leave attorneys taking 40%–50%. He characterized the successful challenge by a Tesla holder with ten shares against Musk’s compensation package as “a shakedown.”
With both entities priced publicly, Chamath expects fewer arguments about Musk allocating time among companies. Nobody objects to Zuckerberg, Nadella, Pichai, or Huang overseeing different internal projects; Musk’s robots, xAI systems, Terafab, and manufacturing capabilities increasingly serve the same underlying stack.

3. Musk’s companies are converging into one industrial system

Chamath’s industrial thesis is that Elon’s robots are used inside SpaceX, Terafab serves Tesla, and xAI is used across both. Combining the teams—and potentially knowledge from Boring Company and Neuralink—would make that cross-disciplinary learning “compound and compound and compound.”
He considers Musk’s factory knowledge a genuine competitive advantage, while acknowledging specialists such as Foxconn may know more about particular aspects. Engineers already move between Musk companies, and bringing those teams together would compound their cross-disciplinary learning.
Friedberg’s 20-year-forward view is that Tesla may ultimately be the more interesting company: it started as an electric-car company. Jason added that it became an autonomous-car company, and that its autonomy competency led to the robotics revolution. Those robots provide the missing labor force for industrializing environments where humans remain difficult to sustain.

4. The moon could become a robotic factory floor

Friedberg called the Artemis 2 launch an important milestone: four astronauts would orbit Earth, travel around the moon, and return, ahead of a landing in about two years. Beyond symbolism and competition with China, he sees access to the moon as the gateway to humanity’s “next industrial frontier.”
The moon contains aluminum, silicon, palladium, platinum, gold, and most industrially relevant materials, Friedberg argued. Its main deficits are carbon, nitrogen, hydrogen, and oxygen—the gases its weak gravity cannot retain—making local mining and processing more plausible than supporting large human workforces.
One-sixth Earth gravity and no atmosphere make outbound freight radically cheaper. Friedberg described a magnetic mass driver accelerating packages at 100 G, sending processed goods toward an exact point on Earth, then using a parachute for landing; unlike asteroid mining, this permits continuous mining, refining, and manufacturing around fixed infrastructure.
His speculative calculation: 500 square meters of solar panels could run a four-kilometer mass driver, dispatching one load every 10–15 minutes. A 15-centimeter layer of moon rock could absorb re-entry heat before the package lands at an industrial site. “I’m speculating in a bunch of different ways,” he emphasized.

5. SpaceX is laying rails for an economy it will not own alone

Jason compared SpaceX with the railroads that opened the American West: transportation infrastructure enables value creation far beyond the carrier. Friedberg said Starlink adds another layer by creating a parallel, extraterrestrial communications network—effectively a “backup internet” that could persist alongside terrestrial fiber and copper networks.
Friedberg offered Vast Space as the enabling example: founder Jed McCaleb committed hundreds of millions of dollars, prepaid SpaceX for launch capacity, and began developing modular station components. Lower launch costs make privately built stations for customers such as Google or Amazon at least conceivable.
Chamath distinguished bulk launch from orbital logistics. A Falcon ride may drop cargo near 500–550 kilometers, but reaching the correct orbital plane still requires the equivalent of FedEx after a container reaches Long Beach; debris collection, specialized solar power, and last-mile delivery remain open categories.
His broader call was expansive: “There’ll be a FedEx of space. There’ll be a Maersk of space,” plus waste management and mining equivalents. This is “the beginning of the beginning of the beginning,” with autonomy making the frontier productive rather than merely explorable.

6. The first IPOs may eat the market’s entire appetite

Jason’s potential 2026 slate included SpaceX, PropTech, Anthropic, OpenAI, Databricks, Stripe, Cerebras, Canva, and Discord. Polymarket assigned respective IPO probabilities of 94% for SpaceX, 41% for Anthropic, 38% for OpenAI, and 32% for Databricks.
Chamath compared investors with Thanksgiving diners whose eyes initially exceed their stomachs: SpaceX should be consumed first, the second issuer may perform well, and each later company faces fuller plates. “It’s good to be first. It’s pretty decent to be second,” but late issuers should fortify their balance sheets quickly.
The principal event risk, in his view, is not Iran but the unresolved meaning of AGI. If it exists, most corporate moats lack durability; if it does not, the AI labs’ enormous capital requirements deserve much harder inspection. “History will sort out which one is right, but both cannot be right.”
That makes mature “high asset, low obsolescence” businesses unusually competitive with speculative issuance. Chamath said investors could buy hundreds of millions of dollars in annual cash flow for two-to-five-times multiples, raising the question: why move outward on the risk curve for something priced at 200 times revenue?

7. Liquidity could reprice even extraordinary AI companies

Friedberg rejected the assumption that an IPO automatically drives value higher. Long-time holders finally receive liquidity, creating large selling pressure; meanwhile, many institutions capable of buying at scale already acquired shares privately. “I don’t know who the big buyers are that everyone’s expecting are going to show up.”
Jason cited OpenAI holders seeking buyers for $600 million of shares at an $850 billion valuation, while Anthropic—then valued at $300 billion—was attracting secondary bids around $600 billion. Against OpenAI’s cited $24 billion of revenue, $850 billion represented roughly 35 times sales.
Chamath nevertheless called both companies unprecedented and deserving of trillion-dollar status: “I’ve never seen a business like this.” His uncertainty remained explicit—he did not know their profitability, terminal values, or what IPO investors would pay—but that uncertainty is precisely why both should raise quickly after SpaceX.
His source-of-funds answer is rotation: SpaceX, OpenAI, and Anthropic will first attack technology incumbents, eroding the moats supporting tech’s valuation premium. Investors may still buy the first five or six years of listed software earnings, but not year 15; tech multiples should converge toward non-tech multiples.

8. Iran could produce a delayed capital shock

Jason’s day-34 tally included 13 dead and more than 200 injured US service members, 3,500 Iranian deaths—including 1,600 civilians and over 200 children—and 1,200 deaths in Lebanon. He cited 50,000 US troops deployed, $70 billion spent at roughly $2 billion daily, and a Pentagon request for another $200 billion.
Prediction markets gave a ceasefire 25% odds by the end of April and 47% by May, while pricing a ground invasion at 63% by April and 71% by December. Jason’s concern was that an escalation or recession could derail the IPO pipeline and allow force-majeure arguments against existing funding commitments.
Friedberg’s subtler risk is that Qatar, Saudi, UAE, and Omani capital becomes less available to capital-intensive US technology. Sovereign and family-office money often reaches companies indirectly through fund commitments, bank loans, and SoftBank-like intermediaries, so reduced allocations may take a cycle before markets discover that “the reliable go-tos are gone.”
Chamath disagreed on market primacy: “I think the markets will shake that off.” For him, Iran is secondary to the “sword of Damocles” hanging over equities—whether AI is real, and therefore what every incumbent company is actually worth.

9. Energy independence is becoming strategic optionality

Trump’s address framed the US as independent of Middle Eastern oil while still supporting allies. Chamath’s takeaway was categorical: countries unable to control their energy infrastructure remain exposed to conflicts they cannot shape, while US independence preserves “complete and total optionality.”
Europe had weakened that optionality through nuclear exits, dependence on expensive imports, and solar-policy reversals. Chamath saw an emerging correction in Italy’s and Spain’s investment tax credits and Germany’s nuclear restart, calling European energy independence a critical positive consequence of the crisis.
The regional financing relationship complicates the picture. Chamath called the UAE, Saudi Arabia, Qatar, and Kuwait America’s essential future banking partners; they need security and the ability to monetize hydrocarbons before rapidly deployable solar and slower nuclear expansion reduce long-run demand. He therefore sees them as more motivated than America to put “boots on the ground.”

10. Fertilizer turns Hormuz into a food-security chokepoint

Friedberg broke fertilizer into nitrogen, phosphorus, and potassium, with nitrogen comprising about 60%–65% of global fertilizer and driving crop productivity. Ammonia production requires natural gas and atmospheric nitrogen compressed to roughly 200 times normal pressure, which concentrates production around major gas-processing regions.
Roughly 35% of global nitrogen fertilizer passes through the Strait of Hormuz, he said, while China—the swing producer at about 15%—halted exports shortly after the war began. Urea consequently rose from about $350 per ton to more than $700.
Corn requires around 200 pounds of urea per acre, making the crop unprofitable at the cited input price while China’s halted corn purchases suppress output prices. Two-thirds of US farmers had secured spring fertilizer; many of the remaining third were switching to soybeans, with another vulnerable planting cycle approaching in the fall.
There is little spare capacity: damaged Qatari infrastructure could require three-to-five years to repair, a new plant about seven years to build, and existing facilities already run continuously. With global calorie inventories below 30 days, Friedberg said the crisis “could be even more severe” than the post-Ukraine shock that pushed 400 million people into malnutrition.

11. The war’s political and nuclear endgame remains unresolved

Jason predicted Trump would “100%” wrap up the conflict quickly, citing net approval of negative 17, gasoline above $4, an expected return to three-handle inflation, and Polymarket odds of 86% for Democrats taking the House and 51% for the Senate. Without a turn, he foresaw investigations, impeachment attempts, and political paralysis.
Jason’s epistemic caution—echoed by Sacks—was that outsiders lack the intelligence required to know whether Israel manipulated Trump or whether military action was indispensable. Easy claims in either direction omit “the texture of the relationships and the details and what really went on.”
Sacks called Trump the most consistently anti-war modern president and said the president has shown unusual restraint and a high bar for entering conflict, while emphasizing that there is much outsiders do not know. Chamath argued no additional state should obtain nuclear weapons, while carefully distinguishing Iran’s extremist fringes from the “overwhelming majority” of peaceful Sunnis and Shias.
The unresolved shift is from Israel’s “mowing the lawn”—periodically removing nuclear progress—to regime change. Chamath’s preferred longer-run equilibrium is bipolar US-China stewardship; China’s Hormuz-driven energy pain and a summit delayed six weeks into mid-May might create leverage for operating rules and de-escalation.

12. Quantum makes cryptographic migration urgent

Friedberg said the horizon for a functional quantum chip may have moved from 25–30 years to perhaps five-to-seven years. Bitcoin’s leaders therefore need an organized answer on quantum resistance because crypto is the most visible “honeypot” if a non-state actor can defeat SHA-256, ECDSA, and related schemes.
Friedberg’s attack sequence answers the objection that banks would also fail: first drain the obvious crypto stores, then reveal the breach, let prices collapse, and use the captured wealth to buy assets. “Yes, you’re right, all this other stuff goes kaput,” but the attacker’s incentives determine the order.
Friedberg traced the software side from Peter Shor’s 1994 factoring algorithm to a paper he believed was published around 2023 by Oded Regev. One cited workload fell from 28 million quantum operations to 500,000, showing that algorithms are becoming less demanding before industrial-scale hardware arrives.
Post-quantum encryption research already spans more than 20 years, but deployment across wallets, processing nodes, transaction flows, internet communications, and institutions remains a heavy lift—and potentially a major business. Jason’s symmetry was blunt: if AGI is real, software is mispriced; if quantum is real, many crypto projects are mispriced. “Pick one.”

How Hims & Hers Reached a $4.3BN Market Cap on $2.3BN of Revenue | Andrew Dudum Episode 178 - April 3, 2026