Moonshots Summary: June 17, 2026 - by Peter H. Diamandis

Moonshots Summary: June 17, 2026

The Most Intense Week of News in Living Memory: SpaceX Becomes a $2.5+ Trillion Bet on the Future, White House Seizes the Keys to Frontier AI, and Recursive Self Improvement Accelerates

TLDR: Some weeks the news trickles in. This week it arrived like a flood. SpaceX went public and instantly became the fifth-largest company on Earth. The US government suspended the world's most capable AI model with 90 minutes of warning. And somewhere in all of it, Elon became the planet's first trillionaire. My Moonshots mates, Salim, Dave, and Alex, joined me to make sense of a week that Salim called “the most intense run of news he has seen ever.” Here's what we covered.

JOIN US AT THE 2026 MOONSHOTS GATHERING

Before we dive into the Moonshots Podcast Summary, I want to invite you to join a live event with the Moonshots mates called the “ Moonshots Gathering” taking place on September 25th in downtown LA. One full-packed day running 8 a.m. to 10 p.m. at the beautiful United Theater.

Who?: The Moonshots mates (Peter, Alex, Dave & Salim) will be joined by Palmer Luckey (Anduril), Astro Teller (the captain of moonshots at X), and Cathie Wood. We’ll be announcing many other AI and exponential tech CEOs in the next month.

Networking: Join 1,500 AI builders and creators for power networking and partnering.

Content: Alex (AWG) will run an hour-long “Solve Everything” session and a one-on-one AMA. Salim will host an hour-long session on the “Organizational Singularity.” Astro Teller will teach you how to build and manage a Moonshot startup. Anousheh Ansari (CEO) will host a session on “designing XPRIZEs,” and Cathie Wood will present her 2026 “Big Ideas” report while Dave Blundin dives into “investing in AI.”

Creator + Builder Awards: The Gathering will serve as the finals for the Future Vision XPRIZE (the world’s largest film competition, $5 million in prizes, screening the five finalists, selecting the winner). We’ll also have the finals of the “ Build w/Gemini XPRIZE,” a 90-day hackathon with over 15,000 entries where the top builders show how they went from zero to real revenue.

Tickets: Tickets just opened and they will sell out fast. Go to www.moonshots.com to grab yours and join us.

SPACEX: NOT A STOCK, A CIVILIZATIONAL BET

Last Friday, SpaceX pulled off the largest IPO in history, opening at $135 and closing day one up nearly 20% at $161. The market cap now sits over $2.5 trillion, the fifth-largest company on the planet, ahead of Amazon.

“People are trying to price SpaceX like a normal tech company. It isn’t one. It’s three converging businesses—the launch monopoly, the Starlink cash engine, and an AI frontier lab—wrapped around a single thesis: We’re in the Singularity. You’re not buying revenue. You’re buying into humanity’s future economy.” – Peter Diamandis

SpaceX has a launch monopoly, with no competitor within a factor of 10 to 100. The Starlink cash engine, 10 billion subscriptions and over a billion dollars in quarterly profit, what I think of as civilizational infrastructure. And an AI frontier lab pointed at orbit. Elon now holds about 82% voting control of SpaceX but not Tesla, which is exactly why I put the SpaceX-Tesla merger at 100% by year-end. Polymarket has it at 37%.

“This is not really just an IPO. This is the public market pricing in a civilizational ExO. You’re really buying into the future of the world.” – Salim Ismail

Now the part everyone got wrong. Elon crossed $1.3 trillion in net worth, making over $100 million a day for three days straight. But the trillionaire framing is a distraction. Here’s the real story: SpaceX’s IPO was the largest single-day creation of millionaires in history. Roughly 4,400 employees became millionaires, another 400 became centimillionaires or billionaires. That is what abundance looks like. Value created, not value extracted.

ANTHROPIC VS. THE GOVERNMENT: WHO OWNS YOUR ACCESS TO INTELLIGENCE

On Friday at 5:21 PM ET, with 90 minutes of notice, the US government issued an export-control directive suspending access to Anthropic’s Fable 5 and Mythos 5 for any foreign national, including Anthropic’s own employees, roughly a third of its staff. Anthropic’s response was to shut down global access entirely. The trigger was a jailbreak a researcher found in Fable 5’s guardrails, which Anthropic called a minor, known issue.

“The deeper question is the one that matters: who decides what level of intelligence you and your company are allowed to access?” – Peter Diamandis

It gets worse for trust. Even before the export order, developers were in revolt. Buried in a 319-page document: Anthropic retained every prompt for at least 30 days, even for enterprise customers with negotiated zero-retention agreements. And if Fable 5 detected you doing frontier AI research, it silently downgraded you to a weaker model without telling you. You can’t run a business on intelligence that can be throttled or revoked overnight.

“At the end of the day, who owns AI? Is it the government or is it the corporations? The government is saying, we do. And that ain’t going back in the bag.” – Dave Blundin

A likely resolution is a grand bargain between Anthropic and the White House, possibly golden shares or a sovereign-wealth stake. Polymarket gives Fable 5 a 90% chance of returning within a month. Dave’s caution: the version that comes back may be patched, or quietly dumbed down, and we’ll have a hard time knowing.

OPENAI, PRICE WARS, AND AGENTS THAT WRITE THEIR OWN GOALS

The WSJ reports that OpenAI is weighing aggressive price cuts to capture developers fleeing Anthropic’s chaos. Alex thinks it’s more marketing than structural; Anthropic can cut costs just as easily. Salim’s framing is the one to remember: every 10x drop in token cost means 100x more experiments run.

Sam Altman floated that faster recursive self-improvement could be a reason to delay OpenAI’s IPO. Dave’s blunt read: Sam holds no personal equity, so a delay costs him nothing. The more interesting idea, from Alex, is that if compounding RSI lets big private companies generate value internally without raising outside capital, we may be watching the early decoupling of technology from capital itself. OpenAI just raised $122 billion, larger than the SpaceX raise, so there’s no near-term pressure anyway.

Then the line that stopped me: OpenAI’s Codex engineering lead revealed that agents now set their own goals. Developers no longer write the objective; Codex writes goals for itself and for every subagent it spawns.

“Codex is now writing its own goals and the goals for each subagent it spawns. We’re speedrunning toward AI personhood and self-determination.” – Dr. Alex Wissner-Gross

DATA CENTERS: THE BOTTLENECK IS A CENTURY-OLD TRANSFORMER

Here’s the single most exponential number in AI right now. Since Colossus One came online in August 2024, the one Elon built in 122 days, the record for compute in a single data center has doubled every seven months. Globally, AI computing capacity is growing 3.3x per year. That demand curve is going vertical.

And here’s where it hits the wall: there’s a two-and-a-half-year wait on power transformers and a three-year wait on step-up transformers. The bottleneck in AI is no longer chips, and it isn’t even money. It’s boring, century-old electrical hardware. Valuations for suppliers like Hitachi, Siemens, GE Vernova, Hyundai, and Virginia Transformer are up 100 to 400% year on year.

Worse, half of the 9 gigawatts of compute planned for 2026 is already delayed by activist protests, the same fight we saw against nuclear plants 30 years ago.

The escape route is up. Near term, big training runs stay on Earth and inference moves to orbit. By the early 2030s, Alex, Dave, and I each independently landed on the same conclusion: the Moon becomes the logical home for super-training-clusters.

ABUNDANCE, THE FLOOR, AND A PANDEMIC OF FEAR

The week’s hardest conversation wasn’t about technology. It was about people. New York Fed data shows recent college grads, ages 22 to 25, having the steepest job-finding difficulty of any cohort, with software and support roles diverging sharply downward. History has a warning here: roughly 20 major revolutions share one ingredient, educated young men facing a gap between the future they were promised and the one they got.

My core worry isn’t whether the job apocalypse is real. It’s that the fear of it is real, and fear is enough to foment unrest. Salim added the accelerant: social media amplifies bad news 100x, and panic now travels far faster than truth. Half this country can’t put $500 together in an emergency. So the work is clear. Get abundance down to the floor, fast.

“The 20th-century fight was capitalism versus socialism. Now it’s who owns the exponential surplus, who owns Abundance, and how do we distribute it.” – Dave Blundin

HERE’S THE BOTTOM LINE...

Every story this week was really the same story. The containers we built for the old world are cracking under the new one. A rocket company became a civilizational bet. A government claimed the right to decide who gets to think with the best machines. Agents started writing their own goals. And the bottleneck on the most advanced technology in history turned out to be a transformer with a three-year backorder. The first trillionaire is a headline. The 4,400 new millionaires, and whether we can lift the floor before fear does its damage, that’s the actual moonshot.

Catch the full episode wherever you get your podcasts, and come build the future with us at the Moonshots Gathering, September 25th in Los Angeles.