The Builders Are Back

Tech's most proven founders are launching their next acts, all at once, and the market is pricing something new: proof of scale. One interpretation is that investors are placing unusually high value on founders with demonstrated execution histories.

‍

Deep Dive

Last week Travis Kalanick raised $1.7 billion for Atoms, in a round reportedly led by Andreessen Horowitz. Ben Horowitz joined the board, fifteen years after a16z came up short on Uber's Series B. His announcement essay was titled "Travis Is Back." Even Uber, the company Kalanick built and famously left, came back as an investor. The debt came from Goldman Sachs, JPMorgan, Wells Fargo, Barclays and Bank of America.

Kalanick isn't the whole story. He's just the tip of the iceberg.

Second acts are a Valley tradition. Steve Jobs coming back to Apple is the most famous version. Even Adam Neumann, after everything at WeWork, raised $350 million from a16z for his next act. Capital has always had a soft spot for people who have done it before.

What's new is the scale, the speed, and that it's all happening at once. Four of the most well-known founders came back within nine months of each other. Calling these second acts undersells it. Bezos built Blue Origin alongside Amazon. Luckey exited Oculus and founded Anduril before defense tech was cool. Kalanick had two startups before Uber. Musk is on company seven or eight. For this group, starting over is routine.

Four founders, nine months, roughly $50 billion

Starting with the richest man in the Milky Way. Musk founded xAI in March 2023 and closed a $20 billion Series E in January at a reported $230 billion valuation. In February, SpaceX absorbed xAI in an all-stock deal that valued xAI at roughly $250 billion and the combined company at $1.25 trillion, the largest private merger on record, according to CNBC. Few founders have converted a track record into capital at comparable speed.

Four years atop the Forbes list, four years of fun, and Bezos is back at it. Prometheus launched in November with $6.2 billion, his first operating role since Amazon. The mission is AI for the physical economy. By June it had reportedly closed a $12 billion Series B at a $41 billion valuation, with JPMorgan, BlackRock, Goldman Sachs, DST Global and Arch Venture among the backers. Roughly $18 billion in seven months, aimed at $16 trillion of global manufacturing.

Kalanick's version took eight years, most of it quiet. He built CloudKitchens, running food operations across dozens of cities, while everyone assumed he was done. That work now sits under Atoms, spanning food, mining, and autonomous ground transport. Uber digitized how people move. Atoms wants to digitize how everything else gets made, moved, and stored.

Then there's Palmer Luckey, who sold Oculus to Facebook for $2 billion at 21 and built Anduril into a defense company reportedly last marked at $61 billion in May, with reports this month of talks near $100 billion. All before turning 35. Erebor is his third act, a national bank aimed at the innovation economy, founded with Trevor Capozza, Jacob Hirshman, Aaron Pelz and Owen Rapaport, and backed by Joe Lonsdale and Peter Thiel. It raised $350 million at $4.35 billion in December, won the first new national bank charter of this administration, and reportedly grew deposits from $1.1 billion in late March to $4.05 billion at the end of Q2.

None of these billionaires needs to work another day. None of these founders appears financially constrained.They're going back in anyway, into the hardest, most capital-intensive problems they can find. When founders with records like these raise, the private market pays attention.

The allocators are crossing over too. Chamath Palihapitiya spent two decades allocating capital. In June he took the CEO seat at 8090 Labs, a company he founded in 2024, alongside a $135 million Series A led by Salesforce Ventures. His stated reason: "the technological ground is moving so ferociously underneath all of us." He waited fifteen years for a moment like this to get back in the trenches.

The existing companies are raising too. This month Bezos opened Blue Origin to outside money for the first time since founding it in 2000, reportedly raising $10 billion at a $130 billion valuation and writing $2 billion of the check himself. Anduril is reportedly in talks near $100 billion, against the $61 billion mark it set in May. The Boring Company is reportedly negotiating $4 billion at $20 billion, up from roughly $5.7 billion in 2022. Those last two were reported a day apart. Same founders, more vehicles, all of them raising at once.

Reported private-company financials and secondary-market indications may be unaudited, incomplete, non-standard, or based on limited transaction activity. They should not be relied upon as fair value, executable pricing, or a basis for any investment decision.

The money had to go somewhere. One explanation is that institutional capital has increasingly concentrated in a relatively small number of companies capable of absorbing very large investments.

There's a name for companies like these. Kyle Harrison, a general partner at Contrary, calls them capital absorbers, the "hungry, hungry hippos of raising money." They complete what he calls the Unholy Trinity of venture capital: allocators, agglomerators, and absorbers. The allocators are the pensions and sovereign wealth funds with hundreds of billions to park. The agglomerators are the mega funds on the other side, raising tens of billions they can't deploy in $20 million checks.

The allocators aren't chasing venture returns. A big pension spends 4 to 5 percent of its assets a year, so a 7 to 8 percent return keeps the fund whole. Smaller funds may put up better numbers, but a $60 billion allocator can't use them. Writing $30 million checks means conducting due diligence and backing more managers than the market produces in a year. The mega funds solve the problem. Park $250 million, hit the yield target, move on. And once that money fills the mega funds, it has to go to work. PitchBook put AI at about 65 percent of US venture deal value through the first three quarters of 2025. The capital needed to be deployed somewhere. The open question was who could credibly absorb it.

The market is pricing the person

Prometheus raised its first $6.2 billion before it had a product. Erebor was reportedly marked at $4.35 billion before it opened. Amazon, for perspective, raised $54 million at a reportedly $438 million valuation in its 1997 IPO. When the check runs that far ahead of the business, what's being underwritten is the founder. Venture used to price proof of concept, then proof of traction, then product-market fit. This market appears to be pricing proof of scale.

Almost nobody has a track record that compares to these four founders. Bezos turned a bookstore into a two-trillion-dollar company. Musk built several of the most valuable companies in history. Kalanick took Uber from an idea to a global network. Luckey shipped a $2 billion exit at 21 and has Anduril at a reported $61 billion mark. Zero to one is common. One to a hundred billion is not. The four are not equivalent, either. The companies differ in business mix, capital intensity, regulatory exposure and disclosure quality, so what they share is a pattern rather than a comparable set of businesses.

And none of it requires a spotless record. Kalanick lost Uber in a boardroom. Neumann's first act ended on magazine covers. Horowitz's line is that a16z invests in strength, not lack of weakness. Nobody is buying a clean resume. They're buying the spike.

Reported private-company financials and secondary-market indications may be unaudited, incomplete, non-standard, or based on limited transaction activity. They should not be relied upon as fair value, executable pricing, or a basis for any investment decision.

The market got bigger, so the first check did too

Why now? Dan Ives, the Wedbush analyst, calls AI the fourth industrial revolution, and he means it literally: a decades-long rebuild of industries, labor, and capital. He has said more may be spent on it in the next two years than in the last ten combined. To the bubble callers, his line is that "this is a 1996 moment, not 1999."

Three of the four founders are focused on the physical economy. Prometheus at manufacturing. Atoms at food, mining, and heavy transport. xAI at compute, satellites, and autonomy. Erebor is building the bank for the next era of builders. When the addressable market jumps from software budgets to global industrial output, a multi-billion dollar first round stops looking reckless. Whether it is rational is a different question, and the results will answer it.

The speed matches. xAI went from a reported $24 billion to $250 billion in under eighteen months. Prometheus went from zero to a reported $41 billion in seven. Entry prices at those levels assume a great deal about what comes next.

The capital needed somewhere to go. The opportunity required founders who have already built at this scale. And the builders needed a prize worth coming back for. All three showed up at the same time.

None of it happened on an exchange

Almost none of this activity touched an exchange, and none of it was available to a non-accredited investor. xAI's $230 billion mark was set privately, then folded into SpaceX at a reported $1.25 trillion, also privately. SpaceX listed in June and raised roughly $86 billion including the underwriters' option, the largest IPO on record. Public investors got their first look at a company already valued above two trillion dollars on its opening day. The entire ride from zero to a trillion happened in private. The IPO used to be where public investors bought growth. Increasingly, that growth is priced before the listing.

Reported private-company financials and secondary-market indications may be unaudited, incomplete, non-standard, or based on limited transaction activity. They should not be relied upon as fair value, executable pricing, or a basis for any investment decision.

The jury is out

Is this the obvious strategy? The jury is still out. No P/E, no EBITDA multiple gets you to $41 billion pre-product. These are venture bets, in the biggest markets anyone has addressed, backing founders with exceptional track records, at entry prices that already assume greatness.

The work ahead is staggering. Prometheus has to grow into one of the most valuable companies on earth, from a product that doesn't exist yet. Atoms has to automate mines and win in several other trillion-dollar markets. Erebor, reportedly in talks near $8 billion, has to show that a bank can produce venture-scale returns. That is a structurally different proposition from the other three: bank returns are constrained by capital requirements and regulatory oversight in ways software and defense companies are not.

Roughly $50 billion of reported fresh capital went into these companies in nine months. Whether that price was right won't be visible for years, and it won't be visible on an exchange.

‍