Blue Origin Raises $10B: What Is It Worth?

Blue Origin is raising outside capital for the first time ever — a $10 billion round at an expected valuation of $130 billion. Jeff Bezos, who has been the sole financier of the company for its entire 25-year existence, is doubling down personally with a $2 billion direct investment. Venture firm Coatue, which Bezos's family office already backs through its Innovative Strategies Fund, is taking an even larger $4 billion allocation. The remaining capital opens the door for true outside investors to get their first slice of the rocket company. The scoop came from Andrew Ross Sorkin at the New York Times, and the timing — right in the middle of Sun Valley — makes you wonder how much was whispered over cocktails at the lodge.

Breaking down the Blue Origin $10B raise and who is participating 00:45 Breaking down the Blue Origin $10B raise and who is participating Watch at 00:45 →

This is technically the first time Blue Origin has taken outside money, which makes it historically unusual. The company has burned an estimated $27 billion to date — roughly $1 billion per year over 25 years — and is expected to burn another $5 billion in 2025 alone. At $10 billion raised, this looks like a clean 12-to-18 month runway, not a panic raise. Bezos isn't selling. He's loading up. And when the founder is marking himself up alongside outside investors, it's hard to complain too loudly about the price.

Is Blue Origin Undervalued Compared to SpaceX?

The Elon bulls are already making noise that SpaceX looks cheap by comparison. The argument goes like this: Blue Origin's revenue is almost certainly a fraction of SpaceX's, the multiple is astronomically high on any cash flow basis, and yet the market is pricing in capability and trajectory rather than current financials. And on that front, Blue Origin has made genuine strides. It became just the second company in the world — after SpaceX — to bring a rocket to orbit, land it successfully, and prove reusability. It beat China, which has been working to replicate SpaceX's model for years. That's not nothing.

For context on where private market valuations have gone: when Gavin Baker at Fidelity wanted to value Uber at $14 billion, everyone thought he was out of his mind. That round closed at $17 billion and felt unprecedented at the time. Today, $60 billion looks almost quaint. Several private companies are valued in the hundreds of billions, and the framework for pricing deep-tech, long-horizon bets has fundamentally shifted. Other comps investors are likely looking at: Rocket Lab is sitting just under $50 billion, and AST SpaceMobile — which has barely gotten anything into operational orbit — is valued around $28 to $30 billion. A fully operational, reusability-proven rocket company at $130 billion starts to look reasonable in that context.

What Is GPT-5.6 Soul and Why Does It Matter?

OpenAI is launching GPT-5.6 Soul — along with Terara and Luna — publicly this Thursday, with preview access expanding globally. Early testers have been sitting on their opinions for weeks, and the embargo lifting has produced a wave of takes. The consensus from people who've actually used it: Soul and Fable represent genuine jumps over every previous model, and they've opened a large gap with the next-best AI options available today. Researcher Ethan put it plainly — if you're doing any work where intelligence quality actually matters, these two models are now your only real choices.

Early reactions to GPT-5.6 Soul from testers who had access 08:20 Early reactions to GPT-5.6 Soul from testers who had access Watch at 08:20 →

What makes Soul distinct from Fable, even though both are top-tier? One early user described it like this: Fable is light-speed across the galaxy, while Soul is a Porsche on the Nürburgring — more engaging, more precise, built for interaction rather than raw output. Soul is also reportedly world-leading in computer use, which sent early testers into a productivity spiral they apparently couldn't sustain once access was briefly pulled. The AI fragmentation story is real: we're entering an era where the right tool for the job matters, and two flagship models from the same company can be meaningfully different in feel and function.

OpenAI's New Real-Time Voice Mode Is Also Live

Separate from the Soul launch, OpenAI has rolled out GP Live — a new generation of voice models built for natural, real-time human-AI conversation. It features full-duplex architecture, meaning the model can listen and speak simultaneously and jump into conversation appropriately, rather than waiting for you to finish. It can also route complex background tasks to smarter models while maintaining the conversation in the foreground. The vision being floated: you're talking to a model, you ask it to run a deep research report, and it goes off and does that while you keep talking about other things. When the report's done, it flags you. That kind of parallel, continuous interaction is genuinely new territory for how people work with AI tools day-to-day.

Discussion on why Apple scrapped the cheaper Vision Pro display 04:15 Discussion on why Apple scrapped the cheaper Vision Pro display Watch at 04:15 →

Why Did Apple Kill the Cheaper Vision Pro?

Apple has scrapped plans for a lower-cost Apple Vision Pro display and is winding down its work with Samsung Display, with the project expected to be formally closed by September. For the small but passionate community of Vision Pro daily users, this lands hard. The original thesis was simple: the display technology Apple used was extraordinary — sourced by essentially jumping years ahead in Samsung's development roadmap and accepting brutal low manufacturing yields at a premium price — and that technology would naturally commoditize over time, enabling a cheaper version that could actually achieve scale.

That commoditization apparently hasn't solved the fundamental problem. Apple seems to be signaling that the near-term market for people sitting down to watch a full movie in a headset just isn't large enough to build a consumer product around. And the argument has some uncomfortable logic to it: the nature of how people actually consume content today — short-form, interactive, social, scrollable — doesn't map well to a fully immersive, eyes-closed, sit-still-for-two-hours experience. The number of taps and interactions per minute in a doom scroll session dwarfs those in a film viewing session. Apple may be bearish on that use case, even if they're still very much building toward everyday AR glasses. The Vision Pro may have been a technology showcase more than a product roadmap commitment.

Why Did the Getty Images Shutterstock Deal Collapse?

Getty Images has officially terminated its planned $3.7 billion merger with Shutterstock after the UK's Competition and Markets Authority placed conditions on the deal that Getty wasn't willing to accept. Specifically, the CMA required Getty to sell off Shutterstock's editorial business to preserve competition for UK media outlets — and Getty's board voted not to proceed on those terms. The DOJ in the US cleared the deal in April with no such conditions, making this a UK-specific block.

The irony here is hard to miss. Both companies exist in a category being disrupted in real time by AI image generation. Shutterstock's market cap is around $329 million. Getty's is around $340 million. The $3.7 billion headline number was based on January 2025 share prices — both stocks have been crushed since. The CMA's logic — that consolidation would reduce competition and raise prices for UK media outlets — seems to ignore the bigger competitive threat: essentially unlimited AI-generated images from Meta, OpenAI, and others that are increasingly indistinguishable from photographs. Blocking this merger looks less like protecting consumers and more like kicking two struggling companies on their way down. The UK has form here — they also blocked Meta from acquiring Giphy in 2022 to prevent a supposed monopoly in GIFs.

Why Are Urban Explorers Obsessed With IBM's Abandoned Campus?

On the lighter end of the news cycle: the Wall Street Journal ran a front-page story on the former IBM campus in Somers, New York, which has become a magnet for urban explorers, graffiti artists, and apparently groups of teenagers running from state troopers in the woods. The site — vacant for years — fits the classic profile for the urban exploration community: a large, decaying structure with a storied history, now slowly being reclaimed. The phenomenon has been turbocharged by Instagram and TikTok, where atmospheric walkthrough videos inspire the next wave of curiosity seekers. Similar warnings have been issued around closed malls in New Jersey and abandoned amusement parks in Florida. If you're a scrappy startup looking for an atmospheric — if legally complicated — headquarters, Watson's old home is apparently about an hour from Manhattan.