Bought or Buried

Nebius paid $643 million for Eigen AI this month. Eigen AI has twenty people.

Do the division. That's about $32 million per human, and the product they bought, an inference-optimization stack, mostly came along for the ride. What Nebius actually wrote the check for was twenty engineers who know how to make models run cheaper, plus a best-paper award. That's not the price of a company. That's the price of a roster.

I keep running into this exact shape lately, and once you see it you can't unsee it. The startup ending you were sold, the one where you build something that stands on its own two feet and somebody pays a premium for the business, has quietly split into two. Your team gets bought. Or your company gets buried. The middle, the stretch where you actually grow into a real company, has narrowed into a goat path.

The acqui-hire stopped being the sad ending.

For a long time "soft landing" was the polite phrase for failure. You couldn't raise, you couldn't sell the product, so you sold the people and everyone clapped and pretended it was a plan. Now it's the plan.

Google paid around $2.4 billion for Windsurf, and what it wanted was the founders, not the code editor. Microsoft handed Inflection $650 million and walked off with Mustafa Suleyman and his research team, leaving the chatbot behind like a returned rental car. HP bought Humane's assets for about $116 million and killed the Pin almost the same week. Meta scooped up Limitless and folded the crew into its wearables group while the pendant itself got a quiet funeral. Tally it up: between early 2024 and now, the four biggest tech companies spent north of $20 billion hiring away founding teams without technically acquiring a single company.

The product is the wrapper, the team is the asset, and the acquisition is just the costume the hiring wears to work.

The other pile is bigger, and a lot quieter.

For every team that gets bought, there is a much longer line getting buried. More than 220 former unicorns are now worth less than half their peak. Calendly. Glossier. Names you have typed into a browser. Companies that last raised in 2021 are marked down about 68% on average, the 2022 class down 52%, and roughly half the unicorn herd hasn't raised in three years. In venture, "hasn't raised in three years" is a gentle way of saying nobody will set a new price.

SaaS is taking the worst of it. Seventy-five software companies on the fallen-unicorn list, double the count of fintechs. The logic is cold. The partner who used to write a follow-on check to a SaaS company growing 40% a year now writes that same check to an AI-native company growing 200%. AI swallowed 81% of venture dollars last quarter. A former DoorDash engineering lead put it flat: every workflow-driven SaaS company is either disrupted or dead inside a decade.

Buried doesn't mean bankrupt. It means stranded. Still alive, still shipping features, still answering support tickets, just permanently un-fundable, idling on a cap table nobody will touch.

Power users figured this out early: speaking a prompt gives you 10x more context in half the time. You include the edge cases, the examples, the tone you want — because talking is fast enough that you don't skip them.

Wispr Flow captures everything you say and turns it into clean, structured text for any AI tool. Speak messy. Get polished input. Paste into ChatGPT, Claude, Cursor, or wherever you work.

89% of messages sent with zero edits. 4x faster than typing. Works system-wide on Mac, Windows, and iPhone.

Start flowing free

What's on the table is your team, not your company.

Here's the thing. Once you accept that the priced asset is the people plus one sharp capability, the whole year stops looking random.

The valuable unit is small now. Twenty people with a genuine edge, an inference optimizer or a retrieval pipeline that actually returns the right document, are worth more to an acquirer than a 150-person company with a broad product and a moat that turned out to be a speed bump. The $32 million a head Nebius paid isn't an equity valuation. It's a comp package with an acquisition stapled around it for tax reasons and a press release.

From the cap-table seat it's uglier than the headline. The soft landing is a lifeboat with room for the crew and nobody else. The founders and the key engineers walk away with retention packages and a shiny new badge. The investors who funded the original dream usually get cents on the dollar, if that. And the people who didn't make the keep-list? A Slack-pocalypse and a severance doc. "We got acquired" and "we won" stopped being the same sentence somewhere in the last two years, and most of the celebration posts haven't gotten the memo.

So if you're building right now, quit pretending the middle road is wide. Pick your pile on purpose. If you're playing to get bought, stay small, go embarrassingly deep on the one capability some giant will eventually need, and don't burn a year on the company scaffolding nobody is paying you for. If you're playing to stay independent, you'd better be AI-native enough to ride the 200% lane, because the 40% lane is where unicorns go to get stranded.

And one honest thing, because you've read this far. I watched a team I genuinely admired get "bought" last year. They announced it like a trophy, and for the founders, it was. Six months on, the product they'd bled four years into was a sunset banner and a redirect, and the people I knew were three layers down in someone else's org chart, shipping someone else's roadmap. They're fine. They're paid well. But the thing they made is just gone, and I don't think it's what they were chasing the night they started. So build the thing you'd still be glad you built if the only buyer who ever shows up wants you, and not it.

— SWEdonym