A recent Crunchbase analysis shows early-stage AI unicorns are being created at record speed. But many are “unicorns at birth,” valued on talent rather than products. TheMarketAI Take: as seen with Thinking Machines, when that talent shifts, so does the value. In a market where capital is flooding AI, speed is rising — but so is fragility.

In traditional startups, early departures are often survivable because value accrues to product, customers, or revenue. In unicorn-at-birth AI startups, valuation is frequently anchored to people. When those people leave — especially co-founders — the asset itself is diminished.

Thinking Machines Labs is reportedly seeking a $50B valuation, just months after raising at a $12B mark. But with co-founder departures and a first product—Tinker—that feels incremental rather than transformative, the jump raises questions. As we’ve covered before, the company still looks more like a research lab than a revenue engine. The widening gap between ambition and output makes this valuation hard to justify.

Meta is cutting around 600 roles from its Superintelligence Labs, impacting FAIR and other AI units in what appears to be a major strategic reset. The move follows a $27B financing deal with Blue Owl Capital and comes just days after Meta hired Andrew Tulloch, co-founder of Thinking Machines Labs. The layoffs suggest Meta is consolidating its AI ambitions — possibly shutting down entire R&D teams to refocus on next-generation foundation models.