Once upon a time in Silicon Valley, a band of ex–OpenAI stars led by former OpenAI CTO Mira Murati launched Thinking Machines Lab. With a $2 billion seed round and a $12 billion valuation — a number so high it made other startups blush — investors celebrated the future of AI. All they had to show for it? A lone fine-tuning API in private beta and a pitch deck full of slides that said “AI for everyone” but didn’t actually explain how or when.
Valuation without revenue: It’s like pricing a unicorn based on how shiny its horn might be. And in the case of Thinking Machines? The horn hasn’t even left the stable.
Management Shake-Up or AI-themed Reality TV?
If the product roadmap felt vague, the HR strategy was cartoonishly explicit.
First, the company’s CTO, Barret Zoph, was fired after internal disputes and a workplace romance saga that sounded more like early-2000s office sitcom plotlines than the scripted brilliance of an AI powerhouse. Then a wave of departures followed: co-founders and top talent trotted back to OpenAI, while others scattered to Meta and elsewhere, like ants fleeing a sinking picnic table.
It’s ironic: a company founded on “thinking machines” couldn’t quite figure out how to keep its human ones.
AI Failure as a Business Model?
Critics might argue the startup had things backwards. Rather than build things people want, they collected cash people feared missing out on — a valuation bubble of epic proportions that made “unicorn” feel like an understatement. Some whispers out of industry circles even suggested Thinking Machines eyes a $50+ billion valuation without any real product traction to justify the math.
In a world where revenue still matters (yes, even in AI), Thinking Machines has leaned into hype over hard engineering:
- Raise billions 🟡
- Ship virtually nothing 🟡
- Lose co-founders back to competitors 🔴
- Attend industry panels and talk about “vision” 🟡
Does this sound like a business model — or the plot of the next Apple TV+ tech satire?
Talent Exodus: The Ultimate Backfire
Here’s the punchline: some of the people who helped found Thinking Machines have now returned to OpenAI — the very company they left in search of something new. That’s the equivalent of bold explorers returning to the village because they realised the trail map was upside-down.
Meanwhile, the product pipeline is still essentially in “you’ll hear from us soon” mode, and investors — who once threw hush money at the company like confetti — may be quietly asking whether any of it was grounded in actual deliverables.
So What Did We Learn?
If there’s a moral here, it’s this:
In the AI gold rush, hype can get you a valuation, but execution gets you revenue. And perhaps, eventually, talent retention.
Thinking Machines Lab may still pivot, ship, and succeed — but right now, it reads less like the next big AI innovator and more like a cautionary tale on how not to build a company. For now, the real “thinking machines” seem to be the investors who are quietly evaluating whether this rocket ship has an actual engine.
