2 hours ago
23
Mira Murati, the former CTO of OpenAI who quietly departed the company and largely vanished from public view, reappeared on June 4 at Bloomberg Tech 2026 in San Francisco. It was her first major media appearance in 18 months.
She wasn’t there to reminisce. Murati used the stage to lay out the roadmap for Thinking Machines Lab, the AI startup she co-founded and now leads as CEO, a company that has already secured a record-breaking $2 billion seed round and is reportedly in discussions that could push its valuation to $50 billion.
From OpenAI exile to $50 billion ambitions
Thinking Machines Lab launched in February 2025 with a $10 billion valuation, making its $2 billion seed round the largest in history at the time.
The company has since locked in heavyweight backing. Nvidia announced a multiyear chip supply agreement with Thinking Machines in March 2026, committing to provide its upcoming Vera Rubin accelerators. Google Cloud is also listed among the startup’s significant partners.
Now, funding discussions are reportedly targeting a $50 billion valuation. That would represent a 5x jump from its initial valuation in roughly 16 months.
Murati’s pitch centers on what the company calls “interaction models.” Think of them as AI systems designed not for one-off prompts and responses, but for continuous, real-time dialogue with humans. The target latency is 200 milliseconds, which is roughly the speed of a human blink.
Tinker and the open-source play
The startup’s first product is Tinker, an API built for fine-tuning open-source AI models. It’s a deliberate strategic choice that positions Thinking Machines differently from OpenAI’s increasingly closed ecosystem.
Murati has been vocal about maintaining human oversight in AI development, a stance that carries extra weight given her front-row seat to OpenAI’s own internal debates about safety, commercialization, and control.
The talent problem nobody wants to talk about
Thinking Machines has already experienced notable talent departures. Several founding-team members returned to OpenAI in early 2026. Others left for Meta, which has been aggressively hiring AI researchers as it builds out its own model ecosystem.
Murati didn’t dwell on the exits during her Bloomberg appearance, focusing instead on the company’s technical direction and her conviction that AI development needs to prioritize real-time human-AI collaboration over raw model capability.
What this means for investors watching the AI sector
The Nvidia partnership is arguably the most significant signal here. Vera Rubin accelerators are Nvidia’s next-generation chips, and a multiyear supply agreement suggests Nvidia sees Thinking Machines as a long-term player worth prioritizing.
The outcome of those funding discussions will be worth watching closely. If Thinking Machines closes at or near $50 billion, it validates Murati’s thesis that there’s room for a major independent AI lab outside the OpenAI-Google-Meta oligopoly.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·