Mira Murati is not a natural for the conference stage. As the CTO of OpenAI, he was present but rarely seen in the company. As the CEO of his company, Think Machines Lab, he became even more difficult to find. So when he sat down with Bloomberg in San Francisco on Thursday — his first media appearance in nearly 18 months — it was worth paying attention to, even if he was careful not to say too much.
Time makes sense. Thinking Machines has spent the better part of a year and a half working hard in the background: raising money, hiring researchers, and shipping one product, Tinker, an API for fine-tuning open source AI models.
Meanwhile, companies competing for the same talent, customers, and headlines have become ubiquitous. OpenAI, where Murati spent six years as CTO, is constantly in the news cycle. Anthropic momentum is all anyone can talk about right now. And xAI, Elon Musk’s AI, has been folded into SpaceX ahead of what is expected to be its largest public offering, generating its own gravitas for attention and investment. In that environment, sitting upside down has diminishing returns; sometimes, you have to make noise just to remind the market that you are there.
Murati used Bloomberg’s appearance to do just that and not much more. He previewed what Thinking Machines calls “interaction models,” which he described as a completely different type of AI interface. Instead of the turn-based, fast-response dynamics that define most AI products today, he told interviewer Emily Chang, the company’s models are designed to process continuous streams of audio, text, and video at 200-millisecond intervals. The idea is that they can learn the texture of people’s communication — interruptions, mid-thought fixations, even pauses to think — in something close to real-time. But Murati was careful to frame it as a first step, not a finished product, and refused to put a specific release date on anything.
He also answered questions about the episode that first thrust him into the public eye: the tumultuous week in November 2023 when OpenAI’s board fired Sam Altman and became interim CEO. Inside OpenAI it was called “blip.” Murati said he felt clear about his decisions at all times – that protecting the mission and the team was the way that made the decisions seem obvious as the situation seemed to be falling apart from the outside. He said the company would have “contributed” had it not been for his involvement in that five-day period and its recent results. But he admitted that clarity of purpose is not the same thing as clarity of results. In retrospect, he said, he would have pushed harder for more information, a better transition plan, and more transparency. He didn’t say, at least not directly, whether he thought things went well.
Asked if he still trusted his former boss, he shrugged off the question, directing the conversation to a larger concern he returned to several times: the concentration of key decisions in too few hands — not just at OpenAI but across the industry. His concern, he said, is less about the character of any individual leader (although he admits it is important) and more about the absence of structural checks. Good people make bad calls. Well-intentioned organizations are drifting. Too much attention to beauty and too little to governance, he suggested.
Chang also politely pressed him on the departure of several high-profile researchers from Think Tank in recent months, a topic Murati has largely avoided publicly and downplayed Thursday. First, he said, building a frontier AI lab from scratch compresses years of typical organizational flexibility into months. He also acknowledged that compensation – nine-figure packages have become the norm in the battle for AI talent – is taking people’s minds, but suggested it’s not the whole story. In some audience laughter, he said about his own competitive feelings, “When I wake up in the morning, I don’t think about how I will kill my competitor.”
Naturally, Chang asked what’s next for AI more broadly, including people who AI companies once said would be empowered by AI but who have recently grown fearful of talk of mass deportation, not to mention a future where AI is used to create chemical weapons.
Mrati, who was born in Albania and speaks with a slight Eastern European accent, is measured in his response. He has pushed back against the formulation of an inevitable dystopia or an inevitable utopia, arguing that there is no predetermined outcome and that the present moment will determine how things turn out. Nevertheless, he said – not for the first time during the discussion – that if people take their hands off quickly, the future will look very different, and not better.
If you shop through links in our articles, we may earn a small commission. This does not affect our editorial independence.
