Mira Murati is not a natural on the conference stage. Although she was present as OpenAI’s CTO, she was rarely the public face of the company. As the CEO of her own company, Thinking Machines Lab, she’s even harder to find. So when she spoke with Bloomberg in San Francisco on Thursday, her first major media appearance in nearly 18 months, it was worth watching, even if she was careful not to say too much.
Timing matters. Thinking Machines has spent the better part of a year and a half operating largely in the background. We raised capital, hired researchers, and shipped a product called Tinker, an API for fine-tuning open source AI models.
Meanwhile, companies competing for the same talent, customers, and headlines are becoming more and more ubiquitous. OpenAI, where Murati spent six years as CTO, is constantly in the news cycle. All anyone can talk about right now is Anthropic’s momentum. And Elon Musk’s AI venture, xAI, is being folded into SpaceX ahead of an expected major initial public offering, creating its own gravitational pull for attention and investment. In such an environment, keeping your head down will reduce your profits. At some point, you have to make some noise to remind the market of your existence.
Murati used his Bloomberg appearance to do just that, but he didn’t do much more than that. She previewed what Thinking Machines calls an “interaction model.” This is a fundamentally different kind of AI interface, she explained. He told interviewer Emily Chang that rather than the turn-based prompt-and-response dynamics that define most AI products today, the company’s models are designed to process continuous streams of audio, text, and video at 200-millisecond intervals. The idea is to be able to capture the texture of human communication (interruptions, mid-thought corrections, pauses to think, etc.) in near real-time. However, Murati was cautious about positioning this as a first step rather than a finished product, and declined to give a specific release date.
She also answered questions about the episode that first brought her to the public eye: the chaotic week in November 2023 when OpenAI’s board fired Sam Altman and she became interim CEO. Inside OpenAI, this is now referred to as a “blip.” Murati said he had a clear feeling about his decision at each moment. Even if the situation looked like it was falling apart from the outside, protecting the mission and the team was the turning point that made the choice seem clear. She said the company would have “fallen apart” if she hadn’t been involved in those strange five days and its immediate aftermath. But she acknowledged that clarity of intent is not the same as clarity of outcome. Looking back, she said she would have asked for more information, a better transition plan and more transparency. What she didn’t say, at least not directly, was whether she thought things went well.
Asked if she still trusts her former boss, she sidesteps the question and directs the conversation to a big concern she keeps coming back to: the concentration of big decisions in a few hands, not just at OpenAI but across the industry. Her concerns, she said, were not about the character of individual leaders (though she acknowledged that was important), but rather about the lack of structural checks. Good people make bad calls. Well-intentioned organizations drift away. Too much attention, she suggested, was paid to virtue and too much to governance.
Chan also politely pressed her about the departure of several high-profile researchers from Thinking Machines in recent months, a topic Murati has largely avoided in public and downplayed on Thursday. First, she said, building a Frontier AI Lab from scratch compresses years of normal organizational change into months. She also acknowledged that compensation (the nine-figure packages that have become the standard currency in the AI talent war) captures people’s imaginations, but suggested that’s not usually the be-all and end-all. “When I wake up in the morning, I don’t think about how I’m going to beat my competition,” she said of her own competitive instincts, drawing laughter from some in the audience.
Unsurprisingly, Mr. Chan asked the public, including humans, who have previously said that AI companies would be empowered by AI, but who have recently begun to be frightened by talk of mass unemployment, what will happen next for AI, not to mention a future where AI makes chemical weapons.
The reaction from Murati, who was born in Albania and speaks with a slight Eastern European accent, was immeasurable. She pushed back on the paradigm of inevitable dystopia or inevitable utopia, arguing that neither outcome is predetermined, and that the moment we are in now is the moment that will determine the direction things will take. Still, she said, not for the first time in an interview, that if humans took their hands off the wheel too soon, the future would look very different, not better.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
