Mira Murati Steps Back Into Spotlight With Careful Approach

Mira Murati has never been a natural fit for the conference circuit. As OpenAI’s CTO, she remained largely behind the scenes, rarely serving as the company’s public face. Now leading her own venture, Thinking Machines Lab, she has been even more elusive. Her recent sit-down with Bloomberg in San Francisco on Thursday — marking her first major media appearance in about 18 months — demanded attention, even as she carefully avoided over-sharing.

The timing aligns with her strategy. For roughly eighteen months, Thinking Machines has operated quietly in the background: securing funding, recruiting top researchers, and launching a single product, Tinker, an API designed for fine-tuning open-source AI models.

Meanwhile, competitors vying for the same talent, clients, and media coverage have become increasingly dominant. OpenAI, where Murati served as CTO for six years, remains a constant fixture in news cycles. Anthropic’s rapid rise dominates current industry conversations. Meanwhile, Elon Musk’s xAI has merged with SpaceX ahead of a anticipated massive public offering, drawing significant attention and investment. In such a crowded landscape, staying silent yields diminishing returns; eventually, companies must make noise to remind the market of their presence.

Murati used this Bloomberg appearance to achieve exactly that, with minimal additional disclosure. She introduced what Thinking Machines terms “interaction models,” describing them as a fundamentally new type of AI interface. Moving beyond the standard turn-based, prompt-and-response structure of current AI products, she explained to interviewer Emily Chang that these models process continuous streams of audio, text, and video in 200-millisecond intervals. The goal is to capture the nuances of human communication — interruptions, mid-sentence corrections, and thoughtful pauses — in near real-time. However, Murati framed this as an initial step rather than a final product, declining to commit to a specific release date.

She also addressed the incident that first thrust her into the spotlight: the chaotic week in November 2023 when OpenAI’s board dismissed Sam Altman, leading to her role as interim CEO. Internally, this period was dubbed “the blip.” Murati stated she felt confident in her decisions at the time, emphasizing that protecting the company’s mission and team was the consistent priority that made her choices seem clear, even as the situation appeared to unravel externally. She claimed the company would have “imploded” without her leadership during that five-day period and its immediate aftermath. Yet, she acknowledged that clear intent does not guarantee clear outcomes. Looking back, she admitted she would have advocated for more information, a smoother transition plan, and greater transparency. Notably, she did not directly state whether she believes the final outcome was positive.

When asked if she still trusts her former boss, Murati deflected, redirecting the conversation to a broader concern she revisited multiple times: the concentration of critical decision-making power in too few hands — not only at OpenAI but across the entire industry. She expressed worry less about the character of individual leaders (though she acknowledged this matters) and more about the lack of structural safeguards. She argued that even good people can make poor decisions and that well-intentioned organizations can drift. Too much focus has been placed on individual virtue, and too little on robust governance.

Chang also gently pressed her on the recent departures of several high-profile researchers from Thinking Machines, a topic Murati has largely avoided publicly and downplayed during the interview. She explained that building a frontier AI lab from scratch compresses years of typical organizational volatility into mere months. She also acknowledged that compensation — the nine-figure packages now standard in the competition for AI talent — attracts attention, but suggested it is rarely the sole factor. To some audience laughter, she remarked on her own competitive nature, stating, “When I wake up in the morning, I am not thinking about how to kill the competitor.”

Naturally, Chang inquired about the future of AI broadly, including its impact on humans. While AI companies initially claimed technology would empower workers, recent fears of mass job displacement have grown, alongside concerns about AI being used to create chemical weapons.

Murati, born in Albania and speaking with a slight Eastern European accent, responded with measured caution. She rejected the narrative of inevitable dystopia or utopia, arguing that neither outcome is predetermined. Instead, she emphasized that the current period is critical in determining which path humanity will take. She reiterated — not for the first time during the interview — that if humans remove themselves from the equation too soon, the future will look vastly different, and likely worse.