Mira Murati's Measured Return to the Spotlight

Mira Murati, a figure not typically inclined towards public speaking, recently made a notable return to the spotlight. As the former CTO of OpenAI, she maintained a low public profile, a characteristic that became even more pronounced as CEO of her own venture, Thinking Machines Lab. Her appearance with Bloomberg in San Francisco on Thursday marked her first significant public engagement in approximately 18 months, drawing considerable attention despite her measured commentary.

This re-emergence is strategically timed. Thinking Machines Lab has operated largely in stealth mode for the past year and a half, dedicating its efforts to securing funding, recruiting top researchers, and developing its initial product, Tinker—an API designed for fine-tuning open-source AI models.

During this period, the landscape of AI companies vying for talent, customers, and media coverage has become increasingly competitive. OpenAI, where Murati served for six years as CTO, consistently dominates news cycles. Anthropic's rapid ascent is a frequent topic of discussion. Furthermore, Elon Musk's AI initiative, xAI, has been integrated into SpaceX in anticipation of a substantial public offering, thereby commanding significant attention and investment. In such a dynamic environment, maintaining a low profile eventually yields diminishing returns; at some point, a company must assert its presence to remain relevant in the market.

Murati leveraged her Bloomberg interview to achieve precisely that, offering a glimpse into what Thinking Machines is developing: "interaction models." She described these as a fundamentally different type of AI interface, moving beyond the conventional turn-based, prompt-and-response dynamic prevalent in most current AI products. Murati explained to interviewer Emily Chang that these models are engineered to continuously process streams of audio, text, and video in rapid 200-millisecond intervals. The objective is to enable them to discern the subtle nuances of human communication—including interruptions, mid-thought corrections, and even reflective pauses—in near real-time. This aligns with the lab's core philosophy that advanced AI capabilities stem from deeper human collaboration. However, Murati carefully presented this as an initial step rather than a finished product, refraining from providing any specific release dates.

She also addressed the tumultuous events of November 2023, referred to internally at OpenAI as "the blip," when the company's board dismissed Sam Altman and she assumed the role of interim CEO. Murati asserted that her decisions during that chaotic week were guided by a clear imperative to protect the company's mission and team, making her choices seem obvious even as the situation appeared to unravel externally. She contended that OpenAI would have "imploded" without her involvement during that critical five-day period and its immediate aftermath. Nevertheless, she conceded that clarity of intent does not equate to foresight regarding consequences, stating in retrospect that she would have advocated more strongly for additional information, a more robust transition plan, and greater transparency. She stopped short, however, of directly commenting on whether she believes the situation ultimately resolved favorably.

When pressed on whether she still trusts her former boss, Murati skillfully redirected the conversation towards a broader concern she emphasized repeatedly: the excessive concentration of critical decision-making power in too few hands, a problem she sees not only at OpenAI but across the entire AI industry. Her primary apprehension, she clarified, is less about the character of individual leaders (though she acknowledged its importance) and more about the absence of robust structural checks. She argued that even well-intentioned individuals can make poor judgments, and organizations with good intentions can lose their way. Murati suggested that too much emphasis has been placed on virtue and insufficient attention on effective governance.

Emily Chang also gently probed Murati about the recent departures of several high-profile researchers from Thinking Machines, a topic Murati had largely avoided publicly and downplayed during the interview. Murati explained that establishing a cutting-edge AI lab from scratch inherently condenses years of typical organizational volatility into mere months. She further acknowledged that the allure of nine-figure compensation packages—now standard in the intense competition for AI talent—is powerful, but suggested it rarely tells the complete story. To the amusement of the audience, she remarked on her own competitive drive, stating, "When I wake up in the morning, I am not thinking about how to kill the competitor."

Naturally, Chang concluded by asking about the broader future of AI, specifically concerning humanity. She highlighted the shift from earlier promises of human empowerment through AI to contemporary anxieties about widespread job displacement and the potential misuse of AI, such as in the creation of chemical weapons.

Murati, who hails from Albania and speaks with a subtle Eastern European accent, offered a thoughtful response. She rejected the binary framing of an inevitable dystopia or utopia, asserting that neither outcome is predetermined and that the current era is pivotal in shaping the future trajectory of AI. However, she reiterated—a point made multiple times during the interview—that if humanity relinquishes control too prematurely, the future will indeed look significantly different, and not for the better.