Mira Murati spent 18 months largely out of sight. Then on June 5 she sat down with Bloomberg’s Emily Chang in San Francisco. The former OpenAI chief technology officer spoke as CEO of Thinking Machines Lab. She previewed technology built for real-time conversation. She reflected on the chaos that briefly made her interim boss at her old employer. And she issued a pointed caution about how few people hold decisive sway over artificial intelligence’s direction.
The appearance marked a deliberate step back into public view. Murati had left OpenAI in September 2024. In the time since, her new company raised $2 billion. It secured access to a full gigawatt of Nvidia’s next-generation Vera Rubin systems. It shipped one product. It watched several key researchers walk out the door. The industry she re-entered looks fiercer, richer and more concentrated than the one she left.
Thinking Machines Lab launched in February 2025. Murati assembled a founding team drawn from OpenAI, Meta and other labs. Names included John Schulman, Barret Zoph, Lilian Weng, Andrew Tulloch and Luke Metz. Early funding valued the startup at $12 billion after a round led by Andreessen Horowitz. Reuters reported the details. Investors bet on Murati’s track record. She had played a central role in releasing ChatGPT, DALL-E and Codex.
Yet progress proved slower than the valuation suggested. The company’s first offering, Tinker, arrived in October 2025. It is an API that lets researchers and developers fine-tune open-source models. Forbes described it as useful for experimentation but not a breakout hit. For months that remained the only public product. Talent began to drift. In January 2026, Zoph, Metz and Sam Schoenholz returned to OpenAI. Five other early employees joined Meta, drawn by compensation packages that reached nine figures. Fortune covered the exits.
Murati addressed the departures directly. She told Chang that building a frontier lab from scratch compresses years of normal upheaval into a few intense months. Compensation matters, she acknowledged. But she pushed back on the idea that rivalry alone drives her. When I wake up in the morning, I am not thinking about how to kill the competitor. The line drew laughs. It also underscored a tension. OpenAI dominates headlines. Anthropic has raised tens of billions. Elon Musk’s xAI continues to expand. Quiet conviction faces a short runway.
The new technology Murati chose to highlight aims to change the texture of human-machine exchange. She described what Thinking Machines calls interaction models. These systems handle continuous streams of audio, video and text. They update every 200 milliseconds. The first example, TML-Interaction-Small, returns responses in 0.40 seconds. That speed matches natural conversation. The models support full duplex interaction. Users can interrupt. The system can adjust midstream. It picks up on pauses, corrections and tone. TechCrunch detailed the approach.
Murati framed the work as a first step. She gave no firm release schedule. She positioned it beside Tinker as part of a broader effort to make AI feel more like a collaborator and less like a question-and-answer box. The thesis is clear. Powerful systems require closer human partnership, not detachment. If people step away from the controls too early, outcomes may suffer. Murati repeated the point. Neither dystopian nor utopian futures are locked in. The choices made right now will shape which path prevails.
Chang asked about the episode still known inside OpenAI as the blip. In November 2023 the board fired Sam Altman. Murati became interim CEO for five turbulent days. She testified later that she feared the organization faced catastrophic risk. During the Bloomberg conversation she stood by her actions. She said she felt clear at each moment about protecting the mission and the team. Without her involvement, she added, the company would have imploded. Yet she admitted room for improvement. She would have pressed for more information, a smoother transition and greater transparency.
When asked whether she still trusts Altman, Murati sidestepped. Instead she widened the lens. Her worry centers less on any single personality and more on the concentration of consequential decisions in too few hands. This pattern exists across the industry, she said. Good people still make bad calls. Well-intentioned organizations can drift. Structural checks remain scarce. Bloomberg captured the exchange. The critique lands at a moment when a handful of labs control the most advanced models, the largest compute allocations and the biggest capital inflows.
Recent coverage shows the pressure mounting. Puck examined the gap between Thinking Machines’ early promise and its limited public output after 15 months. The publication noted that despite stratospheric funding the company had little to show beyond one API. Built In reviewed the lab’s origins and its emphasis on flexible, adaptable systems. Its profile highlighted the original 30-person team drawn from elite labs. On X, discussion mixed skepticism about timing with recognition that the post-OpenAI alumni wave now forms its own category of startups.
Murati’s return arrives as the sector wrestles with speed versus caution. Compute deals grow ever larger. Talent raids intensify. Valuations balloon before revenue appears. Her call for better governance does not propose slowing progress. It argues for mechanisms that prevent drift. Her product direction tries to embed collaboration directly into architecture. Interaction models treat conversation as a live loop rather than a series of discrete prompts. The bet is that systems designed this way will prove more reliable partners.
Whether that vision can hold against competitors racing to scale remains unproven. Thinking Machines has one commercial offering. Its research team has thinned at the top. Yet Murati sounded measured. Heads-down execution still matters, she suggested. At the same time, she concluded that silence carries its own risks. The industry has moved fast without her. Now she aims to help set its next direction. With real-time models in testing and a governance critique on record, the former OpenAI leader has drawn the contours of her second act.
Analysts will watch the next release closely. Investors will measure whether interaction models deliver measurable advantages in latency, coherence and user retention. Regulators and observers will listen for whether her warnings translate into concrete proposals for oversight. Murati has stepped forward. The conversation she rejoined will test both her technology and her thesis about power, partnership and accountability in AI.
WebProNews is an iEntry Publication