OpenAI would’ve ‘imploded’ if Altman didn’t return, ex-CTO says

Former OpenAI CTO Mira Murati has stated that the company would have 'imploded' if Sam Altman hadn’t returned as CEO following his brief ouster in November 2023. Murati, who briefly served as interim CEO before being replaced by Emmett Shear, said she acted quickly after Altman’s firing to prevent the company’s collapse, fearing instability and mass defections. Over 700 employees threatened to quit and join Microsoft, accelerating Altman’s reinstatement within five days alongside a reconstituted board. Murati’s account aligns with her earlier testimony in Elon Musk’s lawsuit against OpenAI, where she warned of the company’s potential unraveling post-Altman. She acknowledged concerns about his leadership but emphasized the need for continuity and stability during the crisis. Former board member Helen Toner, however, depicted Murati’s role as more hesitant, suggesting she initially deferred to others’ decisions. Now leading Thinking Machines Lab, Murati advocates for a shift in AI development toward 'interaction models,' which capture nuanced human communication like pauses and interruptions. Her startup, valued at $12 billion after a $2 billion funding round last year, has released a developer tool called Tinker but has yet to launch a consumer product. Murati contrasts this approach with current AI models, which she describes as rigidly turn-based. She also criticized the industry’s reliance on individual leaders, arguing that AI governance should prioritize transparent decision-making structures over personal trust. When asked about competing with Altman or Anthropic’s Dario Amodei, Murati dismissed competitive drive as a motivator, instead emphasizing collaboration between humans and machines. Her vision centers on preventing AI from outpacing human control, framing governance as essential to ethical progress.