Mira Murati, former CTO of OpenAI, details her new venture, Thinking Machines, which focuses on developing interaction models that prioritize human-AI collaboration. Unlike traditional turn-based AI, these models process continuous, multi-modal data in real-time, enabling nuanced, high-bandwidth interactions that keep humans actively engaged. Murati emphasizes that advancing AI requires intentional institutional design, transparency, and governance rather than relying solely on individual leadership. Reflecting on the 2023 OpenAI board crisis, she highlights the necessity of prioritizing mission continuity and team stability during periods of intense organizational volatility. Ultimately, she advocates for building AI systems that function as "tools for thought," expanding human agency and dignity rather than merely pursuing autonomous capabilities. This approach seeks to align AI development with human values while navigating the competitive, high-stakes landscape of frontier AI labs.
Sign in to continue reading, translating and more.
Open full episode in Podwise