Mira Murati's New AI Aims for Real-Time Human Collaboration

Former OpenAI CTO Mira Murati is back in the AI game with her new venture, Thinking Machines. The company announced this week that it's developing what they're calling "interaction models." The core idea behind these models is to create AI that can collaborate with humans in a way that feels as natural as our interactions with each other.

According to Thinking Machines, these new models are built to continuously process a wide range of inputs, including audio, video, and text. This multi-modal approach is intended to allow the AI to think, respond, and act in real time, mirroring how humans naturally communicate and work together. Unlike current AI models that often wait for a user to finish a specific input before processing, interaction models aim to have a constant perception of the user's actions and context.

This development from Murati’s new company suggests a significant step towards more intuitive and integrated AI experiences. The goal is to move beyond the current limitations of AI, where interactions are often sequential and lack a deeper understanding of real-time human behavior. By enabling AI to 'see,' 'hear,' and 'understand' in a more holistic and immediate manner, Thinking Machines hopes to pave the way for AI that can truly partner with us, rather than just respond to commands.

The implications of such real-time, multi-modal AI are vast, potentially transforming everything from personal assistants to complex professional workflows. The ability for AI to grasp nuances in tone, body language, and spoken words simultaneously could lead to more effective problem-solving and creative endeavors. It’s an ambitious vision that could redefine our relationship with artificial intelligence, making it a more seamless and integral part of our daily lives and work.