Former OpenAI CTO Mira Murati Unveils "Interaction Model" with Thinking Machines
Thinking Machines has announced its "Interaction Model," an AI capable of real-time processing and dialogue, aiming to reshape human-AI collaboration.
Former OpenAI Chief Technology Officer (CTO) Mira Murati has announced that her AI company, Thinking Machines, is developing a groundbreaking concept called the “Interaction Model.”
Toward “Natural Collaboration” Between Humans and AI
The core of the Interaction Model proposed by Thinking Machines is to create an environment where humans and AI can collaborate as seamlessly as humans work together. Most current AI models operate in a “single-threaded” manner, waiting for users to finish typing text or speaking before starting to process the input.
Thinking Machines likens this approach to “holding important discussions over email,” highlighting how such limited communication bandwidth can hinder the AI’s understanding of user intentions or make the AI’s processes less transparent to users.
In contrast, the Interaction Model aims to continuously process audio, video, and text inputs, enabling the AI to think, respond, and act in real time. The company explains that this allows the AI interface to adapt to humans, eliminating the need for humans to adjust their behavior to accommodate the AI.
Demonstrations and Future Plans
To showcase the potential of this technology, Thinking Machines has released several demonstration scenarios. These include detecting and responding immediately when an animal is mentioned in a story, translating spoken words in real-time, and alerting users when they exhibit poor posture, such as slouching.
However, the Interaction Model is not yet available to the general public. Thinking Machines plans to release a limited research preview “within the next few months,” followed by a broader release “later this year.”
From Founding to Current Developments
After leaving OpenAI in February 2025, Mira Murati founded Thinking Machines. Since its inception, the company has faced challenges, such as key team members moving to competitors like Meta or returning to OpenAI. The announcement of this new technological concept represents a significant milestone amid these challenges.
If the Interaction Model becomes practical, it has the potential to transform AI applications across various fields, including voice assistants, real-time translation, education, and healthcare.
Frequently Asked Questions
- What sets Thinking Machines' Interaction Model apart?
- The biggest differences are "real-time capability" and "multimodal input." Unlike traditional AI, which waits for users to complete their input before processing, the Interaction Model continuously processes audio, video, and text inputs to engage in ongoing dialogue. For example, it aims to understand the context of a conversation while the user is still speaking and provide immediate, relevant responses.
- When will Thinking Machines launch its services?
- An official public release date has not yet been announced. However, the company plans to offer a limited research preview "within the next few months," with a broader release expected by the end of 2026. It appears the service will initially target researchers and developers.
- Why did Mira Murati leave OpenAI?
- Specific reasons for Murati's departure from OpenAI have not been disclosed. She left the company in February 2025 and shortly thereafter founded Thinking Machines, seemingly to pursue her own vision for AI research and development.
Comments