Thinking Machines, the AI startup co-founded by Mira Murati, former Chief Technology Officer of OpenAI, has unveiled its latest innovation: interaction models. Announced on Monday, these models are engineered to transform how humans collaborate with artificial intelligence.

According to the company, the core concept behind interaction models is to replicate natural human collaboration. Unlike traditional AI systems, which process inputs in a linear, single-threaded manner, interaction models are designed to continuously ingest and analyze audio, video, and text in real time. This enables the AI to think, respond, and act without waiting for user input to complete.

The announcement highlights a critical limitation in current AI models: "Today's models experience reality in a single thread. Until the user finishes typing or speaking, the model waits with no perception of what the user is doing or how the user is doing it."

Thinking Machines emphasizes that interaction models aim to bridge this gap by providing a more dynamic and responsive AI interaction framework. The company suggests that this approach will allow for seamless, real-time collaboration, closely resembling how humans naturally interact with one another.

Source: The Verge