Back to List
Mira Murati’s New Venture Thinking Machines Unveils Vision for Next-Generation AI Interaction Models
Industry NewsMira MuratiThinking MachinesArtificial Intelligence

Mira Murati’s New Venture Thinking Machines Unveils Vision for Next-Generation AI Interaction Models

Former OpenAI CTO Mira Murati has officially announced the core focus of her new AI startup, Thinking Machines. The company is pivoting away from traditional AI interfaces to develop what it calls "interaction models." These models are designed to facilitate a more natural form of collaboration between humans and artificial intelligence, mimicking the way people interact with one another. According to the announcement, Thinking Machines' technology will continuously process multimodal inputs, including audio and video, to create a seamless collaborative environment. This move marks a significant step for Murati following her departure from OpenAI, signaling a new frontier in how AI systems perceive and respond to human behavior in real-time.

The Verge

Key Takeaways

  • New Venture Identity: Mira Murati, the former Chief Technology Officer of OpenAI, has launched a new AI company named Thinking Machines.
  • Core Product Focus: The company is specializing in the development of "interaction models," a shift from standard large language models.
  • Natural Collaboration: The primary goal is to enable humans to collaborate with AI in the same natural manner they do with other humans.
  • Multimodal Inputs: Thinking Machines' technology is built to continuously ingest and process audio and video data for real-time interaction.

In-Depth Analysis

The Shift Toward Interaction Models

The announcement from Thinking Machines introduces a critical terminology shift in the AI landscape: the "interaction model." While the industry has been dominated by generative models that primarily respond to static text or image prompts, Mira Murati’s new venture is focusing on the process of collaboration itself. By defining their work through the lens of interaction, Thinking Machines suggests a move toward AI that does not just perform tasks on command but participates in a continuous, fluid exchange. This approach aims to bridge the gap between human intent and machine execution by creating a framework where the AI is an active participant in a shared environment.

The concept of an interaction model implies a departure from the "turn-based" nature of current AI tools. Instead of a user providing an input and waiting for an output, the model is designed to exist within a collaborative loop. This aligns with the company's stated goal of allowing people to work with AI as they would with a human colleague, where communication is often non-linear, contextual, and ongoing.

Continuous Multimodal Processing

A defining technical characteristic of the interaction models proposed by Thinking Machines is their ability to continuously take in audio and video. Traditional AI models often process data in discrete batches—a single audio file, a specific image, or a block of text. Thinking Machines is moving toward a "continuous" intake system. This suggests that the AI will be "always-on" in terms of perception, allowing it to pick up on subtle cues that are present in human-to-human communication but lost in traditional AI interfaces.

By integrating audio and video as primary, continuous streams, these interaction models can potentially understand context, tone, and visual environment in real-time. This capability is essential for the "natural collaboration" that Murati’s company envisions. If an AI can see what a user is doing and hear the nuances of their speech as they work, the level of friction in the collaboration is significantly reduced. This technical direction points toward a future where AI is integrated into the physical and digital workspace as a perceptive entity rather than a reactive tool.

Redefining Human-AI Collaboration

The philosophy behind Thinking Machines centers on the word "natural." In the context of the company's announcement, natural collaboration means moving away from the specialized syntax or "prompt engineering" currently required to get high-quality results from AI. If the AI can interact like a human, it must be able to handle the ambiguities and multi-layered communication styles that humans use instinctively.

This vision for Thinking Machines suggests a focus on the user experience as much as the underlying intelligence. By prioritizing the interaction layer, Murati is addressing one of the most significant hurdles in AI adoption: the interface. If Thinking Machines succeeds in creating models that can truly collaborate, the AI becomes less of a software application and more of a digital partner. This focus on the "how" of AI interaction could redefine the standards for the next generation of AI startups entering the market.

Industry Impact

The emergence of Thinking Machines and its focus on interaction models signals a new phase in the AI industry. As the foundational capabilities of large models become more commoditized, the competitive frontier is shifting toward how these models interface with the real world and human users. Murati’s emphasis on continuous audio and video processing sets a high bar for multimodality, pushing other players in the industry to move beyond text-heavy interfaces.

Furthermore, the launch of Thinking Machines highlights the ongoing trend of high-profile talent from established AI giants like OpenAI striking out to form independent ventures. This decentralization of expertise fosters a diverse range of approaches to AI development. Thinking Machines' specific focus on the "interaction" aspect may force a re-evaluation of what constitutes a "state-of-the-art" model, moving the metric from mere parameter count to the fluidity and naturalness of human-machine engagement.

Frequently Asked Questions

Question: What is the main goal of Mira Murati's new company, Thinking Machines?

Thinking Machines aims to develop "interaction models" that allow humans to collaborate with artificial intelligence in a natural way, similar to how humans collaborate with each other.

Question: How do Thinking Machines' interaction models differ from traditional AI?

Unlike traditional models that often rely on static prompts, Thinking Machines' models are designed to continuously process audio and video inputs to facilitate real-time, fluid collaboration.

Question: Who is leading Thinking Machines?

Thinking Machines was founded by Mira Murati, who previously served as the Chief Technology Officer (CTO) of OpenAI.

Related News

Meituan LongCat Team Releases General 365 Benchmark Revealing Reasoning Gaps in Leading AI Models
Industry News

Meituan LongCat Team Releases General 365 Benchmark Revealing Reasoning Gaps in Leading AI Models

The Meituan LongCat team has officially introduced General 365, a new evaluation benchmark designed to test the reasoning capabilities of large language models. In a recent assessment of 26 mainstream models, the benchmark revealed a significant performance gap across the industry. Gemini 3 Pro, currently identified as the strongest model in the test, achieved an accuracy rate of 62.8%. However, the results indicate a broader struggle within the field, as the vast majority of the 26 models tested failed to reach the 60% accuracy threshold, which is considered the passing mark. This release by Meituan's technical team establishes a new standard for measuring AI reasoning, highlighting that even top-tier models have substantial room for improvement in complex cognitive tasks.

Managing AI Coding Through Agent Evaluation: A 310,000-Line Code Refactoring Case Study
Industry News

Managing AI Coding Through Agent Evaluation: A 310,000-Line Code Refactoring Case Study

As AI-generated code begins to account for over 90% of system development, the primary challenge shifts from increasing coding speed to managing and constraining AI output. Meituan's technical team has shared a comprehensive practice involving the refactoring of 310,000 lines of code using an 'Agent evaluation' mindset. By implementing a structured framework—including technical debt sorting, rule construction, standardized operating procedures (SOP), and a Pre-PR (Pull Request) mechanism—the team successfully transitioned code refactoring from a high-cost, specialized project into a sustainable, daily iterative process. This approach addresses the risk of AI-driven development amplifying system chaos and emphasizes the necessity of unified standards in the era of AI-native programming.

Meituan BI Evolution: Building a Next-Generation Architecture with Metrics Platforms and Enhanced Calculation Engines
Industry News

Meituan BI Evolution: Building a Next-Generation Architecture with Metrics Platforms and Enhanced Calculation Engines

Meituan's data platform team has pioneered a new generation of Business Intelligence (BI) architecture, placing a centralized metrics platform at its core. This strategic shift addresses critical limitations found in traditional BI systems, which often suffer from inconsistent data definitions—commonly known as "data caliber confusion"—and sluggish query performance when handling personalized datasets. By developing and implementing two primary technical capabilities, automatic semantics and enhanced calculation, Meituan has successfully streamlined its data processing workflows. This evolution marks a significant transition from dataset-driven analytics to a more robust, metrics-centric model, ensuring higher data reliability and faster insights for the organization's diverse business operations. The practice underscores Meituan's commitment to solving complex data engineering challenges through architectural innovation.