In one of the most audacious moves in the artificial intelligence world this decade, Mira Murati—the former Chief Technology Officer at OpenAI—has secured a staggering $2 billion in funding for her new venture: Thinking Machines Lab. With ambitions to push the frontiers of Artificial General Intelligence (AGI), Murati is not just building another AI startup—she’s engineering a seismic shift in how humanity interacts with intelligent systems.
This marks a pivotal moment in Murati’s career and the broader AI landscape. Long regarded as one of the most influential yet understated figures behind some of OpenAI’s most revolutionary models, she has now stepped fully into the spotlight, leading a new company with a singular mission: to build machines that can reason, learn, and think like humans—without compromising on safety or ethics.
From OpenAI to the Open Frontier
Mira Murati played a crucial role in the development of GPT-4 and GPT-4o at OpenAI, helping turn generative AI into a consumer revolution. But insiders say Murati’s vision extended far beyond large language models. Her interests were deeply rooted in cognitive architectures, long-term memory systems, and adaptive reasoning—domains that many believe hold the key to AGI.
After stepping down as CTO in early 2025, Murati quietly assembled a team of top AI researchers, neuroscientists, mathematicians, and roboticists. Within six months, she had launched Thinking Machines Lab, headquartered in a minimalist campus in Northern California. The lab’s mission? To build a cognitive engine that doesn’t just simulate intelligence—but embodies it.
The $2 Billion Bet
The announcement of the $2 billion raise shocked Silicon Valley—not just because of the amount, but because of who the backers are. While Murati has declined to name all investors, sources close to the matter reveal that a mix of sovereign wealth funds, top-tier venture capitalists, and AI-focused philanthropists are behind the effort.
According to industry observers, this is the largest single round ever raised for an AI company pre-product. What makes it even more extraordinary is that the lab has yet to release a public-facing demo or even a technical paper.
But the investors aren’t concerned. For them, backing Murati isn’t a gamble—it’s a calculated investment in someone they consider the spiritual successor to the early pioneers of AI.
A New AGI Paradigm: “Synthetic Cognition”
At the heart of Thinking Machines Lab is a new architecture Murati calls Synthetic Cognition—a system inspired by the brain’s ability to integrate reasoning, perception, and memory into a unified model.
Unlike transformer-based models that rely on statistical prediction, Synthetic Cognition combines symbolic reasoning with probabilistic learning and multimodal perception. It aims to create systems that can think through problems, reflect on prior knowledge, and adapt to novel situations with minimal human instruction.
Internally, the lab is working on a prototype known only as “ALMA”, reportedly an acronym for Adaptive Learning and Memory Architecture. ALMA is said to integrate three major breakthroughs:
- Dynamic episodic memory systems that allow models to recall specific past interactions and evolve over time.
- Goal-oriented planning modules that can simulate multiple outcomes before making a decision.
- Sensorimotor grounding, enabling the model to learn from real-world robotic interfaces rather than just text or image data.
Engineering AGI for the Real World
Thinking Machines Lab is not just building algorithms; it’s building a platform for the deployment of AGI in real-world settings. From healthcare diagnostics to autonomous exploration systems, Murati envisions a future where AGI can assist, not replace, human expertise.
To that end, the lab is creating a modular framework for safety alignment. Rather than treating alignment as a secondary feature, Murati’s team has integrated it into the core learning objectives of the system. Every model output is evaluated through multi-layered ethical constraints, dynamic reinforcement signals, and user feedback loops.
The lab has also set up a Bioethics & Governance Council, composed of philosophers, ethicists, and legal scholars, who will oversee the deployment and testing of every major release.
The Team: Minds from MIT to DeepMind
Murati has attracted an all-star team, pulling talent from top institutions like MIT, Stanford, and the Max Planck Institute. Several former engineers from DeepMind, Anthropic, and Meta AI have also joined the lab, drawn by what insiders describe as a “mission-first culture.”
The team is structured into three core divisions:
- Neuro-Inspired Architectures: Focused on translating insights from neuroscience into machine learning algorithms.
- Cognitive Safety Systems: Dedicated to alignment, value learning, and interpretability.
- Synthetic Embodiment: Working on AGI systems that learn through robotics, virtual environments, and embodied simulation.
What sets this team apart is its transdisciplinary approach. Researchers don’t just work on code—they’re encouraged to study philosophy, read psychology literature, and explore the frontiers of human cognition.
Silence, Strategy, and the Element of Surprise
Murati’s low-profile launch contrasts sharply with the hype culture dominating the tech world. There have been no splashy keynotes, no teaser demos, and no aggressive social media campaigns. The lab’s website is sparse—just a mission statement and an abstract vision.
But those who know Murati say this is entirely by design. She is said to believe that AGI is too important to be reduced to hype cycles or product rollouts. Her strategy is to build quietly but with velocity, and to only unveil the system when it passes a threshold of capability that she defines as “emergent competence.”
Reactions Across the AI Industry
The AI community is abuzz. While some see Murati’s venture as the most credible AGI initiative since DeepMind’s AlphaFold success, others are skeptical. Critics argue that the AGI label is overused, and question whether any current model can truly generalize in the way humans do.
Still, even the skeptics concede that Murati’s involvement gives the lab instant credibility. Former colleagues at OpenAI have publicly expressed support, noting her “unshakable integrity” and “clear long-term vision.”
Meanwhile, a quiet unease is spreading through Big Tech. With $2 billion in the bank and no pressure to generate short-term revenue, Thinking Machines Lab is free to pursue bold, unorthodox paths. This could put pressure on companies like OpenAI, Anthropic, and Meta to rethink their own AGI roadmaps.
What’s Next: A 24-Month Countdown
Sources close to Murati suggest that the lab has set an internal milestone to demonstrate a working AGI prototype within 24 months. If successful, this would make Thinking Machines Lab the fastest lab in history to reach AGI-level capabilities.
To reach this goal, the lab is scaling compute at unprecedented levels. Rumors suggest partnerships with major chipmakers to develop specialized hardware optimized for synthetic cognition systems. Some even believe the lab may be designing its own silicon.
In parallel, small-scale pilot programs are being prepared in domains such as:
- Advanced tutoring systems for underserved schools
- Cognitive assistants for elderly care
- Autonomous scientific research agents
Each of these use-cases is being built with a dual objective: to test the system’s generalization capacity and to refine safety mechanisms in real-world settings.
Mira Murati: The Reluctant Visionary
Despite the enormous attention, Mira Murati remains characteristically humble. She rarely gives interviews and prefers collaborative whiteboard sessions to boardrooms. But those who’ve worked with her describe her as a “quiet visionary” who possesses a rare blend of technical brilliance and philosophical depth.
She has spoken in closed sessions about the “moral responsibility” of building AGI—a responsibility not just to prevent harm, but to maximize human flourishing. Her speeches are said to be grounded as much in literature and ethics as in computer science.
Conclusion: A New Chapter in Intelligence
The launch of Thinking Machines Lab under Mira Murati’s leadership is more than just a startup story—it is the beginning of a new chapter in the quest for artificial intelligence. With $2 billion, an elite team, and a radically ambitious vision, the lab is poised to shape the next era of computing.
But AGI is not just a technical problem—it’s a societal one. How we choose to develop, align, and distribute this intelligence will define the trajectory of the 21st century.