Why does a "productless" company have a valuation of 12 billion dollars? An interview with a Silicon Valley expert: Take you to understand the largest seed round in history led by OpenAI's former CTO.

The AI battlefield in Silicon Valley has never been as red - hot as it is today. Just this week, a bombshell news detonated the entire tech circle: Thinking Machines Lab, a rising AI startup founded by Mira Murati, the former Chief Technology Officer (CTO) of OpenAI, announced the completion of a seed - round financing of up to $2 billion. The company's valuation soared to $12 billion. This round of financing was led by top venture capital firm a16z, with industry giants such as chip titans Nvidia, AMD, and Cisco following suit.

It should be noted that TML was just founded in February this year and has not released any products yet. How can a "PPT company" set one of the largest seed - round financing records in Silicon Valley history? This is not only the frenzy of capital but also the ultimate bet on "people".

Mira Murati's Entrepreneurial Legend

Who is Mira Murati? She is no ordinary entrepreneur. As the former CTO of OpenAI, she led the development of groundbreaking technologies such as GPT - 3, GPT - 4, DALL - E, and ChatGPT. She once served as the interim CEO of OpenAI briefly and is hailed as the "Mother of ChatGPT".

Besides her tenure at OpenAI, she was also responsible for the development of Model S, Model X, and Autopilot at Tesla, accumulating profound AI and product experience.

In February 2025, Murati founded Thinking Machines Lab, an AI company focused on building "Collaborative General Intelligence".

The company's vision is to make AI more understandable, customizable, and adaptable to human unique needs through multi - modal AI systems (such as those supporting dialogue, vision, and collaboration).

Currently, the company has not launched an official product yet, but its official website reveals that they are developing multi - modal AI, emphasizing human - AI collaboration, personalized adaptation, and open - source components.

Murati said on social media that the first product will be launched within a few months, which will include an important open - source part to help researchers and startups develop customized models.

Why can this "zero - product" company be valued at $12 billion?

The answer lies in its "super - luxury" founding team! The team brings together many big names from OpenAI, including co - founder John Schulman (the inventor of the PPO algorithm and the person in charge of post - training of ChatGPT), Barret Zoph (Vice President of Research), Bob McGrew (former Chief Research Officer), Alec Radford (the first author of the GPT paper), Lilian Weng (the person in charge of AI safety), etc.

These talents have led the development of key models such as the GPT series, CLIP, DALL - E, and Whisper, and can be regarded as the "dream team" in the AI field.

The Largest Seed - Round in History: AI Talent Becomes the Most Scarce Resource

This round of financing was led by a16z, with participation from giants such as NVIDIA, AMD, Accel, ServiceNow, Cisco, and Jane Street. The amount reached up to $2 billion, and the valuation was $12 billion.

This financing broke the record in the technology history. Initially, the plan was to raise $1 billion with a valuation of $9 billion, but the figures increased significantly within a few months.

Why are investors so generous? Because in the AI era, talent, computing power, and data are the three core elements, and talent is being pushed to the top.

As the industry observes: The AI industry is shifting from "computing - power supremacy" to "talent - centric".

Mark Zuckerberg of Meta launched a "talent war", poaching many experts from OpenAI and DeepMind; Google acquired Character.AI for $2.7 billion, mainly targeting its founding team.

The financing of Thinking Machines Lab is similar to Ilya Sutskever's Safe Superintelligence (valued at $32 billion), highlighting the strong influence of the "OpenAI Mafia" (startups spun off from OpenAI). These companies, although without products, are highly sought after because of the gathering of top - notch talents.

Expert's Front - Line Perspective

For a more in - depth and professional analysis, we specifically contacted an expert in Silicon Valley's expert network recommended by Silicon Rabbit King - A (a pseudonym with no specific meaning), an AI architect who once served as a senior researcher at Google Brain and NVIDIA and is now a consultant for several AI startups. A shared his unique insights through our platform:

"The core reason why Thinking Machines Lab has obtained such a high valuation is not just the personal aura of the team members, but the 'chemical reaction' of this specific team has been verified by the market.

They are not only the engineers who developed foundation models such as GPT and DALL - E but also the researchers who led key alignment work such as RLHF (Reinforcement Learning from Human Feedback).

It can be said that they are one of the teams in the world that best understand how to build, train, and 'tame' ultra - large models. Investors are betting on the high certainty that they will create miracles again.

Secondly, attention should be paid to the concept of 'Collaborative General Intelligence' they proposed. This is different from the AGI (Artificial General Intelligence) pursued by OpenAI or the 'Beneficial AI' emphasized by Anthropic.

'Collaborative' means that the core of its product design may be to create an AI partner that can seamlessly cooperate with humans in complex workflows, rather than just a Q&A tool. This indicates that its technical path may focus more on multi - modal fusion understanding, task decomposition and execution, and personalized adaptation to users' thinking patterns, which has huge potential in the fields of enterprise services and professional creation.

Finally, Murati mentioned that the product will include an 'important open - source part', which is an extremely smart strategy. In the current large - model ecosystem, a completely closed - source approach will face huge pressure from open - source communities such as Llama.

By strategically open - sourcing some core components, Thinking Machines Lab can quickly build a developer ecosystem, attracting global talents to contribute to its platform, thus occupying a favorable ecological niche in the competition with giants such as OpenAI and Google.

Of course, the challenges are still huge. Firstly, whether the collaboration efficiency of the 'dream team' can be maintained at a high level in the startup environment remains to be seen. Secondly, in the current situation of extremely high computing - power costs, how to balance the rhythm between technological exploration and commercialization will be their first real test."

This article is from the WeChat official account "Silicon Rabbit King". The author is Silicon Rabbit King. It is published by 36Kr with authorization.