Ex-DeepMind Founders Secure $600M Total for Robot 'Brain' Startup Now Worth $3B

Aug 261 views

Key takeaways

Generalist raised ~$200M in a Series B extension led by 8VC, reaching a $3B valuation.
Founded in 2024 by ex-Google DeepMind and Boston Dynamics engineers; total round now $600M.
Rivals Physical Intelligence ($11B) and Skild AI ($14B) highlight an intensely funded robotics AI race.

Generalist has closed an extension to its Series B round worth nearly $200 million, led by 8VC, lifting the startup's valuation from $2 billion to $3 billion, according to two people with knowledge of the deal. The fresh capital extends the $400 million Series B that Radical Ventures led and that Generalist publicly announced in June, bringing the total round size to $600 million. Neither Generalist nor 8VC responded to requests for comment.

The company was founded in 2024 by Pete Florence and Andy Zeng, both former Google DeepMind researchers, alongside Andrew Barry, a former engineer at Boston Dynamics. That founding team gave Generalist early credibility in the robotics AI space, drawing backing from Nvidia, Union Square Ventures, Bezos Expeditions, and AI pioneer Fei-Fei Li, alongside lead investors 8VC and Radical Ventures. Despite this high-profile lineup, the startup kept a deliberately low public profile for most of its early existence.

Generalist's core product is an AI foundation model designed to operate across a wide variety of robotic hardware, rather than being purpose-built for a single machine or task. Its most recent release, Gen 1.5, reportedly allows robots to learn new tasks by watching video demonstrations lasting as little as three to twelve seconds. This capability — rapid task acquisition from minimal instruction — is central to the company's pitch that it can serve diverse industrial and commercial customers.

The startup is currently working with a small number of customers, using their real-world feedback to refine the model for specific applications, one source said. This approach mirrors how early LLM companies used enterprise pilots to shape general-purpose models before broader deployment, suggesting Generalist is still in a market-discovery phase.

Generalist operates in a crowded and richly funded field. Physical Intelligence holds a reported $11 billion valuation, while SoftBank-backed Skild AI sits at $14 billion. Genesis AI was reportedly in talks as recently as last month to raise capital at a $3 billion valuation, the same mark Generalist just hit. Investors driving these rounds appear to be betting that robotics is approaching its own inflection point — a moment where machines can handle general tasks without explicit per-task training — though some venture capitalists caution that the data constraints facing robotics are far more severe than those LLMs encountered, and a truly general robotics model may remain years away.

The bigger picture

The $3 billion valuation Generalist has now achieved is notable less for the number itself and more for how quickly it arrived. The company was founded in 2024 and has already stacked $600 million in Series B funding, which places it in the same conversation as Physical Intelligence and Skild AI — both of which have been building longer and carry valuations several times higher. That Generalist is closing this gap so rapidly signals that investors are not waiting for proof-of-concept deployments before writing very large checks.

The competitive picture here matters enormously. Physical Intelligence, Skild AI, and now Generalist are all chasing what amounts to the same prize: a foundation model layer that sits between raw robotic hardware and task-specific software. Whoever establishes themselves as the default 'brain' for third-party robot manufacturers holds extraordinary leverage — the kind that Nvidia currently enjoys in GPU compute. Hardware makers like Boston Dynamics, Agility Robotics, and a generation of humanoid startups would effectively become dependent on whichever software layer wins this race. That dynamic explains why investors are willing to fund multiple competitors simultaneously.

The honest caveat, noted even by some of the VCs participating in these rounds, is that robotics AI faces a data wall that language models never encountered. LLMs trained on the breadth of the internet; robots must learn from physical interaction data that is expensive and slow to generate. Gen 1.5's claim of task learning from three-to-twelve-second videos is intriguing, but it will need to hold up at scale across genuinely varied environments before investors' 'ChatGPT moment' framing becomes anything more than aspiration. Readers should watch how Generalist's small customer base expands — or doesn't — over the next twelve months.

LagPing's take

We're covering Generalist's funding round because it captures something important happening right now at the intersection of AI and physical hardware. The people building this company came from Google DeepMind and Boston Dynamics — organizations that spent years pushing the frontier of what robots and AI models can do — and they're now working with serious capital to combine those two worlds. That founding pedigree, combined with backers like Nvidia and Fei-Fei Li, tells us this isn't a speculative moonshot; it's a well-resourced attempt to solve a real and difficult problem. We also think the broader race to build a universal robot 'brain' is underreported relative to how consequential it could be for manufacturing, logistics, and everyday automation. Generalist's Gen 1.5 model and its rapid-learning claims deserve scrutiny, and we'll be watching how those capabilities perform outside of controlled demonstrations. This story fits squarely into LagPing's ongoing coverage of where AI moves beyond software and starts reshaping the physical world.

X / Twitter WhatsApp

View at TechCrunch AI

Find "Google DeepMind" on Amazon

As an Amazon Associate, LagPing earns from qualifying purchases. Product links are affiliate links.