Strip away the corporate restructuring language and this is a straightforward strategic bet: Google has decided the most valuable place for its robotics software division is inside the mothership, not at arm's length in Alphabet's experimental portfolio. Intrinsic announced on 25 February that it is formally joining Google after five years as an independent Alphabet company, with the goal of accelerating what the industry is calling "physical AI" — software that enables machines to perceive and act in the real world.
The numbers that motivate this decision are hard to ignore. McKinsey projects the market for general-purpose robots could reach $370 billion by 2040, a figure that has concentrated minds at every major technology company. For Google, bringing Intrinsic into its core business is less about a dramatic pivot and more about removing the structural barriers that were slowing down the work.
From Moonshot to Mainstream
After five and a half years developing its technology at Alphabet's "moonshot factory" X, Intrinsic became an independent Alphabet company in its "Other Bets" division in July 2021. That independence gave it room to experiment, and experiment it did: in April 2022 the company acquired Vicarious, a robotics software firm, and then purchased several for-profit divisions of Open Robotics, a nonprofit that develops hardware and software platforms for the robotics industry. Those moves came at a cost. Despite these acquisitions, Intrinsic laid off 20 per cent of its workforce in January 2023.
What emerged from that leaner period was Flowstate, Intrinsic's flagship product. Flowstate is a web-based development environment and simulation engine that enables users to build applications with ready-to-use "building blocks of robotic behaviours" known as skills, without the need for deep expertise and hundreds of programming hours. The pitch is essentially democratisation: while hardware like robotic arms has become cheaper, programming them remains incredibly complex, often requiring hundreds of hours of manual coding by specialised engineers. Flowstate is designed to fix that.
Much like Android provides developers a universal platform for building apps that work across different mobile devices, Intrinsic supports app building across different robots, cameras, sensors, AI models and supporting hardware. The Android analogy is deliberate and commercially savvy. Google built enormous enterprise and developer leverage through Android without manufacturing a single handset. The same logic applies here: create a software layer that is agnostic to hardware and work across different brands of industrial arms and sensors, allowing Google to act as a partner to manufacturers rather than a direct competitor in hardware.
Why Now, and Why Inside Google?
When a subsidiary is independent, it often struggles to access the full sales force and enterprise reach of the Google Cloud machine. That structural friction is precisely what this move resolves. As a distinct group within Google, Intrinsic will continue evolving its platform, utilising the capabilities of Gemini models and Cloud, while working closely with Google DeepMind. CEO Wendy Tan White will remain in her position, reporting to Hiroshi Lockheimer, the chief product officer of Other Bets at Google.
The DeepMind connection is significant. In mid-2025, Google debuted two new AI models, Gemini Robotics and Gemini Robotics-ER (extended reasoning), bringing generative AI into physical action commands to control robots. Gemini Robotics is an advanced vision-language-action model built on Gemini 2.0, with physical actions added as a new output modality for the purpose of directly controlling robots. Merging Intrinsic's developer tools and manufacturing relationships with those models creates a more coherent product than either could offer separately.
The competitive pressure is real. Competitors like Amazon are deploying thousands of warehouse robots and Tesla is pushing its humanoid "Optimus" platform, making the race to embed AI into physical systems one of the defining technology contests of the decade. Nvidia's Jensen Huang and Qualcomm's Cristiano Amon have both publicly identified physical AI as the next wave of technology monetisation.
A History Worth Acknowledging
Google's robotics track record deserves scrutiny alongside the optimism. In 2013, Alphabet bought Boston Dynamics and Schaft, a Japanese humanoid robotics company, along with several vision startups. After several years spent trying to build a clear business in the space, Google sold Boston Dynamics and Schaft in 2017 to SoftBank for an undisclosed amount. That retreat is a legitimate data point for sceptics, and it would be unwise to ignore it.
Critics of big tech consolidation will also raise fair questions about what happens to an independent robotics ecosystem when Google becomes the dominant platform provider. The Android analogy cuts both ways: Android created enormous value for developers, but it also concentrated power with Google in ways that have since attracted sustained regulatory scrutiny globally, including from the Australian Competition and Consumer Commission. An "Android of robotics" could, over time, raise similar concerns about lock-in and market concentration in industrial automation.
There are also legitimate questions about the workforce impact. Intrinsic claims that skills can be manually developed or AI-enabled, without the need for deep expertise and hundreds of programming hours. For manufacturers, that sounds like efficiency. For the specialised engineers who currently write that code, it sounds like job displacement. The productivity gains from physical AI are likely to be unevenly distributed, at least in the short term.
The Bigger Picture for Investors and Industry
What the market hasn't fully priced in yet is the compounding effect of combining a mature developer platform, a growing library of Gemini-powered robotics models, and Google Cloud's enterprise distribution. Intrinsic launched a joint venture with Foxconn last year to build an AI-driven intelligent factory of the future, combining Intrinsic's AI robotics platform with Foxconn's electronics manufacturing expertise. That kind of anchor customer relationship, now backed by Google's full commercial infrastructure, is a materially stronger proposition than it was twelve months ago.
The honest assessment is that Google is making a considered, well-resourced push into a space where it has failed before. The difference this time is the maturity of the underlying AI models and a clearer software-first strategy that avoids the hardware pitfalls of the Boston Dynamics era. Whether the Android playbook truly translates to factory floors, assembly lines, and logistics centres is still an open question. The opportunity is real, the competition is fierce, and the outcome is far from certain. For now, the smart money is watching whether Intrinsic's developer community grows as fast as Google DeepMind's Gemini Robotics models improve, and whether a consolidated Google can move faster than the leaner challengers already running in the same direction. Australian Bureau of Statistics data on local manufacturing automation investment will be worth watching as this technology matures and begins to reach Pacific markets.