Physical AI: When Intelligence Takes Physical Form
When we invented machines stronger than us, we thought it was great. When we built machines faster than us, we celebrated. Now we’re witnessing machines that could be more intelligent than us, and suddenly we’re not so thrilled about it.
Why? Perhaps because that intelligence will be valuable, and someone will want to control it. For decades, humanoid robots were pure science fiction. Today, with the rise of general-purpose robots, we’re talking about bipedal machines with computer vision, tactile skin, natural language processing, and the ability to perform multiple tasks in human environments. They’re no longer just arms in a factory. Automation has started to take physical form.
The Third Generation: Figure 03 and the New Reality
The most significant advances in physical robotics and AI integration are concentrated in Figure 03, the third-generation humanoid created by Figure AI. Unlike previous models, this one has been designed directly for mass production and seems to learn more naturally in human environments than its predecessors. It’s no longer a laboratory prototype. It’s a machine ready for serial production and integration into factories, warehouses, or even homes.
Figure AI has secured over $675 million in funding and reached a valuation of $2.6 billion, reflecting the enormous scale of investment in this new form of automation. With this momentum, Figure AI has built BotQue, a facility capable of producing humanoids at scale. In just 4 years, they could manufacture 100,000 units annually from that single factory alone.
Companies like UPS are already testing Figure robots in their warehouses, loading and sorting packages alongside human employees. An automotive manufacturer has them on staff. This isn’t science fiction anymore. They’re operating and occupying positions that were once exclusively human.
The Competitive Landscape
Figure isn’t alone in this race. Tesla competes with its Optimus humanoids. Still in development, but it’s a project Elon Musk considers even more important than Tesla’s automotive business. Musk’s plan is clear: use them first in Tesla factories for repetitive or dangerous tasks, then start selling them to companies and homes. Musk has talked about producing 1 million units per year by 2030 with a target cost of $20,000 per robot. A price that would be affordable in many cases.
Meanwhile, Boston Dynamics explores a different path with its Atlas robot, which went viral for its superhuman agility. Running, jumping, doing parkour. Their Spot and Stretch models are used commercially for inspections and logistics. Their approach focuses on perfecting the physical body that others will endow with intelligence.
In Asia, Xiaomi and UBTech lead the domestic race with robots like Cyber One or Walker, while Japan pushes humanoids for elderly care. Addressing a significant demographic challenge.
The Economic Reality: When Robots Become Cheaper Than Humans
The next consequence will be a drastic drop in labor costs. Mass production reduces prices, and companies like Figure and Tesla are already planning to manufacture tens of thousands of units. In a short time, we’ll face a robot worker that will stop being a luxury and become a very profitable option for companies. Much more profitable than human employment.
When maintaining a humanoid costs less than a salary, replacement won’t be a moral debate but a simple question of economic efficiency. And that always finds a way.
The Numbers Don’t Lie
Recent studies estimate that automation could displace between 400 and 800 million people by 2030, with about 375 million unable to find employment matching their skills. That’s 14% of the global workforce without a place in the economy. We’re heading toward that reality.
We’re no longer talking about the future. We’re talking about the present. Self-checkout kiosks replace cashiers, algorithms write reports and analyze data, and autonomous trucks are beginning to threaten millions of drivers.
The Structural Challenge
The problem isn’t technological advancement. It’s the lack of structural anticipation from those in charge. Mass unemployment won’t be the worst part. What will be truly decisive is who controls the new robotized workforce.
The big tech companies designing these systems will become the new owners of the means of production. They’ll manufacture the machines, rent them out, manage their software, and control the data that makes them function. Tesla, Figure, Amazon, Alibaba could dominate millions of robots and the AI networks that direct them.
The Open Source Imperative
As these robots become more prevalent in our daily lives, there’s a critical question: Who controls the code that runs the machines walking among our children?
The current landscape is dominated by proprietary systems from major tech companies. But there’s a growing movement toward open source robotics. And it’s not just about transparency. It’s about safety, innovation, and democratic control of technology.
Linux is becoming the de facto platform for physical AI systems. Most robots today run on Linux-based operating systems, from industrial robots to research platforms. This isn’t coincidental. Linux provides the flexibility, real-time capabilities, and security features that physical AI demands.
Edge Computing: The Perfect Storm for Physical AI
Physical AI represents one of the most compelling use cases for edge computing. Unlike traditional cloud-based AI, physical robots need to make split-second decisions in real-time. They can’t afford the latency of sending data to a distant server and waiting for a response.
This is where edge computing becomes crucial:
- Real-time processing: Robots need to process sensor data and make decisions in milliseconds
- Bandwidth efficiency: Streaming all sensor data to the cloud would be impractical
- Reliability: Physical systems can’t depend on network connectivity
- Privacy: Local processing keeps sensitive data on-device
The convergence of edge computing, open source software, and physical AI creates an unprecedented opportunity for innovation. Projects like ROS (Robot Operating System) are already building the foundation for open source robotics, while edge computing platforms like Kubernetes on the edge are making it easier to deploy and manage AI workloads locally.
The Need for Action Over Regulation
The current regulatory approach focuses heavily on AI ethics and safety frameworks. Important, but insufficient. The real challenge isn’t just regulating AI. It’s preparing for a world where physical AI becomes ubiquitous.
Instead of over-regulating before understanding the technology, we need:
- Open source development that allows community review of robot code
- Transparent algorithms that can be audited and improved
- Collaborative research between academia, industry, and government
- Investment in education to prepare the workforce for this transition
The goal isn’t to stop progress. It’s to ensure that when robots walk among our children, we can trust the code that controls them. Open source isn’t just about transparency. It’s about democratic control of the technology that will shape our future.
The Historical Perspective
In 1930, John Maynard Keynes spoke of a new disease. Technological unemployment. He said it would be caused by advances that reduce the need for labor faster than new occupations emerge, and he predicted that by 2030 we would live with much shorter workdays, almost instantly, because machines would assume most necessary tasks. He wasn’t wrong about that.
What Happens When Work Disappears?
For centuries, work has been the axis of the economy and our identity. Getting a job, building a career, retiring. But what happens when the economy no longer needs most human work?
The promise of unlimited progress mixes with a disturbing question for those of us who defend technology. What will humans do when we’re no longer needed to sustain the economy?
The Real Questions
While governments take refuge in the old optimism that technology will create as many jobs as it destroys, history and current data tell us that happens in the long term, not immediately. If this trend continues, the next decade will bring profound social changes.
Many people might stop working to live, but not in a scenario of freedom and creative leisure. But one marked by the emptiness of purpose and economic dependence.
The Path Forward
We’re approaching a historical inflection point. The combination of physical robots and artificial intelligence is replacing not only manual tasks but also cognitive functions. A general-purpose humanoid can already learn almost any everyday task, from loading boxes to preparing coffee, and threatens even jobs that were considered safest.
The old argument that there are things only a human will always be able to do is starting to lose validity.
Conclusion: The Choice Is Ours
We’re literally on the edge of one of the greatest civilizational transformations, comparable to the invention of the steam engine and the changes that followed with electricity. Yet institutions move with exasperating slowness, trapped between denial and fear, as if they don’t want to see that the future is already knocking at the door.
The outcome will depend on our attitude. It can be a utopia of creativity and shared well-being or a dystopia of control and exclusion. The difference will be in how much we defend our freedoms and our privacy. Because that will be the key they’ll try to take from us to subdue us.
The future won’t be decided by those useless, short-sighted politicians incapable of looking beyond the next headline, nor by the algorithms they fear and venerate at the same time. We’ll decide it with our values and our actions, starting today.
Instead of fearing displacement, we should rethink our purpose. Maybe we’ll discover that our value never lay in being just another piece of an economic machine. But in our creativity, our empathy, our infinite capacity to reinvent ourselves.
Because that’s what this is about: reinventing ourselves.
What do you think about the rise of Physical AI? Are we ready for a world where robots work alongside, or instead of, humans? Let me know your thoughts in the comments below.