Physical AI is bringing artificial intelligence into the real world, combining robotics and intelligence to transform industries. From warehouses to healthcare, here’s how this new wave of AI can drive positive global impact.
Artificial intelligence is no longer confined to computer screens and data models. A new wave of innovation is pushing AI into the physical world through robotics, giving rise to what many are calling Physical AI, machines that can think, move, and adapt in real environments.
Already in 2023, the “AI in robotics” market was worth about USD $12.77 billion, with projections placing it at nearly USD $124.77 billion by 2030.
Meanwhile there were over 4.28 million industrial robots in factories globally by the end of 2023, and companies installed more than 541,000 new robots just that year.
Even more specifically, the broader Physical AI market (including service robots, humanoids, cobots, sensor/control systems) was valued at around USD $3.78 billion in 2024, and analysts expect it to grow to nearly USD $67.91 billion by 2034.
As investors pour billions into the space and startups push prototypes into pilot programs, Physical AI is shaping up to be one of the most transformative industries of the next decade.
Physical AI refers to systems where artificial intelligence is embedded into robots or machines that can interact with the physical world.
Unlike generative AI tools such as ChatGPT, Claude, Gemini, or image generators, Physical AI goes beyond software.
It combines perception, reasoning, and action in real environments.
Think of it as AI leaving the digital realm to take on physical tasks, from lifting packages in a warehouse to assisting doctors in a hospital.
The concept is closely tied to embodied intelligence, where robots learn and adapt through interaction with their surroundings.
Several factors are driving the momentum behind Physical AI:
These forces are making Physical AI not just possible, but commercially viable.
Several startups and tech companies are leading the charge in bringing Physical AI to market.
While better known for electric cars, Tesla has made headlines with its Optimus humanoid robot project. The company envisions a future where robots handle dangerous or repetitive work, complementing Tesla’s broader mission of advancing automation.
These companies represent just a fraction of the activity in the US and EU, where universities, labs, and startups are working to make Physical AI practical and scalable.
Based in California, Figure AI is building humanoid robots designed to perform labor in industries facing worker shortages.
The company has raised hundreds of millions of dollars and attracted partnerships with major corporations. Their vision is to create versatile robots capable of handling tasks across logistics, manufacturing, and beyond.
Oregon-based Agility Robotics is focused on warehouse and logistics automation. Its bipedal robot, Digit, is already being tested in supply chain environments. Unlike stationary robotic arms, Digit can walk, carry, and place items, offering flexibility in dynamic spaces.
Canadian startup Sanctuary AI is developing robots with advanced cognitive abilities. Their approach integrates reasoning, planning, and memory into physical machines, aiming to create robots that can think through problems and not just execute pre-programmed motions.
One short email a week with the best impact reads, interviews, and tools.
Unsubscribe anytime.
Photo credit: Humanity AI modified by Causeartist
Photo credit: Spiro
Stories, interviews, and tools for founders, nonprofits, and investors building a better economy.
Privacy-first updates. Unsubscribe anytime.
