Physical AI is attracting a surge of investment as AI moves from screens into robots, drones and industrial machines. Investor Anton Alikov argues that the opportunity comes with a less fashionable constraint: hardware, testing and factory integration move at the speed of the physical world, not a software update.
Watch Desk analysis
What happened
In a Forbes Councils essay published on 8 October, Arctic Ventures founder Anton Alikov says startups in physical AI raised $47.4 billion across 521 deals in the first half of 2026, citing Crunchbase data. He compares that with $12 billion across 470 deals in the second half of 2025. Those are figures cited in Alikov’s essay, not independently established market outcomes here.
Alikov describes physical AI as systems that perceive the real world, model it and take action through machines such as robots and autonomous vehicles. He argues that investment is shifting towards businesses combining software with hardware, data and industrial integration. His essay also cites PwC forecasts that the physical AI market could reach about $450 billion by 2030, or up to $1 trillion as a broader ecosystem. Those are forecasts, not receipts from the future.
Why it matters
The investment case is that robotics and industrial AI are harder to copy than software alone: they require specialised hardware, real-world data, testing, certification and integration into existing operations. Those barriers could make successful deployments defensible, but they also make expansion slower and more capital-intensive.
Alikov’s own caution is the useful counterweight. Robot fleets need maintenance and logistics, training data can be costly and scarce, and changing an industrial process may take years. A promising model is only one part of a system that must work around actual machines, staff and production lines.
Our read
The physical-AI boom is a credible investment thesis, not proof that every robot company is about to mint money. The practical test is whether a system can move beyond a convincing demonstration into reliable, repeatable work at a cost customers will accept. Software can pivot in a week; factories have rather more bolts to undo.
What to watch
- Whether announced investment turns into deployed systems and repeat customer use.
- How companies acquire the real-world data and testing needed to make robots dependable.
- Whether hardware, integration and maintenance costs leave room for sustainable margins.
Discussion spark: When should investors treat a physical-AI company as more than a promising demo: after a successful pilot, or only once customers are using it repeatedly?
Sources and evidence
- From Bits To Molecules: Considerations For Investors About Physical AI – Forbes (8 October 2026, 11:30 UTC)
- How To Design Websites And Apps For The Age Of AI Agents – Forbes (8 October 2026, 11:30 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.