IXSAR Insights · Jul 21, 2026

The physical-AI supercycle has begun

Capital is rotating from screens to the physical world. We think this is the defining venture cycle of the decade, and we built IXSAR to invest in it.

For two decades, the largest venture returns were built on code. That era is not over, but the marginal dollar of frontier capital is now moving toward systems that perceive, reason, and act in the real world. The signals are hard to miss. As of this writing, robotics foundation-model company Physical Intelligence has reportedly discussed a round that would value it above $11 billion, roughly doubling its valuation in months, while robot-brain developer Skild AI closed a $1.4 billion Series C in January at a valuation above $14 billion, reported as the largest robotics-AI round on record.

Money follows capability, and capability has become physical. This is what we mean by physical AI: intelligent systems that close the loop between perception and action in unstructured environments, not a chatbot in a browser, but a machine on a factory floor, a vehicle on a road, or a satellite in orbit.

Three things changed at once. Large models gave machines comprehension, world models began to solve the physics of action, and vision-language-action (VLA) architectures closed the last mile from understanding to execution. Robots can now be instructed in natural language and carry out tasks they were never explicitly programmed for. Foundation models pre-trained on massive datasets are proving more effective than the brittle, task-specific systems that came before.

Explore

This site requires JavaScript. For inquiries, email partners@ixsar.com (investors) or companies@ixsar.com (founders).