Summary
The panel defined physical AI as systems that perceive real-world data, reason over it, and support action, spanning satellite-derived Earth measurements, health/genomics digital twins, and the compute fabric behind them. Speakers emphasized that value comes from turning petabyte-to-exabyte data into precise decisions while preserving sovereignty, privacy, and cost-efficient processing.
Key Topics
5 key topics from Rebecca Schaevitz, Rob Pegoraro, Maria Dunford and Fabrizio Pirondini at HumanX Amsterdam 2026. Thicker branches were mentioned more often in the talk.
Engineering the Physical World
Click a topic to open it
Key Takeaways
- Model real-world change as measurements, not just raw images.
- Keep sensitive health data sovereign; bring computation to it.
- Design satellites and sensors specifically for AI-ready data streams.
- Plan for exabyte-scale storage, ingestion, and processing costs.
- Use high-radix optical connectivity to flatten AI infrastructure networks.
“all data that can be used to cure and prevent disease is used”
Share
Up Next
E
WatchNext in agenda
Don't Skip the "Safe" Switch
EEdith Harbaugh
ai developmentfeature flags
BRS
WatchAlso on physical ai
Physical AI Meets Real-World Autonomy
Bobby Healy, Radha Basu, Shawn Kerrigan, Éanna Kelly
physical aiautonomous systems
AS
WatchAlso on physical ai
The Race to Build Real-World Intelligence
Alex Kendall, Sharon Goldman
autonomous drivingphysical ai