The big picture: Physical Superintelligence (PSI), an AI-focused physics research lab, has launched with $58 million in seed funding led by Breakthrough Energy Ventures. The company is developing computational models and tools for designing and optimizing physical systems.
Why it matters:
- Industrializing discovery: PSI aims to industrialize the discovery of new physics, tackling problems that have been untouched for decades.
- AI’s scientific impact: AI has the potential to fundamentally change the pace of scientific discovery, with profound implications for physics, energy, computing, and materials.
- Efficiency edge: The company’s initial focus provides customers with a measurable edge in designing and running data centers more efficiently.
How it works:
- Emmy platform: PSI is developing Emmy, a core platform featuring AI-based ‘virtual physicists’ built on its reasoning engine and a curated collection of simulations.
- Data center optimization: Emmy will use physics-based reasoning and simulation to examine complex interactions among power, cooling, networking, and computing systems in terrestrial and orbital data centers.
- Infrastructure design: The platform is designed to optimize infrastructure before construction and to enhance the efficiency of existing facilities.
The catch: The market for AI-driven physics optimization, particularly for specialized applications like orbital data centers, is nascent. PSI will need to demonstrate clear ROI against established engineering simulation tools and potential in-house solutions from large tech companies. Scaling the ambition of ‘industrializing new physics discovery’ beyond initial commercial applications will require significant ongoing R&D and broad market adoption.
Key Facts
- Company: Physical Superintelligence
- Amount: $58M
- Round: Seed
- Investors: Breakthrough Energy Ventures (lead), Dragon Global, Robot Ventures, Solari, Susa, SV Angel, Valkyrie
- Founders: Matt Pines, Alex Klokus, Dr. Alexander Wissner-Gross
- Announced: 2026-09-07
- Sector: AI for Physical Systems Optimization


