Project Prometheus eyes $10B megaround

This looks less like normal startup pricing and more like a sovereign-scale bet on who owns the next industrial AI platform.

• –The valuation is hard to justify on current product evidence because Project Prometheus remains secretive, early, and apparently pre-commercial.
• –The bull case is that physics-aware AI could collapse design-test-manufacture cycles in sectors where even small efficiency gains are worth billions.
• –The bear case is that “physical AI” is becoming the new frontier-label for capital concentration before anyone proves durable technical advantage.
• –The General Agents acquisition suggests Prometheus wants agentic research and engineering workflows, not just robotics demos or simulation tooling.
• –Developers should watch for whether this produces usable APIs, simulation infrastructure, foundation models for physical systems, or just another closed mega-lab.