Aether AI

About Aether AI

Aether AI builds causal world models—AI systems that learn mechanisms, reason about interventions, and operate more reliably in real-world environments. Instead of only predicting patterns from passive data, the models connect observation, latent state, action, and outcome so systems can evaluate counterfactuals and decide how to intervene. The company also goes by Aether Labs in public materials and is based around San Diego / UC San Diego research roots.

Physical AI and robotics are a primary proving ground: robots need models of contact, force, constraints, and dynamics, not perception alone. Aether positions its work as a decision brain between perception and control—scene understanding becomes physical reasoning, and physical reasoning becomes action. The same causal approach extends to scientific discovery in biology, medicine, and longevity, where teams must distinguish drivers from markers, predict how interventions propagate, and design experiments that separate competing explanations.

Aether serves teams building real-world intelligent systems that need more than pattern recognition. Its business model is frontier AI research and software for causal reasoning in physical and scientific settings, with commercial focus on Physical AI and robotics deployments.