Had @ericschmidt on the pod to discuss the superintelligence timeline...some key points: -- mutually assured AI destruction is probably the only path to a "safe" AGI world -- superintelligence within 10 years (his prediction) -- china is catching up faster than expected -- AI will generate its own scaffolding in 2025, but the real concern is self-improvement -- 10 new Google-scale companies incoming

full ep--