Ex-Google veteran's startup claims to have built world-first AGI model

Integral AI, a Tokyo-based startup founded by ex-Google veteran Jad Tarifi, claims to have developed the world’s first Artificial General Intelligence (AGI) model. According to the company, their AI system can autonomously learn new skills in novel domains without relying on pre-existing datasets or human intervention. Integral AI defines AGI by three core criteria: autonomous skill learning, safe and reliable mastery without catastrophic risks, and energy efficiency comparable to human learning. The company asserts that these principles guided the development and testing of their AGI system, which has reportedly been trialed successfully in robotic applications. Jad Tarifi, who left Google after a decade to start Integral AI in Japan due to its robotics leadership, describes this achievement as a “fundamental leap” in AI technology and a significant milestone for human civilization. The startup claims their system mimics the multi-layered neocortex of the human brain, responsible for conscious thought and perception, aiming to scale toward embodied superintelligence that enhances freedom