Team led by ex-Googlers generative AI pioneers, Jad Tarifi & Nima Asgharbeygi, create an interactive sandbox showcasing the world’s first AGI-capable model

TOKYO: #AGI--Integral AI, a global leader in the development of embodied AGI, today announced the successful testing of the world’s first AGI-capable model. Led by AGI expert and ex-Google AI veteran, Jad Tarifi, company engineers have successfully developed a system that autonomously learns new skills safely, efficiently, and reliably. Thus, marking a fundamental leap beyond the limits of current AI technologies and establishing a scalable foundation for self-improving generality intelligence moving towards superintelligence.

During the research stage, Integral AI identified clearly defined parameters to effectively measure and determine the arrival of AGI by using three fundamental qualifiers: Autonomous Skill Learning defined as a system that must teach itself entirely new skills in novel domains without pre-existing datasets or human intervention, Safe and Reliable Mastery defined as learning that must occur without catastrophic risks or unintended side effects, and Energy Efficiency defined as the total energy cost of system learning being equal to or lower than that of a human acquiring the same skill. For Integral AI engineers, these principles served as fundamental cornerstones and developmental benchmarks during the inception and testing of this first-in-its-class AGI learning system.

Mirroring the multi-layered neocortex that underpins human thought, Integral AI’s model architecture grows, abstracts, plans, and acts as a unified system. Model testing revealed unprecedented adaptability, particularly in autonomous robotics, where robots were observed acquiring new skills in real-world environments without human supervision.

“Today’s announcement is more than just a technical achievement, it marks the next chapter in the story of human civilization,” said Jad Tarifi, Ph.D., CEO and co-founder, Integral AI. “Our mission now is to scale this AGI-capable model, still in its infancy, toward embodied superintelligence that expands freedom and collective agency.”

Learn more at: https://www.integral.ai/agi

About Integral AI

Integral AI builds embodied artificial general intelligence through new foundation model architectures that learn safely, efficiently, and autonomously. By reimagining intelligence as an interactive, world-modeling process, Integral AI is developing scalable infrastructure for embodied superintelligence. Founded in 2021 by Dr. Jad Tarifi and Nima Asgharbeygi, former Google AI pioneers with decades of experience in large-scale AI systems, Integral AI’s mission is to give humankind a true magic wand.

Fonte: Business Wire