News Summary

Axiom, a groundbreaking AI framework developed by Verse AI, offers a novel approach to machine learning by mimicking the way humans understand and interact with the world. Unlike traditional artificial neural networks, Axiom incorporates the free energy principle, a theory formulated by neuroscientist Karl Friston. This approach allows AI agents to learn efficiently from experience with significantly reduced computational power. Axiom has proven capable of mastering simplified video games such as drive, bounce, hunt, and jump, demonstrating its potential in real-time learning applications. The architecture, described as a ‘digital brain,’ is already being explored by a finance company to model market behavior. The framework’s development is influenced by the work of British Canadian computer scientist Geoffrey Hinton, a pioneer in deep learning, revealing an intersection of diverse AI research paths.