Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Watch a 49-minute lecture exploring neural machines through the lens of both biological systems and artificial devices. Delve into research presented by Sanjoy Dasgupta from the University of California, San Diego, at IPAM's Naturalistic Approaches to Artificial Intelligence Workshop. Discover how neural machines process inputs and produce outputs in constant time while utilizing fixed-size memory, with a focus on lifelong learning and environmental adaptation. Explore two fascinating case studies: first, examine the fly's olfactory system architecture and its applications in similarity search, novelty detection, and continual learning; second, investigate low-power embedded devices based on hyperdimensional computing principles. Learn how these systems, despite using only basic arithmetic operations, can effectively handle various information processing tasks through innovative data representation methods inspired by biological systems.