John Hopfield Explained: How Machines Learned to Remember (Nobel Prize 2024)

John Hopfield Explained: How Machines Learned to Remember | Hopfield Network, Nobel Prize in Physics 2024, Associative Memory, AI Explained. Discover how a physicist taught machines to remember, and why your own brain remembers by falling, not by looking things up. In this video, we explore the idea that earned John Hopfield the 2024 Nobel Prize in Physics, shared with Geoffrey Hinton. How can a crowd of simple, mindless neurons store memories with no librarian, no address and no filing cabinet? And why does the same mechanism make you see faces in clouds? Inside this video: Why memory is content-addressable, not stored at an address How a physicist borrowed ideas from magnets and emergence The energy landscape: memories as valleys a network falls into Hebb’s rule: neurons that fire together wire together Why a Hopfield network must always settle (the 1982 proof) False memories, spurious states and the 0.14N capacity limit What this model explains, and what it doesn’t From the Boltzmann machine to modern Hopfield networks and transformer attention The unsolved question: why does recognizing something feel like anything? Whether you’re curious about deep learning, neuroscience, the history of AI or the physics of the mind, this is the story behind the foundations of modern artificial intelligence. Part of our series on the scientists whose Nobel and Turing Award-winning ideas shaped AI and computing. Next up: Geoffrey Hinton and how machines learned to learn. 👍 Like, subscribe and turn on notifications for more. Disclosure: The narration is written in the style of John Hopfield and voiced by an AI avatar. It is a dramatization and does not contain his actual words. #JohnHopfield #HopfieldNetwork #nobelprize #physics 2024, how #machine remember, associative memory, #deeplearning explained, #geoffreyhinton #historyofai how the brain stores memory, #neuralnetworks #ai