Neural Networks and Thought
Exploring the limitations of simple neural networks reveals their struggle with generating creative thought and handling queries outside their training set. John emphasizes that true understanding and mental exploration require a dynamic system, unlike static models that rely solely on past instances. The conversation highlights the critical difference between human cognition and artificial intelligence, particularly in recognizing unexpected scenarios.In this clip
From this podcast
Lex Fridman Podcast
John Hopfield: Physics View of the Mind and Neurobiology | Lex Fridman Podcast #76
Related Questions