In the late 1950s, two young researchers at Johns Hopkins threaded an electrode into a cat's brain and stumbled onto a discovery that would win them the Nobel Prize: single neurons that fire only for a specific edge of light at a specific angle. This episode traces David Hubel and Torsten Wiesel's work from that lucky afternoon through the mapping of the visual cortex. Along the way we go back into the history of vision research, before following the trail forward into convolutional neural networks and modern computer vision. It's a story about how the brain doesn't just see, it decodes.

We thank the Karolinska institutet for kindly granting permission to use audio recordings of Hubel and Wiesel's Nobel Prize lectures.

The voice used in this podcast was generated using artificial intelligence.

In 1949, a Canadian psychologist with no patience for the strict behaviorism of his day proposed a rule for how connections between neurons change with experience — and used it to imagine how thought could emerge from cell assemblies firing in loops. That idea became "neurons that fire together, wire together," now one of neuroscience's most repeated phrases, even though Donald Hebb himself never wrote it quite that way. This episode follows Hebb from a childhood in Nova Scotia, through failed ambitions as a novelist, into a career that helped found a whole way of thinking about the mind — one whose fingerprints are now all over modern learning algorithms.

Dopamine has long been called the brain's pleasure chemical — the one that surges when you eat sugar or scroll through your phone. That story turns out to be mostly wrong. What dopamine neurons are actually doing is more surprising, and more mathematical. This episode traces how that discovery came together — through a neurophysiologist recording from monkey brains and a pair of computational neuroscientists who had been reading about reinforcement learning. It ends with how the same math is being used to understand depression and addiction.