Learning Science, Learning Magic

This week on Unpopular Neuroscience, Katie and Patrick welcome special guest Matt Baumann with a deep question from his son, Calvin. When a slug remembers something, is it reliving the experience, or just firing a reflex? And what about remembering for humans? Or for robots?

Large attentional transformers like ChatGPT generate text that seems surprisingly human-like. On this episode Katie and Patrick ask which bits of the brain work like large language models, which bits AI still hasn't figured out, and whether the difference can explain why robots can't count the vowels in a strawberry. 

Neuroscientists don't believe in Artificial General Intelligence because they don't believe in General Intelligence. Instead they think the brain has multiple, specialized systems that work together to solve complex problems–more like a swarm of bees than a philosopher king. On this episode Katie and Patrick talk about the hive of brain systems living in your head and why this makes biological intelligence so robust and effective.

What is consciousness? To the general public it's one of the most profound and enduring mysteries of the human psyche. To specialists it's something we don't worry about much. In this episode, Katie and Patrick explore the semantic gap between broad popular conceptions of consciousness and narrower understanding held by neuroscientists and philosophers.

On this episode Katie & Patrick discuss why cognitive biases are so popular—at least when applied to other people—even though these reasoning foibles are not exclusive to humans, probably not caused by the brain, and unlikely to explain much about behavior.

Everyone knows that learning takes hard work, deep reading, study, and repeated practice. But wouldn't it be rad if it didn't? On this episode Katie and Patrick talk about the science of learning and the often more appealing—but less real—magic of learning. 

AI oscillates between impressive breakthroughs and laughable failures. In this episode Katie & Patrick discuss AI booms and winters and what makes the technology so brittle. What can AI learn from neuroscience? 

What's so great about having a "normal" brain? Katie & Patrick discuss a trio of patients from Oliver Sacks' iconic book The Man Who Mistook His Wife For A Hat who have adapted to unusual configurations of their neural hardware. Listen to explore the diverse neuropsychology of the human species.

We sometimes talk about "the brain" as if there's only one of them, but there's tremendous diversity in nervous systems between species, and even between individuals. On this episode of Unpopular Neuroscience Katie and Patrick discuss some of the most interesting brains they've encountered in the neuroscience literature and their own research careers. Learn more about the radically different brains, and surprisingly consistent behavior of jumping spiders, orcas, and french white-collar workers.

How hard is it to build a working, thinking brain from simple parts like wires, sensors, and motors? According to Valentino Braitenberg, author of this week's Book Vehicles, it's actually easier to build a brain than to figure out how one works. This week Katie & Patrick discuss what the weird little robots Braitenberg describes can do and how it can help us think more clearly about real brains.