99 avsnitt • Längd: 90 min • Månadsvis
Neuroscience and artificial intelligence work better together. Brain inspired is a celebration and exploration of the ideas driving our progress to understand intelligence. I interview experts about their work at the interface of neuroscience, artificial intelligence, cognitive science, philosophy, psychology, and more: the symbiosis of these overlapping fields, how they inform each other, where they differ, what the past brought us, and what the future brings. Topics include computational neuroscience, supervised machine learning, unsupervised learning, reinforcement learning, deep learning, convolutional and recurrent neural networks, decision-making science, AI agents, backpropagation, credit assignment, neuroengineering, neuromorphics, emergence, philosophy of mind, consciousness, general AI, spiking neural networks, data science, and a lot more. The podcast is not produced for a general audience. Instead, it aims to educate, challenge, inspire, and hopefully entertain those interested in learning more about neuroscience and AI.
The podcast Brain Inspired is created by Paul Middlebrooks. The podcast and the artwork on this page are embedded on this page using the public podcast feed (RSS).
Support the show to get full episodes, full archive, and join the Discord community.
Today I'm in conversation with Rajesh Rao, a distinguished professor of computer science and engineering at the University of Washington, where he also co-directs the Center for Neurotechnology. Back in 1999, Raj and Dana Ballard published what became quite a famous paper, which proposed how predictive coding might be implemented in brains. What is predictive coding, you may be wondering? It's roughly the idea that your brain is constantly predicting incoming sensory signals, and it generates that prediction as a top-down signal that meets the bottom-up sensory signals. Then the brain computes a difference between the prediction and the actual sensory input, and that difference is sent back up to the "top" where the brain then updates its internal model to make better future predictions.
So that was 25 years ago, and it was focused on how the brain handles sensory information. But Raj just recently published an update to the predictive coding framework, one that incorporates actions and perception, suggests how it might be implemented in the cortex - specifically which cortical layers do what - something he calls "Active predictive coding." So we discuss that new proposal, we also talk about his engineering work on brain-computer interface technologies, like BrainNet, which basically connects two brains together, and like neural co-processors, which use an artificial neural network as a prosthetic that can do things like enhance memories, optimize learning, and help restore brain function after strokes, for example. Finally, we discuss Raj's interest and work on deciphering an ancient Indian text, the mysterious Indus script.
Read the transcript.
0:00 - Intro 7:40 - Predictive coding origins 16:14 - Early appreciation of recurrence 17:08 - Prediction as a general theory of the brain 18:38 - Rao and Ballard 1999 26:32 - Prediction as a general theory of the brain 33:24 - Perception vs action 33:28 - Active predictive coding 45:04 - Evolving to augment our brains 53:03 - BrainNet 57:12 - Neural co-processors 1:11:19 - Decoding the Indus Script 1:20:18 - Transformer models relation to active predictive coding
Support the show to get full episodes, full archive, and join the Discord community.
Joe Monaco and Grace Hwang co-organized a recent workshop I participated in, the 2024 BRAIN NeuroAI Workshop. You may have heard of the BRAIN Initiative, but in case not, BRAIN is is huge funding effort across many agencies, one of which is the National Institutes of Health, where this recent workshop was held. The BRAIN Initiative began in 2013 under the Obama administration, with the goal to support developing technologies to help understand the human brain, so we can cure brain based diseases.
BRAIN Initiative just became a decade old, with many successes like recent whole brain connectomes, and discovering the vast array of cell types. Now the question is how to move forward, and one area they are curious about, that perhaps has a lot of potential to support their mission, is the recent convergence of neuroscience and AI... or NeuroAI. The workshop was designed to explore how NeuroAI might contribute moving forward, and to hear from NeuroAI folks how they envision the field moving forward. You'll hear more about that in a moment.
That's one reason I invited Grace and Joe on. Another reason is because they co-wrote a position paper a while back that is impressive as a synthesis of lots of cognitive sciences concepts, but also proposes a specific level of abstraction and scale in brain processes that may serve as a base layer for computation. The paper is called Neurodynamical Computing at the Information Boundaries, of Intelligent Systems, and you'll learn more about that in this episode.
Read the transcript.
0:00 - Intro 25:45 - NeuroAI Workshop - neuromorphics 33:31 - Neuromorphics and theory 49:19 - Reflections on the workshop 54:22 - Neurodynamical computing and information boundaries 1:01:04 - Perceptual control theory 1:08:56 - Digital twins and neural foundation models 1:14:02 - Base layer of computation
Support the show to get full episodes, full archive, and join the Discord community.
The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
Read more about our partnership.
Sign up for the “Brain Inspired” email alerts to be notified every time a new “Brain Inspired” episode is released: https://www.thetransmitter.org/newsletters/
To explore more neuroscience news and perspectives, visit thetransmitter.org.
Hessam Akhlaghpour is a postdoctoral researcher at Rockefeller University in the Maimon lab. His experimental work is in fly neuroscience mostly studying spatial memories in fruit flies. However, we are going to be talking about a different (although somewhat related) side of his postdoctoral research. This aspect of his work involves theoretical explorations of molecular computation, which are deeply inspired by Randy Gallistel and Adam King's book Memory and the Computational Brain. Randy has been on the podcast before to discuss his ideas that memory needs to be stored in something more stable than the synapses between neurons, and how that something could be genetic material like RNA. When Hessam read this book, he was re-inspired to think of the brain the way he used to think of it before experimental neuroscience challenged his views. It re-inspired him to think of the brain as a computational system. But it also led to what we discuss today, the idea that RNA has the capacity for universal computation, and Hessam's development of how that might happen. So we discuss that background and story, why universal computation has been discovered in organisms yet since surely evolution has stumbled upon it, and how RNA might and combinatory logic could implement universal computation in nature.
Read the transcript.
0:00 - Intro 4:44 - Hessam's background 11:50 - Randy Gallistel's book 14:43 - Information in the brain 17:51 - Hessam's turn to universal computation 35:30 - AI and universal computation 40:09 - Universal computation to solve intelligence 44:22 - Connecting sub and super molecular 50:10 - Junk DNA 56:42 - Genetic material for coding 1:06:37 - RNA and combinatory logic 1:35:14 - Outlook 1:42:11 - Reflecting on the molecular world
Support the show to get full episodes, full archive, and join the Discord community.
The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
Read more about our partnership.
Sign up for the “Brain Inspired” email alerts to be notified every time a new “Brain Inspired” episode is released: https://www.thetransmitter.org/newsletters/
To explore more neuroscience news and perspectives, visit thetransmitter.org.
Tony Zador runs the Zador lab at Cold Spring Harbor Laboratory. You've heard him on Brain Inspired a few times in the past, most recently in a panel discussion I moderated at this past COSYNE conference - a conference Tony co-founded 20 years ago. As you'll hear, Tony's current and past interests and research endeavors are of a wide variety, but today we focus mostly on his thoughts on NeuroAI.
We're in a huge AI hype cycle right now, for good reason, and there's a lot of talk in the neuroscience world about whether neuroscience has anything of value to provide AI engineers - and how much value, if any, neuroscience has provided in the past.
Tony is team neuroscience. You'll hear him discuss why in this episode, especially when it comes to ways in which development and evolution might inspire better data efficiency, looking to animals in general to understand how they coordinate numerous objective functions to achieve their intelligent behaviors - something Tony calls alignment - and using spikes in AI models to increase energy efficiency.
Read the transcript.
0:00 - Intro 3:28 - "Neuro-AI" 12:48 - Visual cognition history 18:24 - Information theory in neuroscience 20:47 - Necessary steps for progress 24:34 - Neuro-AI models and cognition 35:47 - Animals for inspiring AI 41:48 - What we want AI to do 46:01 - Development and AI 59:03 - Robots 1:25:10 - Catalyzing the next generation of AI
Support the show to get full episodes, full archive, and join the Discord community.
The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
Read more about our partnership.
Sign up for the “Brain Inspired” email alerts to be notified every time a new “Brain Inspired” episode is released.
To explore more neuroscience news and perspectives, visit thetransmitter.org.
Karen Adolph runs the Infant Action Lab at NYU, where she studies how our motor behaviors develop from infancy onward. We discuss how observing babies at different stages of development illuminates how movement and cognition develop in humans, how variability and embodiment are key to that development, and the importance of studying behavior in real-world settings as opposed to restricted laboratory settings. We also explore how these principles and simulations can inspire advances in intelligent robots. Karen has a long-standing interest in ecological psychology, and she shares some stories of her time studying under Eleanor Gibson and other mentors.
Finally, we get a surprise visit from her partner Mark Blumberg, with whom she co-authored an opinion piece arguing that "motor cortex" doesn't start off with a motor function, oddly enough, but instead processes sensory information during the first period of animals' lives.
Read the transcript.
Support the show to get full episodes, full archive, and join the Discord community.
The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
This is the second conversation I had while teamed up with Gaute Einevoll at a workshop on NeuroAI in Norway. In this episode, Gaute and I are joined by Cristina Savin and Tim Vogels. Cristina shares how her lab uses recurrent neural networks to study learning, while Tim talks about his long-standing research on synaptic plasticity and how AI tools are now helping to explore the vast space of possible plasticity rules.
We touch on how deep learning has changed the landscape, enhancing our research but also creating challenges with the "fashion-driven" nature of science today. We also reflect on how these new tools have changed the way we think about brain function without fundamentally altering the structure of our questions.
Be sure to check out Gaute's Theoretical Neuroscience podcast as well!
Read the transcript, provided by The Transmitter.
Support the show to get full episodes, full archive, and join the Discord community.
The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
This is the first of two less usual episodes. I was recently in Norway at a NeuroAI workshop called Validating models: How would success in NeuroAI look like? What follows are a few recordings I made with my friend Gaute Einevoll. Gaute has been on this podcast before, but more importantly he started his own podcast a while back called Theoretical Neuroscience, which you should check out.
Gaute and I introduce the episode, then briefly speak with Mikkel Lepperød, one of the organizers of the workshop. In this first episode, we're then joined by Ken Harris and Andreas Tolias to discuss how AI has influenced their research, thoughts about brains and minds, and progress and productivity.
Read the transcript, provided by The Transmitter.
Support the show to get full episodes, full archive, and join the Discord community.
https://youtu.be/lbKEOdbeqHoThe Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
The Transmitter has provided a transcript for this episode.
Vijay Namoodiri runs the Nam Lab at the University of California San Francisco, and Ali Mojebi is an assistant professor at the University of Wisconsin-Madison. Ali as been on the podcast before a few times, and he's interested in how neuromodulators like dopamine affect our cognition. And it was Ali who pointed me to Vijay, because of some recent work Vijay has done reassessing how dopamine might function differently than what has become the classic story of dopamine's function as it pertains to learning. The classic story is that dopamine is related to reward prediction errors. That is, dopamine is modulated when you expect reward and don't get it, and/or when you don't expect reward but do get it. Vijay calls this a "prospective" account of dopamine function, since it requires an animal to look into the future to expect a reward. Vijay has shown, however, that a retrospective account of dopamine might better explain lots of know behavioral data. This retrospective account links dopamine to how we understand causes and effects in our ongoing behavior. So in this episode, Vijay gives us a history lesson about dopamine, his newer story and why it has caused a bit of controversy, and how all of this came to be.
I happened to be looking at the Transmitter the other day, after I recorded this episode, and low and behold, there was an article titles Reconstructing dopamine’s link to reward. Vijay is featured in the article among a handful of other thoughtful researchers who share their work and ideas about this very topic. Vijay wrote his own piece as well: Dopamine and the need for alternative theories. So check out those articles for more views on how the field is reconsidering how dopamine works.
Read the transcript, produced by The Transmitter.
0:00 - Intro 3:42 - Dopamine: the history of theories 32:54 - Importance of learning and behavior studies 39:12 - Dopamine and causality 1:06:45 - Controversy over Vijay's findings
Support the show to get full episodes, full archive, and join the Discord community.
The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists.
Read more about our partnership.
Check out this story: Monkeys build mental maps to navigate new tasks
Sign up for “Brain Inspired” email alerts to be notified every time a new “Brain Inspired” episode is released.
To explore more neuroscience news and perspectives, visit thetransmitter.org.
Kim Stachenfeld embodies the original core focus of this podcast, the exploration of the intersection between neuroscience and AI, now commonly known as Neuro-AI. That's because she walks both lines. Kim is a Senior Research Scientist at Google DeepMind, the AI company that sprang from neuroscience principles, and also does research at the Center for Theoretical Neuroscience at Columbia University. She's been using her expertise in modeling, and reinforcement learning, and cognitive maps, for example, to help understand brains and to help improve AI. I've been wanting to have her on for a long time to get her broad perspective on AI and neuroscience.
We discuss the relative roles of industry and academia in pursuing various objectives related to understanding and building cognitive entities
She's studied the hippocampus in her research on reinforcement learning and cognitive maps, so we discuss what the heck the hippocampus does since it seems to implicated in so many functions, and how she thinks of reinforcement learning these days.
Most recently Kim at Deepmind has focused on more practical engineering questions, using deep learning models to predict things like chaotic turbulent flows, and even to help design things like bridges and airplanes. And we don't get into the specifics of that work, but, given that I just spoke with Damian Kelty-Stephen, who thinks of brains partially as turbulent cascades, Kim and I discuss how her work on modeling turbulence has shaped her thoughts about brains.
Check out the transcript, provided by The Transmitter.
0:00 - Intro 4:31 - Deepmind's original and current vision 9:53 - AI as tools and models 12:53 - Has AI hindered neuroscience? 17:05 - Deepmind vs academic work balance 20:47 - Is industry better suited to understand brains? 24?42 - Trajectory of Deepmind 27:41 - Kim's trajectory 33:35 - Is the brain a ML entity? 36:12 - Hippocampus 44:12 - Reinforcement learning 51:32 - What does neuroscience need more and less of? 1:02:53 - Neuroscience in a weird place? 1:06:41 - How Kim's questions have changed 1:16:31 - Intelligence and LLMs 1:25:34 - Challenges
Support the show to get full episodes, full archive, and join the Discord community.
Àlex Gómez-Marín heads The Behavior of Organisms Laboratory at the Institute of Neuroscience in Alicante, Spain. He's one of those theoretical physicist turned neuroscientist, and he has studied a wide range of topics over his career. Most recently, he has become interested in what he calls the "edges of consciousness", which encompasses the many trying to explain what may be happening when we have experiences outside our normal everyday experiences. For example, when we are under the influence of hallucinogens, when have near-death experiences (as Alex has), paranormal experiences, and so on.
So we discuss what led up to his interests in these edges of consciousness, how he now thinks about consciousness and doing science in general, how important it is to make room for all possible explanations of phenomena, and to leave our metaphysics open all the while.
0:00 - Intro 4:13 - Evolving viewpoints 10:05 - Near-death experience 18:30 - Mechanistic neuroscience vs. the rest 22:46 - Are you doing science? 33:46 - Where is my. mind? 44:55 - Productive vs. permissive brain 59:30 - Panpsychism 1:07:58 - Materialism 1:10:38 - How to choose what to do 1:16:54 - Fruit flies 1:19:52 - AI and the Singularity
Support the show to get full episodes, full archive, and join the Discord community.
Damian Kelty-Stephen is an experimental psychologist at State University of New York at New Paltz. Last episode with Luis Favela, we discussed many of the ideas from ecological psychology, and how Louie is trying to reconcile those principles with those of neuroscience. In this episode, Damian and I in some ways continue that discussion, because Damian is also interested in unifying principles of ecological psychology and neuroscience. However, he is approaching it from a different perspective that Louie. What drew me originally to Damian was a paper he put together with a bunch of authors offering their own alternatives to the computer metaphor of the brain, which has come to dominate neuroscience. And we discuss that some, and I'll link to the paper in the show notes. But mostly we discuss Damian's work studying the fractal structure of our behaviors, connecting that structure across scales, and linking it to how our brains and bodies interact to produce our behaviors. Along the way, we talk about his interests in cascades dynamics and turbulence to also explain our intelligence and behaviors. So, I hope you enjoy this alternative slice into thinking about how we think and move in our bodies and in the world.
0:00 - Intro 2:34 - Damian's background 9:02 - Brains 12:56 - Do neuroscientists have it all wrong? 16:56 - Fractals everywhere 28:01 - Fractality, causality, and cascades 32:01 - Cascade instability as a metaphor for the brain 40:43 - Damian's worldview 46:09 - What is AI missing? 54:26 - Turbulence 1:01:02 - Intelligence without fractals? Multifractality 1:10:28 - Ergodicity 1:19:16 - Fractality, intelligence, life 1:23:24 - What's exciting, changing viewpoints
Support the show to get full episodes, full archive, and join the Discord community.
Luis Favela is an Associate Professor at Indiana University Bloomington. He is part philosopher, part cognitive scientist, part many things, and on this episode we discuss his new book, The Ecological Brain: Unifying the Sciences of Brain, Body, and Environment.
In the book, Louie presents his NeuroEcological Nexus Theory, or NExT, which, as the subtitle says, proposes a way forward to tie together our brains, our bodies, and the environment; namely it has a lot to do with the complexity sciences and manifolds, which we discuss. But the book doesn't just present his theory. Among other things, it presents a rich historical look into why ecological psychology and neuroscience haven't been exactly friendly over the years, in terms of how to explain our behaviors, the role of brains in those explanations, how to think about what minds are, and so on. And it suggests how the two fields can get over their differences and be friends moving forward. And I'll just say, it's written in a very accessible manner, gently guiding the reader through many of the core concepts and science that have shaped ecological psychology and neuroscience, and for that reason alone I highly it.
Ok, so we discuss a bunch of topics in the book, how Louie thinks, and Louie gives us some great background and historical lessons along the way.
0:00 - Intro 7:05 - Louie's target with NEXT 20:37 - Ecological psychology and grid cells 22:06 - Why irreconcilable? 28:59 - Why hasn't ecological psychology evolved more? 47:13 - NExT 49:10 - Hypothesis 1 55:45 - Hypothesis 2 1:02:55 - Artificial intelligence and ecological psychology 1:16:33 - Manifolds 1:31:20 - Hypothesis 4: Body, low-D, Synergies 1:35:53 - Hypothesis 5: Mind emerges 1:36:23 - Hypothesis 6:
Support the show to get full episodes, full archive, and join the Discord community.
Jovo, as you'll learn, is theoretically oriented, and enjoys the formalism of mathematics to approach questions that begin with a sense of wonder. So after I learn more about his overall approach, the first topic we discuss is the world's currently largest map of an entire brain... the connectome of an insect, the fruit fly. We talk about his role in this collaborative effort, what the heck a connectome is, why it's useful and what to do with it, and so on.
The second main topic we discuss is his theoretical work on what his team has called prospective learning. Prospective learning differs in a fundamental way from the vast majority of AI these days, which they call retrospective learning. So we discuss what prospective learning is, and how it may improve AI moving forward.
At some point there's a little audio/video sync issues crop up, so we switched to another recording method and fixed it... so just hang tight if you're viewing the podcast... it'll get better soon.
0:00 - Intro 05:25 - Jovo's approach 13:10 - Connectome of a fruit fly 26:39 - What to do with a connectome 37:04 - How important is a connectome? 51:48 - Prospective learning 1:15:20 - Efficiency 1:17:38 - AI doomerism
Support the show to get full episodes, full archive, and join the Discord community.
Jolande Fooken is a post-postdoctoral researcher interested in how we move our eyes and move our hands together to accomplish naturalistic tasks. Hand-eye coordination is one of those things that sounds simple and we do it all the time to make meals for our children day in, and day out, and day in, and day out. But it becomes way less seemingly simple as soon as you learn how we make various kinds of eye movements, and how we make various kinds of hand movements, and use various strategies to do repeated tasks. And like everything in the brain sciences, it's something we don't have a perfect story for yet. So, Jolande and I discuss her work, and thoughts, and ideas around those and related topics.
0:00 - Intro 3:27 - Eye movements 8:53 - Hand-eye coordination 9:30 - Hand-eye coordination and naturalistic tasks 26:45 - Levels of expertise 34:02 - Yarbus and eye movements 42:13 - Varieties of experimental paradigms, varieties of viewing the brain 52:46 - Career vision 1:04:07 - Evolving view about the brain 1:10:49 - Coordination, robots, and AI
Support the show to get full episodes, full archive, and join the Discord community.
Recently I was invited to moderate a panel at the annual Computational and Systems Neuroscience, or COSYNE, conference. This year was the 20th anniversary of COSYNE, and we were in Lisbon Porturgal. The panel goal was to discuss the relationship between neuroscience and AI. The panelists were Tony Zador, Alex Pouget, Blaise Aguera y Arcas, Kim Stachenfeld, Jonathan Pillow, and Eva Dyer. And I'll let them introduce themselves soon. Two of the panelists, Tony and Alex, co-founded COSYNE those 20 years ago, and they continue to have different views about the neuro-AI relationship. Tony has been on the podcast before and will return soon, and I'll also have Kim Stachenfeld on in a couple episodes. I think this was a fun discussion, and I hope you enjoy it. There's plenty of back and forth, a wide range of opinions, and some criticism from one of the audience questioners. This is an edited audio version, to remove long dead space and such. There's about 30 minutes of just panel, then the panel starts fielding questions from the audience.
Support the show to get full episodes, full archive, and join the Discord community.
Mazviita Chirimuuta is a philosopher at the University of Edinburgh. Today we discuss topics from her new book, The Brain Abstracted: Simplification in the History and Philosophy of Neuroscience.
She largely argues that when we try to understand something complex, like the brain, using models, and math, and analogies, for example - we should keep in mind these are all ways of simplifying and abstracting away details to give us something we actually can understand. And, when we do science, every tool we use and perspective we bring, every way we try to attack a problem, these are all both necessary to do the science and limit the interpretation we can claim from our results. She does all this and more by exploring many topics in neuroscience and philosophy throughout the book, many of which we discuss today.
0:00 - Intro 5:28 - Neuroscience to philosophy 13:39 - Big themes of the book 27:44 - Simplifying by mathematics 32:19 - Simplifying by reduction 42:55 - Simplification by analogy 46:33 - Technology precedes science 55:04 - Theory, technology, and understanding 58:04 - Cross-disciplinary progress 58:45 - Complex vs. simple(r) systems 1:08:07 - Is science bound to study stability? 1:13:20 - 4E for philosophy but not neuroscience? 1:28:50 - ANNs as models 1:38:38 - Study of mind
Support the show to get full episodes, full archive, and join the Discord community.
As some of you know, I recently got back into the research world, and in particular I work in Eric Yttris' lab at Carnegie Mellon University.
Eric's lab studies the relationship between various kinds of behaviors and the neural activity in a few areas known to be involved in enacting and shaping those behaviors, namely the motor cortex and basal ganglia. And study that, he uses tools like optogentics, neuronal recordings, and stimulations, while mice perform certain tasks, or, in my case, while they freely behave wandering around an enclosed space.
We talk about how Eric got here, how and why the motor cortex and basal ganglia are still mysteries despite lots of theories and experimental work, Eric's work on trying to solve those mysteries using both trained tasks and more naturalistic behavior. We talk about the valid question, "What is a behavior?", and lots more.
0:00 - Intro 2:36 - Eric's background 14:47 - Different animal models 17:59 - ANNs as models for animal brains 24:34 - Main question 25:43 - How circuits produce appropriate behaviors 26:10 - Cerebellum 27:49 - What do motor cortex and basal ganglia do? 49:12 - Neuroethology 1:06:09 - What is a behavior? 1:11:18 - Categorize behavior (B-SOiD) 1:22:01 - Real behavior vs. ANNs 1:33:09 - Best era in neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Peter Stratton is a research scientist at Queensland University of Technology.
I was pointed toward Pete by a patreon supporter, who sent me a sort of perspective piece Pete wrote that is the main focus of our conversation, although we also talk about some of his work in particular - for example, he works with spiking neural networks, like my last guest, Dan Goodman.
What Pete argues for is what he calls a sideways-in approach. So a bottom-up approach is to build things like we find them in the brain, put them together, and voila, we'll get cognition. A top-down approach, the current approach in AI, is to train a system to perform a task, give it some algorithms to run, and fiddle with the architecture and lower level details until you pass your favorite benchmark test. Pete is focused more on the principles of computation brains employ that current AI doesn't. If you're familiar with David Marr, this is akin to his so-called "algorithmic level", but it's between that and the "implementation level", I'd say. Because Pete is focused on the synthesis of different kinds of brain operations - how they intermingle to perform computations and produce emergent properties. So he thinks more like a systems neuroscientist in that respect. Figuring that out is figuring out how to make better AI, Pete says. So we discuss a handful of those principles, all through the lens of how challenging a task it is to synthesize multiple principles into a coherent functioning whole (as opposed to a collection of parts). Buy, hey, evolution did it, so I'm sure we can, too, right?
0:00 - Intro 3:50 - AI background, neuroscience principles 8:00 - Overall view of modern AI 14:14 - Moravec's paradox and robotics 20:50 -Understanding movement to understand cognition 30:01 - How close are we to understanding brains/minds? 32:17 - Pete's goal 34:43 - Principles from neuroscience to build AI 42:39 - Levels of abstraction and implementation 49:57 - Mental disorders and robustness 55:58 - Function vs. implementation 1:04:04 - Spiking networks 1:07:57 - The roadmap 1:19:10 - AGI 1:23:48 - The terms AGI and AI 1:26:12 - Consciousness
Support the show to get full episodes, full archive, and join the Discord community.
You may know my guest as the co-founder of Neuromatch, the excellent online computational neuroscience academy, or as the creator of the Brian spiking neural network simulator, which is freely available. I know him as a spiking neural network practitioner extraordinaire. Dan Goodman runs the Neural Reckoning Group at Imperial College London, where they use spiking neural networks to figure out how biological and artificial brains reckon, or compute.
All of the current AI we use to do all the impressive things we do, essentially all of it, is built on artificial neural networks. Notice the word "neural" there. That word is meant to communicate that these artificial networks do stuff the way our brains do stuff. And indeed, if you take a few steps back, spin around 10 times, take a few shots of whiskey, and squint hard enough, there is a passing resemblance. One thing you'll probably still notice, in your drunken stupor, is that, among the thousand ways ANNs differ from brains, is that they don't use action potentials, or spikes. From the perspective of neuroscience, that can seem mighty curious. Because, for decades now, neuroscience has focused on spikes as the things that make our cognition tick.
We count them and compare them in different conditions, and generally put a lot of stock in their usefulness in brains.
So what does it mean that modern neural networks disregard spiking altogether?
Maybe spiking really isn't important to process and transmit information as well as our brains do. Or maybe spiking is one among many ways for intelligent systems to function well. Dan shares some of what he's learned and how he thinks about spiking and SNNs and a host of other topics.
0:00 - Intro 3:47 - Why spiking neural networks, and a mathematical background 13:16 - Efficiency 17:36 - Machine learning for neuroscience 19:38 - Why not jump ship from SNNs? 23:35 - Hard and easy tasks 29:20 - How brains and nets learn 32:50 - Exploratory vs. theory-driven science 37:32 - Static vs. dynamic 39:06 - Heterogeneity 46:01 - Unifying principles vs. a hodgepodge 50:37 - Sparsity 58:05 - Specialization and modularity 1:00:51 - Naturalistic experiments 1:03:41 - Projects for SNN research 1:05:09 - The right level of abstraction 1:07:58 - Obstacles to progress 1:12:30 - Levels of explanation 1:14:51 - What has AI taught neuroscience? 1:22:06 - How has neuroscience helped AI?
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
John Krakauer has been on the podcast multiple times (see links below). Today we discuss some topics framed around what he's been working on and thinking about lately. Things like
Relevant links:
Time stamps 0:00 - Intro 2:07 - It's a podcast episode! 6:47 - Stroke and Sherrington neuroscience 19:26 - Thinking vs. moving, representations 34:15 - What's special about humans? 56:35 - Does cortical reorganization happen? 1:14:08 - Current era in neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
By day, Max Bennett is an entrepreneur. He has cofounded and CEO'd multiple AI and technology companies. By many other countless hours, he has studied brain related sciences. Those long hours of research have payed off in the form of this book, A Brief History of Intelligence: Evolution, AI, and the Five Breakthroughs That Made Our Brains.
Three lines of research formed the basis for how Max synthesized knowledge into the ideas in his current book: findings from comparative psychology (comparing brains and minds of different species), evolutionary neuroscience (how brains have evolved), and artificial intelligence, especially the algorithms developed to carry out functions. We go through I think all five of the breakthroughs in some capacity. A recurring theme is that each breakthrough may explain multiple new abilities. For example, the evolution of the neocortex may have endowed early mammals with the ability to simulate or imagine what isn't immediately present, and this ability might further explain mammals' capacity to engage in vicarious trial and error (imagining possible actions before trying them out), the capacity to engage in counterfactual learning (what would have happened if things went differently than they did), and the capacity for episodic memory and imagination.
The book is filled with unifying accounts like that, and it makes for a great read. Strap in, because Max gives a sort of masterclass about many of the ideas in his book.
0:00 - Intro 5:26 - Why evolution is important 7:22 - Maclean's triune brain 14:59 - Breakthrough 1: Steering 29:06 - Fish intelligence 40:38 - Breakthrough 3: Mentalizing 52:44 - How could we improve the human brain? 1:00:44 - What is intelligence? 1:13:50 - Breakthrough 5: Speaking
Support the show to get full episodes, full archive, and join the Discord community.
Welcome to another special panel discussion episode.
I was recently invited to moderate at discussion amongst 6 people at the annual Aspirational Neuroscience meetup. Aspirational Neuroscience is a nonprofit community run by Kenneth Hayworth. Ken has been on the podcast before on episode 103. Ken helps me introduce the meetup and panel discussion for a few minutes. The goal in general was to discuss how current and developing neuroscience technologies might be used to decode a nontrivial memory from a static connectome - what the obstacles are, how to surmount those obstacles, and so on.
There isn't video of the event, just audio, and because we were all sharing microphones and they were being passed around, you'll hear some microphone type noise along the way - but I did my best to optimize the audio quality, and it turned out mostly quite listenable I believe.
0:00 - Intro 1:45 - Ken Hayworth 14:09 - Panel Discussion
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Laura Gradowski is a philosopher of science at the University of Pittsburgh. Pluralism is roughly the idea that there is no unified account of any scientific field, that we should be tolerant of and welcome a variety of theoretical and conceptual frameworks, and methods, and goals, when doing science. Pluralism is kind of a buzz word right now in my little neuroscience world, but it's an old and well-trodden notion... many philosophers have been calling for pluralism for many years. But how pluralistic should we be in our studies and explanations in science? Laura suggests we should be very, very pluralistic, and to make her case, she cites examples in the history of science of theories and theorists that were once considered "fringe" but went on to become mainstream accepted theoretical frameworks. I thought it would be fun to have her on to share her ideas about fringe theories, mainstream theories, pluralism, etc.
We discuss a wide range of topics, but also discuss some specific to the brain and mind sciences. Laura goes through an example of something and someone going from fringe to mainstream - the Garcia effect, named after John Garcia, whose findings went agains the grain of behaviorism, the dominant dogma of the day in psychology. But this overturning only happened after Garcia had to endure a long scientific hell of his results being ignored and shunned. So, there are multiple examples like that, and we discuss a handful. This has led Laura to the conclusion we should accept almost all theoretical frameworks, We discuss her ideas about how to implement this, where to draw the line, and much more.
0:00 - Intro 3:57 - What is fringe? 10:14 - What makes a theory fringe? 14:31 - Fringe to mainstream 17:23 - Garcia effect 28:17 - Fringe to mainstream: other examples 32:38 - Fringe and consciousness 33:19 - Words meanings change over time 40:24 - Pseudoscience 43:25 - How fringe becomes mainstream 47:19 - More fringe characteristics 50:06 - Pluralism as a solution 54:02 - Progress 1:01:39 - Encyclopedia of theories 1:09:20 - When to reject a theory 1:20:07 - How fringe becomes fringe 1:22:50 - Marginilization 1:27:53 - Recipe for fringe theorist
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Eric Shea-Brown is a theoretical neuroscientist and principle investigator of the working group on neural dynamics at the University of Washington. In this episode, we talk a lot about dynamics and dimensionality in neural networks... how to think about them, why they matter, how Eric's perspectives have changed through his career. We discuss a handful of his specific research findings about dynamics and dimensionality, like how dimensionality changes when one is performing a task versus when you're just sort of going about your day, what we can say about dynamics just by looking at different structural connection motifs, how different modes of learning can rely on different dimensionalities, and more.We also talk about how he goes about choosing what to work on and how to work on it. You'll hear in our discussion how much credit Eric gives to those surrounding him and those who came before him - he drops tons of references and names, so get ready if you want to follow up on some of the many lines of research he mentions.
0:00 - Intro 4:15 - Reflecting on the rise of dynamical systems in neuroscience 11:15 - DST view on macro scale 15:56 - Intuitions 22:07 - Eric's approach 31:13 - Are brains more or less impressive to you now? 38:45 - Why is dimensionality important? 50:03 - High-D in Low-D 54:14 - Dynamical motifs 1:14:56 - Theory for its own sake 1:18:43 - Rich vs. lazy learning 1:22:58 - Latent variables 1:26:58 - What assumptions give you most pause?
Support the show to get full episodes, full archive, and join the Discord community.
I was recently invited to moderate a panel at the Annual Bernstein conference - this one was in Berlin Germany. The panel I moderated was at a satellite workshop at the conference called How can machine learning be used to generate insights and theories in neuroscience? Below are the panelists. I hope you enjoy the discussion!
Organizers:
Support the show to get full episodes, full archive, and join the Discord community.
David runs his lab at NYU, where they stud`y auditory cognition, speech perception, language, and music. On the heels of the episode with David Glanzman, we discuss the ongoing mystery regarding how memory works, how to study and think about brains and minds, and the reemergence (perhaps) of the language of thought hypothesis.
David has been on the podcast a few times... once by himself, and again with Gyorgy Buzsaki.
0:00 - Intro 11:17 - Across levels 14:598 - Nature of memory 24:12 - Using the right tools for the right question 35:46 - LLMs, what they need, how they've shaped David's thoughts 44:55 - Across levels 54:07 - Speed of progress 1:02:21 - Neuroethology and mental illness - patreon 1:24:42 - Language of Thought
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Kevin Mitchell is professor of genetics at Trinity College Dublin. He's been on the podcast before, and we talked a little about his previous book, Innate – How the Wiring of Our Brains Shapes Who We Are. He's back today to discuss his new book Free Agents: How Evolution Gave Us Free Will. The book is written very well and guides the reader through a wide range of scientific knowledge and reasoning that undergirds Kevin's main take home: our free will comes from the fact that we are biological organisms, biological organisms have agency, and as that agency evolved to become more complex and layered, so does our ability to exert free will. We touch on a handful of topics in the book, like the idea of agency, how it came about at the origin of life, and how the complexity of kinds of agency, the richness of our agency, evolved as organisms became more complex.
We also discuss Kevin's reliance on the indeterminacy of the universe to tell his story, the underlying randomness at fundamental levels of physics. Although indeterminacy isn't necessary for ongoing free will, it is responsible for the capacity for free will to exist in the first place. We discuss the brain's ability to harness its own randomness when needed, creativity, whether and how it's possible to create something new, artificial free will, and lots more.
4:27 - From Innate to Free Agents 9:14 - Thinking of the whole organism 15:11 - Who the book is for 19:49 - What bothers Kevin 27:00 - Indeterminacy 30:08 - How it all began 33:08 - How indeterminacy helps 43:58 - Libet's free will experiments 50:36 - Creativity 59:16 - Selves, subjective experience, agency, and free will 1:10:04 - Levels of agency and free will 1:20:38 - How much free will can we have? 1:28:03 - Hierarchy of mind constraints 1:36:39 - Artificial agents and free will 1:42:57 - Next book?
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Alicia Juarrero is a philosopher and has been interested in complexity since before it was cool.
In this episode, we discuss many of the topics and ideas in her new book, Context Changes Everything: How Constraints Create Coherence, which makes the thorough case that constraints should be given way more attention when trying to understand complex systems like brains and minds - how they're organized, how they operate, how they're formed and maintained, and so on. Modern science, thanks in large part to the success of physics, focuses on a single kind of causation - the kind involved when one billiard ball strikes another billiard ball. But that kind of causation neglects what Alicia argues are the most important features of complex systems the constraints that shape the dynamics and possibility spaces of systems. Much of Alicia's book describes the wide range of types of constraints we should be paying attention to, and how they interact and mutually influence each other. I highly recommend the book, and you may want to read it before, during, and after our conversation. That's partly because, if you're like me, the concepts she discusses still aren't comfortable to think about the way we're used to thinking about how things interact. Thinking across levels of organization turns out to be hard. You might also want her book handy because, hang on to your hats, we jump around a lot among those concepts. Context Changes everything comes about 25 years after her previous classic, Dynamics In Action, which we also discuss and which I also recommend if you want more of a primer to her newer more expansive work. Alicia's work touches on all things complex, from self-organizing systems like whirlpools, to ecologies, businesses, societies, and of course minds and brains.
0:00 - Intro 3:37 - 25 years thinking about constraints 8:45 - Dynamics in Action and eliminativism 13:08 - Efficient and other kinds of causation 19:04 - Complexity via context independent and dependent constraints 25:53 - Enabling and limiting constraints 30:55 - Across scales 36:32 - Temporal constraints 42:58 - A constraint cookbook? 52:12 - Constraints in a mechanistic worldview 53:42 - How to explain using constraints 56:22 - Concepts and multiple realizabillity 59:00 - Kevin Mitchell question 1:08:07 - Mac Shine Question 1:19:07 - 4E 1:21:38 - Dimensionality across levels 1:27:26 - AI and constraints 1:33:08 - AI and life
Support the show to get full episodes, full archive, and join the Discord community.
In the intro, I mention the Bernstein conference workshop I'll participate in, called How can machine learning be used to generate insights and theories in neuroscience?. Follow that link to learn more, and register for the conference here. Hope to see you there in late September in Berlin!
Justin Wood runs the Wood Lab at Indiana University, and his lab's tagline is "building newborn minds in virtual worlds." In this episode, we discuss his work comparing the visual cognition of newborn chicks and AI models. He uses a controlled-rearing technique with natural chicks, whereby the chicks are raised from birth in completely controlled visual environments. That way, Justin can present designed visual stimuli to test what kinds of visual abilities chicks have or can immediately learn. Then he can building models and AI agents that are trained on the same data as the newborn chicks. The goal is to use the models to better understand natural visual intelligence, and use what we know about natural visual intelligence to help build systems that better emulate biological organisms. We discuss some of the visual abilities of the chicks and what he's found using convolutional neural networks. Beyond vision, we discuss his work studying the development of collective behavior, which compares chicks to a model that uses CNNs, reinforcement learning, and an intrinsic curiosity reward function. All of this informs the age-old nature (nativist) vs. nurture (empiricist) debates, which Justin believes should give way to embrace both nature and nurture.
0:00 - Intro 5:39 - Origins of Justin's current research 11:17 - Controlled rearing approach 21:52 - Comparing newborns and AI models 24:11 - Nativism vs. empiricism 28:15 - CNNs and early visual cognition 29:35 - Smoothness and slowness 50:05 - Early biological development 53:27 - Naturalistic vs. highly controlled 56:30 - Collective behavior in animals and machines 1:02:34 - Curiosity and critical periods 1:09:05 - Controlled rearing vs. other developmental studies 1:13:25 - Breaking natural rules 1:16:33 - Deep RL collective behavior 1:23:16 - Bottom-up and top-down
Support the show to get full episodes, full archive, and join the Discord community.
David runs his lab at UCLA where he's also a distinguished professor. David used to believe what is currently the mainstream view, that our memories are stored in our synapses, those connections between our neurons. So as we learn, the synaptic connections strengthen and weaken until their just right, and that serves to preserve the memory. That's been the dominant view in neuroscience for decades, and is the fundamental principle that underlies basically all of deep learning in AI. But because of his own and others experiments, which he describes in this episode, David has come to the conclusion that memory must be stored not at the synapse, but in the nucleus of neurons, likely by some epigenetic mechanism mediated by RNA molecules. If this sounds familiar, I had Randy Gallistel on the the podcast on episode 126 to discuss similar ideas, and David discusses where he and Randy differ in their thoughts. This episode starts out pretty technical as David describes the series of experiments that changed his mind, but after that we broaden our discussion to a lot of the surrounding issues regarding whether and if his story about memory is true. And we discuss meta-issues like how old discarded ideas in science often find their way back, what it's like studying non-mainstream topic, including challenges trying to get funded for it, and so on.
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
My guest is Michael C. Frank, better known as Mike Frank, who runs the Language and Cognition lab at Stanford. Mike's main interests center on how children learn language - in particular he focuses a lot on early word learning, and what that tells us about our other cognitive functions, like concept formation and social cognition.
We discuss that, his love for developing open data sets that anyone can use,
The dance he dances between bottom-up data-driven approaches in this big data era, traditional experimental approaches, and top-down theory-driven approaches
How early language learning in children differs from LLM learning
Mike's rational speech act model of language use, which considers the intentions or pragmatics of speakers and listeners in dialogue.
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
In this episode I have a casual chat with Ali Mohebi about his new faculty position and his plans for the future.
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
My guest today is Andrea Martin, who is the Research Group Leader in the department of Language and Computation in Neural Systems at the Max Plank Institute and the Donders Institute. Andrea is deeply interested in understanding how our biological brains process and represent language. To this end, she is developing a theoretical model of language. The aim of the model is to account for the properties of language, like its structure, its compositionality, its infinite expressibility, while adhering to physiological data we can measure from human brains.
Her theoretical model of language, among other things, brings in the idea of low-dimensional manifolds and neural dynamics along those manifolds. We've discussed manifolds a lot on the podcast, but they are a kind of abstract structure in the space of possible neural population activity - the neural dynamics. And that manifold structure defines the range of possible trajectories, or pathways, the neural dynamics can take over time.
One of Andrea's ideas is that manifolds might be a way for the brain to combine two properties of how we learn and use language. One of those properties is the statistical regularities found in language - a given word, for example, occurs more often near some words and less often near some other words. This statistical approach is the foundation of how large language models are trained. The other property is the more formal structure of language: how it's arranged and organized in such a way that gives it meaning to us. Perhaps these two properties of language can come together as a single trajectory along a neural manifold. But she has lots of ideas, and we discuss many of them. And of course we discuss large language models, and how Andrea thinks of them with respect to biological cognition. We talk about modeling in general and what models do and don't tell us, and much more.
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
This is one in a periodic series of episodes with Alex Gomez-Marin, exploring how the arts and humanities can impact (neuro)science. Artistic creations, like cinema, have the ability to momentarily lower our ever-critical scientific mindset and allow us to imagine alternate possibilities and experience emotions outside our normal scientific routines. Might this feature of art potentially change our scientific attitudes and perspectives?
Frauke Sandig and Eric Black recently made the documentary film AWARE: Glimpses of Consciousness, which profiles six researchers studying consciousness from different perspectives. The film is filled with rich visual imagery and conveys a sense of wonder and awe in trying to understand subjective experience, while diving deep into the reflections of the scientists and thinkers approaching the topic from their various perspectives.
This isn't a "normal" Brain Inspired episode, but I hope you enjoy the discussion!
0:00 - Intro 19:42 - Mechanistic reductionism 45:33 - Changing views during lifetime 53:49 - Did making the film alter your views? 57:49 - ChatGPT 1:04:20 - Materialist assumption 1:11:00 - Science of consciousness 1:20:49 - Transhumanism 1:32:01 - Integrity 1:36:19 - Aesthetics 1:39:50 - Response to the film
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Panayiota Poirazi runs the Poirazi Lab at the FORTH Institute of Molecular Biology and Biotechnology, and Yiota loves dendrites, those branching tree-like structures sticking out of all your neurons, and she thinks you should love dendrites, too, whether you study biological or artificial intelligence. In neuroscience, the old story was that dendrites just reach out and collect incoming signals for the all-important neuron cell body to process. Yiota, and people Like Matthew Larkum, with whom I chatted in episode 138, are continuing to demonstrate that dendrites are themselves computationally complex and powerful, doing many varieties of important signal transformation before signals reach the cell body. For example, in 2003, Yiota showed that because of dendrites, a single neuron can act as a two-layer artificial neural network, and since then others have shown single neurons can act as deeper and deeper multi-layer networks. In Yiota's opinion, an even more important function of dendrites is increased computing efficiency, something evolution favors and something artificial networks need to favor as well moving forward.
0:00 - Intro 3:04 - Yiota's background 6:40 - Artificial networks and dendrites 9:24 - Dendrites special sauce? 14:50 - Where are we in understanding dendrite function? 20:29 - Algorithms, plasticity, and brains 29:00 - Functional unit of the brain 42:43 - Engrams 51:03 - Dendrites and nonlinearity 54:51 - Spiking neural networks 56:02 - Best level of biological detail 57:52 - Dendrify 1:05:41 - Experimental work 1:10:58 - Dendrites across species and development 1:16:50 - Career reflection 1:17:57 - Evolution of Yiota's thinking
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Nick Enfield is a professor of linguistics at the University of Sydney. In this episode we discuss topics in his most recent book, Language vs. Reality: Why Language Is Good for Lawyers and Bad for Scientists. A central question in the book is what is language for? What's the function of language. You might be familiar with the debate about whether language evolved for each of us thinking our wonderful human thoughts, or for communicating those thoughts between each other. Nick would be on the communication side of that debate, but if by communication we mean simply the transmission of thoughts or information between people - I have a thought, I send it to you in language, and that thought is now in your head - then Nick wouldn't take either side of that debate. He argues the function language goes beyond the transmission of information, and instead is primarily an evolved solution for social coordination - coordinating our behaviors and attention. When we use language, we're creating maps in our heads so we can agree on where to go.
For example, when I say, "This is brain inspired," I'm pointing you to a place to meet me on a conceptual map, saying, "Get ready, we're about to have a great time again!" In any case, with those 4 words, "This is brain inspired," I'm not just transmitting information from my head into your head. I'm providing you with a landmark so you can focus your attention appropriately.
From that premise, that language is about social coordination, we talk about a handful of topics in his book, like the relationship between language and reality, the idea that all language is framing- that is, how we say something influences how to think about it. We discuss how our language changes in different social situations, the role of stories, and of course, how LLMs fit into Nick's story about language.
0:00 - Intro 4:23 - Is learning about language important? 15:43 - Linguistic Anthropology 28:56 - Language and truth 33:57 - How special is language 46:19 - Choice architecture and framing 48:19 - Language for thinking or communication 52:30 - Agency and language 56:51 - Large language models 1:16:18 - Getting language right 1:20:48 - Social relationships and language
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Jeffrey Bowers is a psychologist and professor at the University of Bristol. As you know, many of my previous guests are in the business of comparing brain activity to the activity of units in artificial neural network models, when humans or animals and the models are performing the same tasks. And a big story that has emerged over the past decade or so is that there's a remarkable similarity between the activities and representations in brains and models. This was originally found in object categorization tasks, where the goal is to name the object shown in a given image, where researchers have compared the activity in the models good at doing that to the activity in the parts of our brains good at doing that. It's been found in various other tasks using various other models and analyses, many of which we've discussed on previous episodes, and more recently a similar story has emerged regarding a similarity between language-related activity in our brains and the activity in large language models. Namely, the ability of our brains to predict an upcoming word can been correlated with the models ability to predict an upcoming word. So the word is that these deep learning type models are the best models of how our brains and cognition work.
However, this is where Jeff Bowers comes in and raises the psychology flag, so to speak. His message is that these predictive approaches to comparing artificial and biological cognition aren't enough, and can mask important differences between them. And what we need to do is start performing more hypothesis driven tests like those performed in psychology, for example, to ask whether the models are indeed solving tasks like our brains and minds do. Jeff and his group, among others, have been doing just that are discovering differences in models and minds that may be important if we want to use models to understand minds. We discuss some of his work and thoughts in this regard, and a lot more.
0:00 - Intro 3:52 - Testing neural networks 5:35 - Neuro-AI needs psychology 23:36 - Experiments in AI and neuroscience 23:51 - Why build networks like our minds? 44:55 - Vision problem spaces, solution spaces, training data 55:45 - Do we implement algorithms? 1:01:33 - Relational and combinatorial cognition 1:06:17 - Comparing representations in different networks 1:12:31 - Large language models 1:21:10 - Teaching LLMs nonsense languages
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Gary Lupyan runs the Lupyan Lab at University of Wisconsin, Madison, where he studies how language and cognition are related. In some ways, this is a continuation of the conversation I had last episode with Ellie Pavlick, in that we partly continue to discuss large language models. But Gary is more focused on how language, and naming things, categorizing things, changes our cognition related those things. How does naming something change our perception of it, and so on. He's interested in how concepts come about, how they map onto language. So we talk about some of his work and ideas related to those topics.
And we actually start the discussion with some of Gary's work related the variability of individual humans' phenomenal experience, and how that affects our individual cognition. For instance, some people are more visual thinkers, others are more verbal, and there seems to be an appreciable spectrum of differences that Gary is beginning to experimentally test.
0:00 - Intro 2:36 - Words and communication 14:10 - Phenomenal variability 26:24 - Co-operating minds 38:11 - Large language models 40:40 - Neuro-symbolic AI, scale 44:43 - How LLMs have changed Gary's thoughts about language 49:26 - Meaning, grounding, and language 54:26 - Development of language 58:53 - Symbols and emergence 1:03:20 - Language evolution in the LLM era 1:08:05 - Concepts 1:11:17 - How special is language? 1:18:08 - AGI
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Ellie Pavlick runs her Language Understanding and Representation Lab at Brown University, where she studies lots of topics related to language. In AI, large language models, sometimes called foundation models, are all the rage these days, with their ability to generate convincing language, although they still make plenty of mistakes. One of the things Ellie is interested in is how these models work, what kinds of representations are being generated in them to produce the language they produce. So we discuss how she's going about studying these models. For example, probing them to see whether something symbolic-like might be implemented in the models, even though they are the deep learning neural network type, which aren't suppose to be able to work in a symbol-like manner. We also discuss whether grounding is required for language understanding - that is, whether a model that produces language well needs to connect with the real world to actually understand the text it generates. We talk about what language is for, the current limitations of large language models, how the models compare to humans, and a lot more.
0:00 - Intro 2:34 - Will LLMs make us dumb? 9:01 - Evolution of language 17:10 - Changing views on language 22:39 - Semantics, grounding, meaning 37:40 - LLMs, humans, and prediction 41:19 - How to evaluate LLMs 51:08 - Structure, semantics, and symbols in models 1:00:08 - Dimensionality 1:02:08 - Limitations of LLMs 1:07:47 - What do linguists think? 1:14:23 - What is language for?
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Earl Miller runs the Miller Lab at MIT, where he studies how our brains carry out our executive functions, like working memory, attention, and decision-making. In particular he is interested in the role of the prefrontal cortex and how it coordinates with other brain areas to carry out these functions. During this episode, we talk broadly about how neuroscience has changed during Earl's career, and how his own thoughts have changed. One thing we focus on is the increasing appreciation of brain oscillations for our cognition.
Recently on BI we've discussed oscillations quite a bit. In episode 153, Carolyn Dicey-Jennings discussed her philosophical ideas relating attention to the notion of the self, and she leans a lot on Earl's research to make that argument. In episode 160, Ole Jensen discussed his work in humans showing that low frequency oscillations exert a top-down control on incoming sensory stimuli, and this is directly in agreement with Earl's work over many years in nonhuman primates. So we continue that discussion relating low-frequency oscillations to executive control. We also discuss a new concept Earl has developed called spatial computing, which is an account of how brain oscillations can dictate where in various brain areas neural activity be on or off, and hence contribute or not to ongoing mental function. We also discuss working memory in particular, and a host of related topics.
0:00 - Intro 6:22 - Evolution of Earl's thinking 14:58 - Role of the prefrontal cortex 25:21 - Spatial computing 32:51 - Homunculus problem 35:34 - Self 37:40 - Dimensionality and thought 46:13 - Reductionism 47:38 - Working memory and capacity 1:01:45 - Capacity as a principle 1:05:44 - Silent synapses 1:10:16 - Subspaces in dynamics
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Hugo Spiers runs the Spiers Lab at University College London. In general Hugo is interested in understanding spatial cognition, like navigation, in relation to other processes like planning and goal-related behavior, and how brain areas like the hippocampus and prefrontal cortex coordinate these cognitive functions. So, in this episode, we discuss a range of his research and thoughts around those topics. You may have heard about the studies he's been involved with for years, regarding London taxi drivers and how their hippocampus changes as a result of their grueling efforts to memorize how to best navigate London. We talk about that, we discuss the concept of a schema, which is roughly an abstracted form of knowledge that helps you know how to behave in different environments. Probably the most common example is that we all have a schema for eating at a restaurant, independent of which restaurant we visit, we know about servers, and menus, and so on. Hugo is interested in spatial schemas, for things like navigating a new city you haven't visited. Hugo describes his work using reinforcement learning methods to compare how humans and animals solve navigation tasks. And finally we talk about the video game Hugo has been using to collect vast amount of data related to navigation, to answer questions like how our navigation ability changes over our lifetimes, the different factors that seem to matter more for our navigation skills, and so on.
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Ole Jensen is co-director of the Centre for Human Brain Health at University of Birmingham, where he runs his Neuronal Oscillations Group lab. Ole is interested in how the oscillations in our brains affect our cognition by helping to shape the spiking patterns of neurons, and by helping to allocate resources to parts of our brains that are relevant for whatever ongoing behaviors we're performing in different contexts. People have been studying oscillations for decades, finding that different frequencies of oscillations have been linked to a bunch of different cognitive functions. Some of what we discuss today is Ole's work on alpha oscillations, which are around 10 hertz, so 10 oscillations per second. The overarching story is that alpha oscillations are thought to inhibit or disrupt processing in brain areas that aren't needed during a given behavior. And therefore by disrupting everything that's not needed, resources are allocated to the brain areas that are needed. We discuss his work in the vein on attention - you may remember the episode with Carolyn Dicey-Jennings, and her ideas about how findings like Ole's are evidence we all have selves. We also talk about the role of alpha rhythms for working memory, for moving our eyes, and for previewing what we're about to look at before we move our eyes, and more broadly we discuss the role of oscillations in cognition in general, and of course what this might mean for developing better artificial intelligence.
0:00 - Intro 2:58 - Oscillations import over the years 5:51 - Oscillations big picture 17:62 - Oscillations vs. traveling waves 22:00 - Oscillations and algorithms 28:53 - Alpha oscillations and working memory 44:46 - Alpha as the controller 48:55 - Frequency tagging 52:49 - Timing of attention 57:41 - Pipelining neural processing 1:03:38 - Previewing during reading 1:15:50 - Previewing, prediction, and large language models 1:24:27 - Dyslexia
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Chris Summerfield runs the Human Information Processing Lab at University of Oxford, and he's a research scientist at Deepmind. You may remember him from episode 95 with Sam Gershman, when we discussed ideas around the usefulness of neuroscience and psychology for AI. Since then, Chris has released his book, Natural General Intelligence: How understanding the brain can help us build AI. In the book, Chris makes the case that inspiration and communication between the cognitive sciences and AI is hindered by the different languages each field speaks. But in reality, there has always been and still is a lot of overlap and convergence about ideas of computation and intelligence, and he illustrates this using tons of historical and modern examples.
0:00 - Intro 2:20 - Natural General Intelligence 8:05 - AI and Neuro interaction 21:42 - How to build AI 25:54 - Umwelts and affordances 32:07 - Different kind of intelligence 39:16 - Ecological validity and AI 48:30 - Is reward enough? 1:05:14 - Beyond brains 1:15:10 - Large language models and brains
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Paul Rosenbloom is Professor Emeritus of Computer Science at the University of Southern California. In the early 1980s, Paul , along with John Laird and the early AI pioneer Alan Newell, developed one the earliest and best know cognitive architectures called SOAR. A cognitive architecture, as Paul defines it, is a model of the fixed structures and processes underlying minds, and in Paul's case the human mind. And SOAR was aimed at generating general intelligence. He doesn't work on SOAR any more, although SOAR is still alive and well in the hands of his old partner John Laird. He did go on to develop another cognitive architecture, called Sigma, and in the intervening years between those projects, among other things Paul stepped back and explored how our various scientific domains are related, and how computing itself should be considered a great scientific domain. That's in his book On Computing: The Fourth Great Scientific Domain.
He also helped develop the Common Model of Cognition, which isn't a cognitive architecture itself, but instead a theoretical model meant to generate consensus regarding the minimal components for a human-like mind. The idea is roughly to create a shared language and framework among cognitive architecture researchers, so the field can , so that whatever cognitive architecture you work on, you have a basis to compare it to, and can communicate effectively among your peers.
All of what I just said, and much of what we discuss, can be found in Paul's memoir, In Search of Insight: My Life as an Architectural Explorer.
0:00 - Intro 3:26 - A career of exploration 7:00 - Alan Newell 14:47 - Relational model and dichotomic maps 24:22 - Cognitive architectures 28:31 - SOAR cognitive architecture 41:14 - Sigma cognitive architecture 43:58 - SOAR vs. Sigma 53:06 - Cognitive architecture community 55:31 - Common model of cognition 1:11:13 - What's missing from the common model 1:17:48 - Brains vs. cognitive architectures 1:21:22 - Mapping the common model onto the brain 1:24:50 - Deep learning 1:30:23 - AGI
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Sarah Robins is a philosopher at the University of Kansas, one a growing handful of philosophers specializing in memory. Much of her work focuses on memory traces, which is roughly the idea that somehow our memories leave a trace in our minds. We discuss memory traces themselves and how they relate to the engram (see BI 126 Randy Gallistel: Where Is the Engram?, and BI 127 Tomás Ryan: Memory, Instinct, and Forgetting).
Psychology has divided memories into many categories - the taxonomy of memory. Sarah and I discuss how memory traces may cross-cut those categories, suggesting we may need to re-think our current ontology and taxonomy of memory.
We discuss a couple challenges to the idea of a stable memory trace in the brain. Neural dynamics is the notion that all our molecules and synapses are constantly changing and being recycled. Memory consolidation refers to the process of transferring our memory traces from an early unstable version to a more stable long-term version in a different part of the brain. Sarah thinks neither challenge poses a real threat to the idea
We also discuss the impact of optogenetics on the philosophy and neuroscience and memory, the debate about whether memory and imagination are essentially the same thing, whether memory's function is future oriented, and whether we want to build AI with our often faulty human-like memory or with perfect memory.
0:00 - Intro 4:18 - Philosophy of memory 5:10 - Making a move 6:55 - State of philosophy of memory 11:19 - Memory traces or the engram 20:44 - Taxonomy of memory 25:50 - Cognitive ontologies, neuroscience, and psychology 29:39 - Optogenetics 33:48 - Memory traces vs. neural dynamics and consolidation 40:32 - What is the boundary of a memory? 43:00 - Process philosophy and memory 45:07 - Memory vs. imagination 49:40 - Constructivist view of memory and imagination 54:05 - Is memory for the future? 58:00 - Memory errors and intelligence 1:00:42 - Memory and AI 1:06:20 - Creativity and memory errors
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Mariam Aly runs the Aly lab at Columbia University, where she studies the interaction of memory, attention, and perception in brain regions like the hippocampus. The short story is that memory affects our perceptions, attention affects our memories, memories affect our attention, and these effects have signatures in neural activity measurements in our hippocampus and other brain areas. We discuss her experiments testing the nature of those interactions. We also discuss a particularly difficult stretch in Mariam's graduate school years, and how she now prioritizes her mental health.
0:00 - Intro 3:50 - Mariam's background 9:32 - Hippocampus history and current science 12:34 - hippocampus and perception 13:42 - Relational information 18:30 - How much memory is explicit? 22:32 - How attention affects hippocampus 32:40 - fMRI levels vs. stability 39:04 - How is hippocampus necessary for attention 57:00 - How much does attention affect memory? 1:02:24 - How memory affects attention 1:06:50 - Attention and memory relation big picture 1:07:42 - Current state of memory and attention 1:12:12 - Modularity 1:17:52 - Practical advice to improve attention/memory 1:21:22 - Mariam's challenges
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Luiz Pessoa runs his Laboratory of Cognition and Emotion at the University of Maryland, College Park, where he studies how emotion and cognition interact. On this episode, we discuss many of the topics from his latest book, The Entangled Brain: How Perception, Cognition, and Emotion Are Woven Together, which is aimed at a general audience. The book argues we need to re-think how to study the brain. Traditionally, cognitive functions of the brain have been studied in a modular fashion: area X does function Y. However, modern research has revealed the brain is highly complex and carries out cognitive functions in a much more interactive and integrative fashion: a given cognitive function results from many areas and circuits temporarily coalescing (for similar ideas, see also BI 152 Michael L. Anderson: After Phrenology: Neural Reuse). Luiz and I discuss the implications of studying the brain from a complex systems perspective, why we need go beyond thinking about anatomy and instead think about functional organization, some of the brain's principles of organization, and a lot more.
0:00 - Intro 2:47 - The Entangled Brain 16:24 - How to think about complex systems 23:41 - Modularity thinking 28:16 - How to train one's mind to think complex 33:26 - Problem or principle? 44:22 - Complex behaviors 47:06 - Organization vs. structure 51:09 - Principles of organization: Massive Combinatorial Anatomical Connectivity 55:15 - Principles of organization: High Distributed Functional Connectivity 1:00:50 - Principles of organization: Networks as Functional Units 1:06:15 - Principles of Organization: Interactions via Cortical-Subcortical Loops 1:08:53 - Open and closed loops 1:16:43 - Principles of organization: Connectivity with the Body 1:21:28 - Consciousness 1:24:53 - Emotions 1:32:49 - Emottions and AI 1:39:47 - Emotion as a concept 1:43:25 - Complexity and functional organization in AI
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Anne Collins runs her Computational Cognitive Neuroscience Lab at the University of California, Berkley One of the things she's been working on for years is how our working memory plays a role in learning as well, and specifically how working memory and reinforcement learning interact to affect how we learn, depending on the nature of what we're trying to learn. We discuss that interaction specifically. We also discuss more broadly how segregated and how overlapping and interacting our cognitive functions are, what that implies about our natural tendency to think in dichotomies - like MF vs MB-RL, system-1 vs system-2, etc., and we dive into plenty other subjects, like how to possibly incorporate these ideas into AI.
0:00 - Intro 5:25 - Dimensionality of learning 11:19 - Modularity of function and computations 16:51 - Is working memory a thing? 19:33 - Model-free model-based dichotomy 30:40 - Working memory and RL 44:43 - How working memory and RL interact 50:50 - Working memory and attention 59:37 - Computations vs. implementations 1:03:25 - Interpreting results 1:08:00 - Working memory and AI
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Carolyn Dicey Jennings is a philosopher and a cognitive scientist at University of California, Merced. In her book The Attending Mind, she lays out an attempt to unify the concept of attention. Carolyn defines attention roughly as the mental prioritization of some stuff over other stuff based on our collective interests. And one of her main claims is that attention is evidence of a real, emergent self or subject, that can't be reduced to microscopic brain activity. She does connect attention to more macroscopic brain activity, suggesting slow longer-range oscillations in our brains can alter or entrain the activity of more local neural activity, and this is a candidate for mental causation. We unpack that more in our discussion, and how Carolyn situates attention among other cognitive functions, like consciousness, action, and perception.
0:00 - Intro 12:15 - Reconceptualizing attention 16:07 - Types of attention 19:02 - Predictive processing and attention 23:19 - Consciousness, identity, and self 30:39 - Attention and the brain 35:47 - Integrated information theory 42:05 - Neural attention 52:08 - Decoupling oscillations from spikes 57:16 - Selves in other organisms 1:00:42 - AI and the self 1:04:43 - Attention, consciousness, conscious perception 1:08:36 - Meaning and attention 1:11:12 - Conscious entrainment 1:19:57 - Is attention a switch or knob?
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Michael L. Anderson is a professor at the Rotman Institute of Philosophy, at Western University. His book, After Phrenology: Neural Reuse and the Interactive Brain, calls for a re-conceptualization of how we understand and study brains and minds. Neural reuse is the phenomenon that any given brain area is active for multiple cognitive functions, and partners with different sets of brain areas to carry out different cognitive functions. We discuss the implications for this, and other topics in Michael's research and the book, like evolution, embodied cognition, and Gibsonian perception. Michael also fields guest questions from John Krakauer and Alex Gomez-Marin, about representations and metaphysics, respectively.
0:00 - Intro 3:02 - After Phrenology 13:18 - Typical neuroscience experiment 16:29 - Neural reuse 18:37 - 4E cognition and representations 22:48 - John Krakauer question 27:38 - Gibsonian perception 36:17 - Autoencoders without representations 49:22 - Pluralism 52:42 - Alex Gomez-Marin question - metaphysics 1:01:26 - Stimulus-response historical neuroscience 1:10:59 - After Phrenology influence 1:19:24 - Origins of neural reuse 1:35:25 - The way forward
Support the show to get full episodes, full archive, and join the Discord community.
Steve Byrnes is a physicist turned AGI safety researcher. He's concerned that when we create AGI, whenever and however that might happen, we run the risk of creating it in a less than perfectly safe way. AGI safety (AGI not doing something bad) is a wide net that encompasses AGI alignment (AGI doing what we want it to do). We discuss a host of ideas Steve writes about in his Intro to Brain-Like-AGI Safety blog series, which uses what he has learned about brains to address how we might safely make AGI.
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Dan Nicholson is a philosopher at George Mason University. He incorporates the history of science and philosophy into modern analyses of our conceptions of processes related to life and organisms. He is also interested in re-orienting our conception of the universe as made fundamentally of things/substances, and replacing it with the idea the universe is made fundamentally of processes (process philosophy). In this episode, we both of those subjects, the why the "machine conception of the organism" is incorrect, how to apply these ideas to topics like neuroscience and artificial intelligence, and much more.
0:00 - Intro 2:49 - Philosophy and science 16:37 - Role of history 23:28 - What Is Life? And interaction with James Watson 38:37 - Arguments against the machine conception of organisms 49:08 - Organisms as streams (processes) 57:52 - Process philosophy 1:08:59 - Alfred North Whitehead 1:12:45 - Process and consciousness 1:22:16 - Artificial intelligence and process 1:31:47 - Language and symbols and processes
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
William B. Miller is an ex-physician turned evolutionary biologist. In this episode, we discuss topics related to his new book, Bioverse: How the Cellular World Contains the Secrets to Life's Biggest Questions. The premise of the book is that all individual cells are intelligent in their own right, and possess a sense of self. From this, Bill makes the case that cells cooperate with other cells to engineer whole organisms that in turn serve as wonderful hosts for the myriad cell types. Further, our bodies are collections of our own cells (with our DNA), and an enormous amount and diversity of foreign cells - our microbiome - that communicate and cooperate with each other and with our own cells. We also discuss how cell intelligence compares to human intelligence, what Bill calls the "era of the cell" in science, how the future of medicine will harness the intelligence of cells and their cooperative nature, and much more.
0:00 - Intro 3:43 - Bioverse 7:29 - Bill's cell appreciation origins 17:03 - Microbiomes 27:01 - Complexity of microbiomes and the "Era of the cell" 46:00 - Robustness 55:05 - Cell vs. human intelligence 1:10:08 - Artificial intelligence 1:21:01 - Neuro-AI 1:25:53 - Hard problem of consciousness
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Gaute Einevoll is a professor at the University of Oslo and Norwegian University of Life Sciences. Use develops detailed models of brain networks to use as simulations, so neuroscientists can test their various theories and hypotheses about how networks implement various functions. Thus, the models are tools. The goal is to create models that are multi-level, to test questions at various levels of biological detail; and multi-modal, to predict that handful of signals neuroscientists measure from real brains (something Gaute calls "measurement physics"). We also discuss Gaute's thoughts on Carina Curto's "beautiful vs ugly models", and his reaction to Noah Hutton's In Silico documentary about the Blue Brain and Human Brain projects (Gaute has been funded by the Human Brain Project since its inception).
0:00 - Intro 3:25 - Beautiful and messy models 6:34 - In Silico 9:47 - Goals of human brain project 15:50 - Brain simulation approach 21:35 - Degeneracy in parameters 26:24 - Abstract principles from simulations 32:58 - Models as tools 35:34 - Predicting brain signals 41:45 - LFPs closer to average 53:57 - Plasticity in simulations 56:53 - How detailed should we model neurons? 59:09 - Lessons from predicting signals 1:06:07 - Scaling up 1:10:54 - Simulation as a tool 1:12:35 - Oscillations 1:16:24 - Manifolds and simulations 1:20:22 - Modeling cortex like Hodgkin and Huxley
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Noah Hutton writes, directs, and scores documentary and narrative films. On this episode, we discuss his documentary In Silico. In 2009, Noah watched a TED talk by Henry Markram, in which Henry claimed it would take 10 years to fully simulate a human brain. This claim inspired Noah to chronicle the project, visiting Henry and his team periodically throughout. The result was In Silico, which tells the science, human, and social story of Henry's massively funded projects - the Blue Brain Project and the Human Brain Project.
0:00 - Intro 3:36 - Release and premier 7:37 - Noah's background 9:52 - Origins of In Silico 19:39 - Recurring visits 22:13 - Including the critics 25:22 - Markram's shifting outlook and salesmanship 35:43 - Promises and delivery 41:28 - Computer and brain terms interchange 49:22 - Progress vs. illusion of progress 52:19 - Close to quitting 58:01 - Salesmanship vs bad at estimating timelines 1:02:12 - Brain simulation science 1:11:19 - AGI 1:14:48 - Brain simulation vs. neuro-AI 1:21:03 - Opinion on TED talks 1:25:16 - Hero worship 1:29:03 - Feedback on In Silico
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Lauren Ross is an Associate Professor at the University of California, Irvine. She studies and writes about causal and non-causal explanations in philosophy of science, including distinctions among causal structures. Throughout her work, Lauren employs Jame's Woodward's interventionist approach to causation, which Jim and I discussed in episode 145. In this episode, we discuss Jim's lasting impact on the philosophy of causation, the current dominance of mechanistic explanation and its relation to causation, and various causal structures of explanation, including pathways, cascades, topology, and constraints.
0:00 - Intro 2:46 - Lauren's background 10:14 - Jim Woodward legacy 15:37 - Golden era of causality 18:56 - Mechanistic explanation 28:51 - Pathways 31:41 - Cascades 36:25 - Topology 41:17 - Constraint 50:44 - Hierarchy of explanations 53:18 - Structure and function 57:49 - Brain and mind 1:01:28 - Reductionism 1:07:58 - Constraint again 1:14:38 - Multiple realizability
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
James Woodward is a recently retired Professor from the Department of History and Philosophy of Science at the University of Pittsburgh. Jim has tremendously influenced the field of causal explanation in the philosophy of science. His account of causation centers around intervention - intervening on a cause should alter its effect. From this minimal notion, Jim has described many facets and varieties of causal structures. In this episode, we discuss topics from his recent book, Causation with a Human Face: Normative Theory and Descriptive Psychology. In the book, Jim advocates that how we should think about causality - the normative - needs to be studied together with how we actually do think about causal relations in the world - the descriptive. We discuss many topics around this central notion, epistemology versus metaphysics, the the nature and varieties of causal structures.
0:00 - Intro 4:14 - Causation with a Human Face & Functionalist approach 6:16 - Interventionist causality; Epistemology and metaphysics 9:35 - Normative and descriptive 14:02 - Rationalist approach 20:24 - Normative vs. descriptive 28:00 - Varying notions of causation 33:18 - Invariance 41:05 - Causality in complex systems 47:09 - Downward causation 51:14 - Natural laws 56:38 - Proportionality 1:01:12 - Intuitions 1:10:59 - Normative and descriptive relation 1:17:33 - Causality across disciplines 1:21:26 - What would help our understanding of causation
Check out my short video series about what's missing in AI and Neuroscience.
Support the show to get full episodes, full archive, and join the Discord community.
Large language models, often now called "foundation models", are the model de jour in AI, based on the transformer architecture. In this episode, I bring together Evelina Fedorenko and Emily M. Bender to discuss how language models stack up to our own language processing and generation (models and brains both excel at next-word prediction), whether language evolved in humans for complex thoughts or for communication (communication, says Ev), whether language models grasp the meaning of the text they produce (Emily says no), and much more.
Evelina Fedorenko is a cognitive scientist who runs the EvLab at MIT. She studies the neural basis of language. Her lab has amassed a large amount of data suggesting language did not evolve to help us think complex thoughts, as Noam Chomsky has argued, but rather for efficient communication. She has also recently been comparing the activity in language models to activity in our brain's language network, finding commonality in the ability to predict upcoming words.
Emily M. Bender is a computational linguist at University of Washington. Recently she has been considering questions about whether language models understand the meaning of the language they produce (no), whether we should be scaling language models as is the current practice (not really), how linguistics can inform language models, and more.
0:00 - Intro 4:35 - Language and cognition 15:38 - Grasping for meaning 21:32 - Are large language models producing language? 23:09 - Next-word prediction in brains and models 32:09 - Interface between language and thought 35:18 - Studying language in nonhuman animals 41:54 - Do we understand language enough? 45:51 - What do language models need? 51:45 - Are LLMs teaching us about language? 54:56 - Is meaning necessary, and does it matter how we learn language? 1:00:04 - Is our biology important for language? 1:04:59 - Future outlook
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Rodolphe Sepulchre is a control engineer and theorist at Cambridge University. He focuses on applying feedback control engineering principles to build circuits that model neurons and neuronal circuits. We discuss his work on mixed feedback control - positive and negative - as an underlying principle of the mixed digital and analog brain signals,, the role of neuromodulation as a controller, applying these principles to Eve Marder's lobster/crab neural circuits, building mixed-feedback neuromorphics, some feedback control history, and how "If you wish to contribute original work, be prepared to face loneliness," among other topics.
0:00 - Intro 4:38 - Control engineer 9:52 - Control vs. dynamical systems 13:34 - Building vs. understanding 17:38 - Mixed feedback signals 26:00 - Robustness 28:28 - Eve Marder 32:00 - Loneliness 37:35 - Across levels 44:04 - Neuromorphics and neuromodulation 52:15 - Barrier to adopting neuromorphics 54:40 - Deep learning influence 58:04 - Beyond energy efficiency 1:02:02 - Deep learning for neuro 1:14:15 - Role of philosophy 1:16:43 - Doing it right
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Cameron Buckner is a philosopher and cognitive scientist at The University of Houston. He is writing a book about the age-old philosophical debate on how much of our knowledge is innate (nature, rationalism) versus how much is learned (nurture, empiricism). In the book and his other works, Cameron argues that modern AI can help settle the debate. In particular, he suggests we focus on what types of psychological "domain-general faculties" underlie our own intelligence, and how different kinds of deep learning models are revealing how those faculties may be implemented in our brains. The hope is that by building systems that possess the right handful of faculties, and putting those systems together in a way they can cooperate in a general and flexible manner, it will result in cognitive architectures we would call intelligent. Thus, what Cameron calls The New DoGMA: Domain-General Modular Architecture. We also discuss his work on mental representation and how representations get their content - how our thoughts connect to the natural external world.
0:00 - Intro 4:55 - Interpreting old philosophy 8:26 - AI and philosophy 17:00 - Empiricism vs. rationalism 27:09 - Domain-general faculties 33:10 - Faculty psychology 40:28 - New faculties? 46:11 - Human faculties 51:15 - Cognitive architectures 56:26 - Language 1:01:40 - Beyond dichotomous thinking 1:04:08 - Lower-level faculties 1:10:16 - Animal cognition 1:14:31 - A Forward-Looking Theory of Content
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Carina Curto is a professor in the Department of Mathematics at The Pennsylvania State University. She uses her background skills in mathematical physics/string theory to study networks of neurons. On this episode, we discuss the world of topology in neuroscience - the study of the geometrical structures mapped out by active populations of neurons. We also discuss her work on "combinatorial linear threshold networks" (CLTNs). Unlike the large deep learning models popular today as models of brain activity, the CLTNs Carina builds are relatively simple, abstracted graphical models. This property is important to Carina, whose goal is to develop mathematically tractable neural network models. Carina has worked out how the structure of many CLTNs allows prediction of the model's allowable dynamics, how motifs of model structure can be embedded in larger models while retaining their dynamical features, and more. The hope is that these elegant models can tell us more about the principles our messy brains employ to generate the robust and beautiful dynamics underlying our cognition.
0:00 - Intro 4:25 - Background: Physics and math to study brains 20:45 - Beautiful and ugly models 35:40 - Topology 43:14 - Topology in hippocampal navigation 56:04 - Topology vs. dynamical systems theory 59:10 - Combinatorial linear threshold networks 1:25:26 - How much more math do we need to invent?
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Jeff Schall is the director of the Center for Visual Neurophysiology at York University, where he runs the Schall Lab. His research centers around studying the mechanisms of our decisions, choices, movement control, and attention within the saccadic eye movement brain systems and in mathematical psychology models- in other words, how we decide where and when to look. Jeff was my postdoctoral advisor at Vanderbilt University, and I wanted to revisit a few guiding principles he instills in all his students. Linking Propositions by Davida Teller are a series of logical statements to ensure we rigorously connect the brain activity we record to the psychological functions we want to explain. Strong Inference by John Platt is the scientific method on steroids - a way to make our scientific practice most productive and efficient. We discuss both of these topics in the context of Jeff's eye movement and decision-making science. We also discuss how neurophysiology has changed over the past 30 years, we compare the relatively small models he employs with the huge deep learning models, many of his current projects, and plenty more. If you want to learn more about Jeff's work and approach, I recommend reading in order two of his review papers we discuss as well. One was written 20 years ago (On Building a Bridge Between Brain and Behavior), and the other 2-ish years ago (Accumulators, Neurons, and Response Time).
0:00 - Intro 6:51 - Neurophysiology old and new 14:50 - Linking propositions 24:18 - Psychology working with neurophysiology 35:40 - Neuron doctrine, population doctrine 40:28 - Strong Inference and deep learning 46:37 - Model mimicry 51:56 - Scientific fads 57:07 - Current projects 1:06:38 - On leaving academia 1:13:51 - How academia has changed for better and worse
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Marc Howard runs his Theoretical Cognitive Neuroscience Lab at Boston University, where he develops mathematical models of cognition, constrained by psychological and neural data. In this episode, we discuss the idea that a Laplace transform and its inverse may serve as a unified framework for memory. In short, our memories are compressed on a continuous log-scale: as memories get older, their representations "spread out" in time. It turns out this kind of representation seems ubiquitous in the brain and across cognitive functions, suggesting it is likely a canonical computation our brains use to represent a wide variety of cognitive functions. We also discuss some of the ways Marc is incorporating this mathematical operation in deep learning nets to improve their ability to handle information at different time scales.
0:00 - Intro 4:57 - Main idea: Laplace transforms 12:00 - Time cells 20:08 - Laplace, compression, and time cells 25:34 - Everywhere in the brain 29:28 - Episodic memory 35:11 - Randy Gallistel's memory idea 40:37 - Adding Laplace to deep nets 48:04 - Reinforcement learning 1:00:52 - Brad Wyble Q: What gets filtered out? 1:05:38 - Replay and complementary learning systems 1:11:52 - Howard Goldowsky Q: Gyorgy Buzsaki 1:15:10 - Obstacles
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Matthew Larkum runs his lab at Humboldt University of Berlin, where his group studies how dendrites contribute to computations within and across layers of the neocortex. Since the late 1990s, Matthew has continued to uncover key properties of the way pyramidal neurons stretch across layers of the cortex, their dendrites receiving inputs from those different layers - and thus different brain areas. For example, layer 5 pyramidal neurons have a set of basal dendrites near the cell body that receives feedforward-like input, and a set of apical dendrites all the way up in layer 1 that receives feedback--like input. Depending on which set of dendrites is receiving input, or neither or both, the neuron's output functions in different modes- silent, regular spiking, or burst spiking. Matthew realized the different sets of dendritic inputs could signal different operations, often pairing feedforward sensory--like signals and feedback context-like signals. His research has shown this kind of coincidence detection is important for cognitive functions like perception, memory, learning, and even wakefulness. We discuss many of his ideas and research findings, why dendrites have long been neglected in favor of neuron cell bodies, the possibility of learning about computations by studying implementation-level phenomena, and much more.
0:00 - Intro 5:31 - Background: Dendrites 23:20 - Cortical neuron bodies vs. branches 25:47 - Theories of cortex 30:49 - Feedforward and feedback hierarchy 37:40 - Dendritic integration hypothesis 44:32 - DIT vs. other consciousness theories 51:30 - Mac Shine Q1 1:04:38 - Are dendrites conceptually useful? 1:09:15 - Insights from implementation level 1:24:44 - How detailed to model? 1:28:15 - Do action potentials cause consciousness? 1:40:33 - Mac Shine Q2
Check out my free video series about what's missing in AI and Neuroscience
Support the show to get full episodes, full archive, and join the Discord community.
Brian Butterworth is Emeritus Professor of Cognitive Neuropsychology at University College London. In his book, Can Fish Count?: What Animals Reveal About Our Uniquely Mathematical Minds, he describes the counting and numerical abilities across many different species, suggesting our ability to count is evolutionarily very old (since many diverse species can count). We discuss many of the examples in his book, the mathematical disability dyscalculia and its relation to dyslexia, how to test counting abilities in various species, how counting may happen in brains, the promise of creating artificial networks that can do math, and many more topics.
0:00 - Intro 3:19 - Why Counting? 5:31 - Dyscalculia 12:06 - Dyslexia 19:12 - Counting 26:37 - Origins of counting vs. language 34:48 - Counting vs. higher math 46:46 - Counting some things and not others 53:33 - How to test counting 1:03:30 - How does the brain count? 1:13:10 - Are numbers real?
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Michel Bitbol is Director of Research at CNRS (Centre National de la Recherche Scientifique). Alex Gomez-Marin is a neuroscientist running his lab, The Behavior of Organisms Laboratory, at the Instituto de Neurociencias in Alicante. We discuss phenomenology as an alternative perspective on our scientific endeavors. Although we like to believe our science is objective and explains the reality of the world we inhabit, we can't escape the fact that all of our scientific knowledge comes through our perceptions and interpretations as conscious living beings. Michel has used phenomenology to resolve many of the paradoxes that quantum mechanics generates when it is understood as a description of reality, and more recently he has applied phenomenology to the philosophy of mind and consciousness. Alex is currently trying to apply the phenomenological approach to his research on brains and behavior. Much of our conversation revolves around how phenomenology and our "normal" scientific explorations can co-exist, including the study of minds, brains, and intelligence- our own and that of other organisms. We also discuss the "blind spot" of science, the history and practice of phenomenology, various kinds of explanation, the language we use to describe things, and more.
0:00 - Intro 4:32 - The Blind Spot 15:53 - Phenomenology and interpretation 22:51 - Personal stories: appreciating phenomenology 37:42 - Quantum physics example 47:16 - Scientific explanation vs. phenomenological description 59:39 - How can phenomenology and science complement each other? 1:08:22 - Neurophenomenology 1:17:34 - Use of language 1:25:46 - Mutual constraints
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Brains are often conceived as consisting of neurons and "everything else." As Elena discusses, the "everything else," including glial cells and in particular astrocytes, have largely been ignored in neuroscience. That's partly because the fast action potentials of neurons have been assumed to underlie computations in the brain, and because technology only recently afforded closer scrutiny of astrocyte activity. Now that we can record calcium signaling in astrocytes, it's possible to relate how astrocyte signaling with each other and with neurons may complement the cognitive roles once thought the sole domain of neurons. Although the computational role of astrocytes remains unclear, it is clear that astrocytes interact with neurons and neural circuits in dynamic and interesting ways. We talk about the historical story of astrocytes, the emerging modern story, and Elena shares her views on the path forward to understand astrocyte function in cognition, disease, homeostasis, and - Elena's favorite current hypothesis - their integrative role in negative feedback control.
0:00 - Intro 5:23 - The changing story of astrocytes 14:58 - Astrocyte research lags neuroscience 19:45 - Types of astrocytes 23:06 - Astrocytes vs neurons 26:08 - Computational roles of astrocytes 35:45 - Feedback control 43:37 - Energy efficiency 46:25 - Current technology 52:58 - Computational astroscience 1:10:57 - Do names for things matter
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Srini is Emeritus Professor at Queensland Brain Institute in Australia. In this episode, he shares his wide range of behavioral experiments elucidating the principles of flight and navigation in insects. We discuss how bees use optic flow signals to determine their speed, distance, proximity to objects, and to gracefully land. These abilities are largely governed via control systems, balancing incoming perceptual signals with internal reference signals. We also talk about a few of the aerial robotics projects his research has inspired, many of the other cognitive skills bees can learn, the possibility of their feeling pain , and the nature of their possible subjective conscious experience.
0:00 - Intro 3:34 - Background 8:20 - Bee experiments 14:30 - Bee flight and navigation 28:05 - Landing 33:06 - Umwelt and perception 37:26 - Bee-inspired aerial robotics 49:10 - Motion camouflage 51:52 - Cognition in bees 1:03:10 - Small vs. big brains 1:06:42 - Pain in bees 1:12:50 - Subjective experience 1:15:25 - Deep learning 1:23:00 - Path forward
Support the show to get full episodes, full archive, and join the Discord community.
Check out my free video series about what's missing in AI and Neuroscience
Ken discusses the recent work in his lab that allows communication with subjects while they experience lucid dreams. This new paradigm opens many avenues to study the neuroscience and psychology of consciousness, sleep, dreams, memory, and learning, and to improve and optimize sleep for cognition. Ken and his team are developing a Lucid Dreaming App which is freely available via his lab. We also discuss much of his work on memory and learning in general and specifically related to sleep, like reactivating specific memories during sleep to improve learning.
0:00 - Intro 2:48 - Background and types of memory 14:44 -Consciousness and memory 23:32 - Phases and sleep and wakefulness 28:19 - Sleep, memory, and learning 33:50 - Targeted memory reactivation 48:34 - Problem solving during sleep 51:50 - 2-way communication with lucid dreamers 1:01:43 - Confounds to the paradigm 1:04:50 - Limitations and future studies 1:09:35 - Lucid dreaming app 1:13:47 - How sleep can inform AI 1:20:18 - Advice for students
I'm releasing my Neuro-AI course April 10-13, after which it will be closed for some time. Learn more here.
Support the show to get full episodes, full archive, and join the Discord community.
Ila discusses her theoretical neuroscience work suggesting how our memories are formed within the cognitive maps we use to navigate the world and navigate our thoughts. The main idea is that grid cell networks in the entorhinal cortex internally generate a structured scaffold, which gets sent to the hippocampus. Neurons in the hippocampus, like the well-known place cells, receive that scaffolding and also receive external signals from the neocortex- signals about what's happening in the world and in our thoughts. Thus, the place cells act to "pin" what's happening in our neocortex to the scaffold, forming a memory. We also discuss her background as a physicist and her approach as a "neurophysicist", and a review she's publishing all about the many brain areas and cognitive functions being explained as attractor landscapes within a dynamical systems framework.
0:00 - Intro 3:36 - "Neurophysicist" 9:30 - Bottom-up vs. top-down 15:57 - Tool scavenging 18:21 - Cognitive maps and hippocampus 22:40 - Hopfield networks 27:56 - Internal scaffold 38:42 - Place cells 43:44 - Grid cells 54:22 - Grid cells encoding place cells 59:39 - Scaffold model: stacked hopfield networks 1:05:39 - Attractor landscapes 1:09:22 - Landscapes across scales 1:12:27 - Dimensionality of landscapes
Support the show to get full episodes, full archive, and join the Discord community.
Sri and Mei join me to discuss how including principles of neuromodulation in deep learning networks may improve network performance. It's an ever-present question how much detail to include in models, and we are in the early stages of learning how neuromodulators and their interactions shape biological brain function. But as we continue to learn more, Sri and Mei are interested in building "neuromodulation-aware DNNs".
0:00 - Intro 3:10 - Background 9:19 - Bottom-up vs. top-down 14:42 - Levels of abstraction 22:46 - Biological neuromodulation 33:18 - Inventing neuromodulators 41:10 - How far along are we? 53:31 - Multiple realizability 1:09:40 -Modeling dendrites 1:15:24 - Across-species neuromodulation
Support the show to get full episodes, full archive, and join the Discord community.
Eve discusses many of the lessons she has learned studying a small nervous system, the crustacean stomatogastric nervous system (STG). The STG has only about 30 neurons and its connections and neurophysiology are well-understood. Yet Eve's work has shown it functions under a remarkable diversity of conditions, and does so is a remarkable variety of ways. We discuss her work on the STG specifically, and what her work implies about trying to study much larger nervous systems, like our human brains.
0:00 - Intro 3:58 - Background 8:00 - Levels of ambiguity 9:47 - Stomatogastric nervous system 17:13 - Structure vs. function 26:08 - Role of theory 34:56 - Technology vs. understanding 38:25 - Higher cognitive function 44:35 - Adaptability, resilience, evolution 50:23 - Climate change 56:11 - Deep learning 57:12 - Dynamical systems
Support the show to get full episodes, full archive, and join the Discord community.
Patryk and I discuss his wide-ranging background working in both the neuroscience and AI worlds, and his resultant perspective on what's needed to move forward in AI, including some principles of brain processes that are more and less important. We also discuss his own work using some of those principles to help deep learning generalize to better capture how humans behave in and perceive the world.
0:00 - Intro 2:22 - Patryk's background 8:37 - Importance of diverse skills 16:14 - What is intelligence? 20:34 - Important brain principles 22:36 - Learning from the real world 35:09 - Language models 42:51 - AI contribution to neuroscience 48:22 - Criteria for "real" AI 53:11 - Neuroscience for AI 1:01:20 - What can we ignore about brains? 1:11:45 - Advice to past self
Support the show to get full episodes, full archive, and join the Discord community.
Hakwan and I discuss many of the topics in his new book, In Consciousness we Trust: The Cognitive Neuroscience of Subjective Experience. Hakwan describes his perceptual reality monitoring theory of consciousness, which suggests consciousness may act as a systems check between our sensory perceptions and higher cognitive functions. We also discuss his latest thoughts on mental quality space and how it relates to perceptual reality monitoring. Among many other topics, we chat about the many confounds and challenges to empirically studying consciousness, a topic featured heavily in the first half of his book. Hakwan was on a previous episode with Steve Fleming, BI 099 Hakwan Lau and Steve Fleming: Neuro-AI Consciousness.
0:00 - Intro 4:37 - In Consciousness We Trust 12:19 - Too many consciousness theories? 19:26 - Philosophy and neuroscience of consciousness 29:00 - Local vs. global theories 31:20 - Perceptual reality monitoring and GANs 42:43 - Functions of consciousness 47:17 - Mental quality space 56:44 - Cognitive maps 1:06:28 - Performance capacity confounds 1:12:28 - Blindsight 1:19:11 - Philosophy vs. empirical work
Support the show to get full episodes, full archive, and join the Discord community.
Tomás and I discuss his research and ideas on how memories are encoded (the engram), the role of forgetting, and the overlapping mechanisms of memory and instinct. Tomás uses otpogenetics and other techniques to label and control neurons involved in learning and memory, and has shown that forgotten memories can be restored by stimulating "engram cells" originally associated with the forgotten memory. This line of research has led Tomás to think forgetting might be a learning mechanism itself, a adaption our brains make based on the predictability and affordances of the environment. His work on engrams has also led Tomás to think our instincts (ingrams) may share the same mechanism of our memories (engrams), and that memories may transition to instincts across generations. We begin by addressing Randy Gallistel's engram ideas from the previous episode: BI 126 Randy Gallistel: Where Is the Engram?
0:00 - Intro 4:05 - Response to Randy Gallistel 10:45 - Computation in the brain 14:52 - Instinct and memory 19:37 - Dynamics of memory 21:55 - Wiring vs. connection strength plasticity 24:16 - Changing one's mind 33:09 - Optogenetics and memory experiments 47:24 - Forgetting as learning 1:06:35 - Folk psychological terms 1:08:49 - Memory becoming instinct 1:21:49 - Instinct across the lifetime 1:25:52 - Boundaries of memories 1:28:52 - Subjective experience of memory 1:31:58 - Interdisciplinary research 1:37:32 - Communicating science
Support the show to get full episodes, full archive, and join the Discord community.
Randy and I discuss his long-standing interest in how the brain stores information to compute. That is, where is the engram, the physical trace of memory in the brain? Modern neuroscience is dominated by the view that memories are stored among synaptic connections in populations of neurons. Randy believes a more reasonable and reliable way to store abstract symbols, like numbers, is to write them into code within individual neurons. Thus, the spiking code, whatever it is, functions to write and read memories into and out of intracellular substrates, like polynucleotides (DNA, RNA, e.g.). He lays out his case in detail in his book with Adam King, Memory and the Computational Brain: Why Cognitive Science will Transform Neuroscience. We also talk about some research and theoretical work since then that support his views.
0:00 - Intro 6:50 - Cognitive science vs. computational neuroscience 13:23 - Brain as computing device 15:45 - Noam Chomsky's influence 17:58 - Memory must be stored within cells 30:58 - Theoretical support for the idea 34:15 - Cerebellum evidence supporting the idea 40:56 - What is the write mechanism? 51:11 - Thoughts on deep learning 1:00:02 - Multiple memory mechanisms? 1:10:56 - The role of plasticity 1:12:06 - Trying to convince molecular biologists
Support the show to get full episodes, full archive, and join the Discord community.
Doris, Tony, and Blake are the organizers for this year's NAISys conference, From Neuroscience to Artificially Intelligent Systems (NAISys), at Cold Spring Harbor. We discuss the conference itself, some history of the neuroscience and AI interface, their current research interests, and a handful of topics around evolution, innateness, development, learning, and the current and future prospects for using neuroscience to inspire new ideas in artificial intelligence.
0:00 - Intro 4:16 - Tony Zador 5:38 - Doris Tsao 10:44 - Blake Richards 15:46 - Deductive, inductive, abductive inference 16:32 - NAISys 33:09 - Evolution, development, learning 38:23 - Learning: plasticity vs. dynamical structures 54:13 - Different kinds of understanding 1:03:05 - Do we understand evolution well enough? 1:04:03 - Neuro-AI fad? 1:06:26 - Are your problems bigger or smaller now?
Support the show to get full episodes, full archive, and join the Discord community.
Robin and I discuss many of the ideas in his book The Self-Assembling Brain: How Neural Networks Grow Smarter. The premise is that our DNA encodes an algorithmic growth process that unfolds information via time and energy, resulting in a connected neural network (our brains!) imbued with vast amounts of information from the "start". This contrasts with modern deep learning networks, which start with minimal initial information in their connectivity, and instead rely almost solely on learning to gain their function. Robin suggests we won't be able to create anything with close to human-like intelligence unless we build in an algorithmic growth process and an evolutionary selection process to create artificial networks.
0:00 - Intro 3:01 - The Self-Assembling Brain 21:14 - Including growth in networks 27:52 - Information unfolding and algorithmic growth 31:27 - Cellular automata 40:43 - Learning as a continuum of growth 45:01 - Robustness, autonomous agents 49:11 - Metabolism vs. connectivity 58:00 - Feedback at all levels 1:05:32 - Generality vs. specificity 1:10:36 - Whole brain emulation 1:20:38 - Changing view of intelligence 1:26:34 - Popular and wrong vs. unknown and right
Support the show to get full episodes, full archive, and join the Discord community.
Irina is a faculty member at MILA-Quebec AI Institute and a professor at Université de Montréal. She has worked from both ends of the neuroscience/AI interface, using AI for neuroscience applications, and using neural principles to help improve AI. We discuss her work on biologically-plausible alternatives to back-propagation, using "auxiliary variables" in addition to the normal connection weight updates. We also discuss the world of lifelong learning, which seeks to train networks in an online manner to improve on any tasks as they are introduced. Catastrophic forgetting is an obstacle in modern deep learning, where a network forgets old tasks when it is trained on new tasks. Lifelong learning strategies, like continual learning, transfer learning, and meta-learning seek to overcome catastrophic forgetting, and we talk about some of the inspirations from neuroscience being used to help lifelong learning in networks.
0:00 - Intro 3:26 - AI for Neuro, Neuro for AI 14:59 - Utility of philosophy 20:51 - Artificial general intelligence 24:34 - Back-propagation alternatives 35:10 - Inductive bias vs. scaling generic architectures 45:51 - Continual learning 59:54 - Neuro-inspired continual learning 1:06:57 - Learning trajectories
Support the show to get full episodes and join the Discord community.
Ko and I discuss a range of topics around his work to understand our visual intelligence. Ko was a postdoc in James Dicarlo's lab, where he helped develop the convolutional neural network models that have become the standard for explaining core object recognition. He is starting his own lab at York University, where he will continue to expand and refine the models, adding important biological details and incorporating models for brain areas outside the ventral visual stream. He will also continue recording neural activity, and performing perturbation studies to better understand the networks involved in our visual cognition.
0:00 - Intro 3:49 - Background 13:51 - Where are we in understanding vision? 19:46 - Benchmarks 21:21 - Falsifying models 23:19 - Modeling vs. experiment speed 29:26 - Simple vs complex models 35:34 - Dorsal visual stream and deep learning 44:10 - Modularity and brain area roles 50:58 - Chemogenetic perturbation, DREADDs 57:10 - Future lab vision, clinical applications 1:03:55 - Controlling visual neurons via image synthesis 1:12:14 - Is it enough to study nonhuman animals? 1:18:55 - Neuro/AI intersection 1:26:54 - What is intelligence?
Support the show to get full episodes, full archive, and join the Discord community.
Mac and I discuss his systems level approach to understanding brains, and his theoretical work suggesting important roles for the thalamus, basal ganglia, and cerebellum, shifting the dynamical landscape of brain function within varying behavioral contexts. We also discuss his recent interest in the ascending arousal system and neuromodulators. Mac thinks the neocortex has been the sole focus of too much neuroscience research, and that the subcortical brain regions and circuits have a much larger role underlying our intelligence.
0:00 - Intro 6:32 - Background 10:41 - Holistic approach 18:19 - Importance of thalamus 35:19 - Thalamus circuitry 40:30 - Cerebellum 46:15 - Predictive processing 49:32 - Brain as dynamical attractor landscape 56:48 - System 1 and system 2 1:02:38 - How to think about the thalamus 1:06:45 - Causality in complex systems 1:11:09 - Clinical applications 1:15:02 - Ascending arousal system and neuromodulators 1:27:48 - Implications for AI 1:33:40 - Career serendipity 1:35:12 - Advice
Support the show to get full episodes, full archive, and join the Discord community.
James, Andrew, and Weinan discuss their recent theory about how the brain might use complementary learning systems to optimize our memories. The idea is that our hippocampus creates our episodic memories for individual events, full of particular details. And through a complementary process, slowly consolidates those memories within our neocortex through mechanisms like hippocampal replay. The new idea in their work suggests a way for the consolidated cortical memory to become optimized for generalization, something humans are known to be capable of but deep learning has yet to build. We discuss what their theory predicts about how the "correct" process depends on how much noise and variability there is in the learning environment, how their model solves this, and how it relates to our brain and behavior.
0:00 - Intro 3:57 - Guest Intros 15:04 - Organizing memories for generalization 26:48 - Teacher, student, and notebook models 30:51 - Shallow linear networks 33:17 - How to optimize generalization 47:05 - Replay as a generalization regulator 54:57 - Whole greater than sum of its parts 1:05:37 - Unpredictability 1:10:41 - Heuristics 1:13:52 - Theoretical neuroscience for AI 1:29:42 - Current personal thinking
Support the show to get full episodes, full archive, and join the Discord community.
Henry and I discuss why he thinks neuroscience is in a crisis (in the Thomas Kuhn sense of scientific paradigms, crises, and revolutions). Henry thinks our current concept of the brain as an input-output device, with cognition in the middle, is mistaken. He points to the failure of neuroscience to successfully explain behavior despite decades of research. Instead, Henry proposes the brain is one big hierarchical set of control loops, trying to control their output with respect to internally generated reference signals. He was inspired by control theory, but points out that most control theory for biology is flawed by not recognizing that the reference signals are internally generated. Instead, most control theory approaches, and neuroscience research in general, assume the reference signals are what gets externally supplied... by the experimenter.
0:00 - Intro 5:40 - Kuhnian crises 9:32 - Control theory and cybernetics 17:23 - How much of brain is control system? 20:33 - Higher order control representation 23:18 - Prediction and control theory 27:36 - The way forward 31:52 - Compatibility with mental representation 38:29 - Teleology 45:53 - The right number of subjects 51:30 - Continuous measurement 57:06 - Artificial intelligence and control theory
Support the show to get full episodes, full archive, and join the Discord community.
Johannes (Yogi) is a freelance philosopher, researcher & educator. We discuss many of the topics in his online course, Beyond Networks: The Evolution of Living Systems. The course is focused on the role of agency in evolution, but it covers a vast range of topics: process vs. substance metaphysics, causality, mechanistic dynamic explanation, teleology, the important role of development mediating genotypes, phenotypes, and evolution, what makes biological organisms unique, the history of evolutionary theory, scientific perspectivism, and a view toward the necessity of including agency in evolutionary theory. I highly recommend taking his course. We also discuss the role of agency in artificial intelligence, how neuroscience and evolutionary theory are undergoing parallel re-evaluations, and Yogi answers a guest question from Kevin Mitchell.
0:00 - Intro 4:10 - Yogi's background 11:00 - Beyond Networks - limits of dynamical systems models 16:53 - Kevin Mitchell question 20:12 - Process metaphysics 26:13 - Agency in evolution 40:37 - Agent-environment interaction, open-endedness 45:30 - AI and agency 55:40 - Life and intelligence 59:08 - Deep learning and neuroscience 1:03:21 - Mental autonomy 1:06:10 - William Wimsatt's biopsychological thicket 1:11:23 - Limtiations of mechanistic dynamic explanation 1:18:53 - Synthesis versus multi-perspectivism 1:30:31 - Specialization versus generalization
Support the show to get full episodes, full archive, and join the Discord community.
Anil and I discuss a range of topics from his book, BEING YOU A New Science of Consciousness. Anil lays out his framework for explaining consciousness, which is embedded in what he calls the "real problem" of consciousness. You know the "hard problem", which was David Chalmers term for our eternal difficulties to explain why we have subjective awareness at all instead of being unfeeling, unexperiencing machine-like organisms. Anil's "real problem" aims to explain, predict, and control the phenomenal properties of consciousness, and his hope is that, by doing so, the hard problem of consciousness will dissolve much like the mystery of explaining life dissolved with lots of good science.
Anil's account of perceptual consciousness, like seeing red, is that it's rooted in predicting our incoming sensory data. His account of our sense of self, is that it's rooted in predicting our bodily states to control them.
We talk about that and a lot of other topics from the book, like consciousness as "controlled hallucinations", free will, psychedelics, complexity and emergence, and the relation between life, intelligence, and consciousness. Plus, Anil answers a handful of questions from Megan Peters and Steve Fleming, both previous brain inspired guests.
0:00 - Intro 6:32 - Megan Peters Q: Communicating Consciousness 15:58 - Human vs. animal consciousness 19:12 - BEING YOU A New Science of Consciousness 20:55 - Megan Peters Q: Will the hard problem go away? 30:55 - Steve Fleming Q: Contents of consciousness 41:01 - Megan Peters Q: Phenomenal character vs. content 43:46 - Megan Peters Q: Lempels of complexity 52:00 - Complex systems and emergence 55:53 - Psychedelics 1:06:04 - Free will 1:19:10 - Consciousness vs. life vs. intelligence
Support the show to get full episodes, full archive, and join the Discord community.
Mike and I discuss his modeling approach to study cognition. Many people I have on the podcast use deep neural networks to study brains, where the idea is to train or optimize the model to perform a task, then compare the model properties with brain properties. Mike's approach is different in at least two ways. One, he builds the architecture of his models using connectivity data from fMRI recordings. Two, he doesn't train his models; instead, he uses functional connectivity data from the fMRI recordings to assign weights between nodes of the network (in deep learning, the weights are learned through lots of training). Mike calls his networks empirically-estimated neural networks (ENNs), and/or network coding models. We walk through his approach, what we can learn from models like ENNs, discuss some of his earlier work on cognitive control and our ability to flexibly adapt to new task rules through instruction, and he fields questions from Kanaka Rajan, Kendrick Kay, and Patryk Laurent.
0:00 - Intro 4:58 - Cognitive control 7:44 - Rapid Instructed Task Learning and Flexible Hub Theory 15:53 - Patryk Laurent question: free will 26:21 - Kendrick Kay question: fMRI limitations 31:55 - Empirically-estimated neural networks (ENNs) 40:51 - ENNs vs. deep learning 45:30 - Clinical relevance of ENNs 47:32 - Kanaka Rajan question: a proposed collaboration 56:38 - Advantage of modeling multiple regions 1:05:30 - How ENNs work 1:12:48 - How ENNs might benefit artificial intelligence 1:19:04 - The need for causality 1:24:38 - Importance of luck and serendipity
Support the show to get full episodes, full archive, and join the Discord community.
Steve and I discuss his book Conscious Mind, Resonant Brain: How Each Brain Makes a Mind. The book is a huge collection of his models and their predictions and explanations for a wide array of cognitive brain functions. Many of the models spring from his Adaptive Resonance Theory (ART) framework, which explains how networks of neurons deal with changing environments while maintaining self-organization and retaining learned knowledge. ART led Steve to the hypothesis that all conscious states are resonant states, which we discuss. There are also guest questions from György Buzsáki, Jay McClelland, and John Krakauer.
0:00 - Intro 2:38 - Conscious Mind, Resonant Brain 11:49 - Theoretical method 15:54 - ART, learning, and consciousness 22:58 - Conscious vs. unconscious resonance 26:56 - Györy Buzsáki question 30:04 - Remaining mysteries in visual system 35:16 - John Krakauer question 39:12 - Jay McClelland question 51:34 - Any missing principles to explain human cognition? 1:00:16 - Importance of an early good career start 1:06:50 - Has modeling training caught up to experiment training? 1:17:12 - Universal development code
Support the show to get full episodes, full archive, and join the Discord community.
Mark and Mazviita discuss the philosophy and science of mind, and how to think about computations with respect to understanding minds. Current approaches to explaining brain function are dominated by computational models and the computer metaphor for brain and mind. But there are alternative ways to think about the relation between computations and brain function, which we explore in the discussion. We also talk about the role of philosophy broadly and with respect to mind sciences, pluralism and perspectival approaches to truth and understanding, the prospects and desirability of naturalizing representations (accounting for how brain representations relate to the natural world), and much more.
0:00 - Intro 5:26 - Philosophy contributing to mind science 15:45 - Trend toward hyperspecialization 21:38 - Practice-focused philosophy of science 30:42 - Computationalism 33:05 - Philosophy of mind: identity theory, functionalism 38:18 - Computations as descriptions 41:27 - Pluralism and perspectivalism 54:18 - How much of brain function is computation? 1:02:11 - AI as computationalism 1:13:28 - Naturalizing representations 1:30:08 - Are you doing it right?
Support the show to get full episodes, full archive, and join the Discord community.
David and John discuss some of the concepts from their recent paper Two Views on the Cognitive Brain, in which they argue the recent population-based dynamical systems approach is a promising route to understanding brain activity underpinning higher cognition. We discuss mental representations, the kinds of dynamical objects being used for explanation, and much more, including David's perspectives as a practicing neuroscientist and philosopher.
Timestamps
0:00 - Intro 3:13 - David's philosophy and neuroscience experience 20:01 - Renaissance person 24:36 - John's medical training 31:58 - Two Views on the Cognitive Brain 44:18 - Representation 49:37 - Studying populations of neurons 1:05:17 - What counts as representation 1:18:49 - Does this approach matter for AI?
Ben has started his new lab and is recruiting grad students.
Check out his lab here and apply!
Ali and Ben discuss the ever-expanding discoveries about the roles dopamine plays for our cognition. Dopamine is known to play a role in learning – dopamine (DA) neurons fire when our reward expectations aren’t met, and that signal helps adjust our expectation. Roughly, DA corresponds to a reward prediction error. The reward prediction error has helped reinforcement learning in AI develop into a raging success, specially with deep reinforcement learning models trained to out-perform humans in games like chess and Go. But DA likely contributes a lot more to brain function. We discuss many of those possible roles, how to think about computation with respect to neuromodulators like DA, how different time and spatial scales interact, and more.
Dopamine: A Simple AND Complex Story
Timestamps:
0:00 – Intro 5:02 – Virtual Dopamine Conference 9:56 – History of dopamine’s roles 16:47 – Dopamine circuits 21:13 – Multiple roles for dopamine 31:43 – Deep learning panel discussion 50:14 – Computation and neuromodulationThis is the 5th in a series of panel discussions in collaboration with Neuromatch Academy, the online computational neuroscience summer school. This is the 2nd of 3 in the deep learning series. In this episode, the panelists discuss their experiences “doing more with fewer parameters: Convnets, RNNs, attention & transformers, generative models (VAEs & GANs).
The other panels:
This is the 4th in a series of panel discussions in collaboration with Neuromatch Academy, the online computational neuroscience summer school. This is the first of 3 in the deep learning series. In this episode, the panelists discuss their experiences with some basics in deep learning, including Linear deep learning, Pytorch, multi-layer-perceptrons, optimization, & regularization.
Timestamps:
Erik, Kevin, and I discuss... well a lot of things.
Erik's recent novel The Revelations is a story about a group of neuroscientists trying to develop a good theory of consciousness (with a murder mystery plot).
Kevin's book Innate - How the Wiring of Our Brains Shapes Who We Are describes the messy process of getting from DNA, traversing epigenetics and development, to our personalities.
We talk about both books, then dive deeper into topics like whether brains evolved for moving our bodies vs. consciousness, how information theory is lending insights to emergent phenomena, and the role of agency with respect to what counts as intelligence.
Timestamps
0:00 - Intro 3:28 - The Revelations - Erik's novel 15:15 - Innate - Kevin's book 22:56 - Cycle of progress 29:05 - Brains for movement or consciousness? 46:46 - Freud's influence 59:18 - Theories of consciousness 1:02:02 - Meaning and emergence 1:05:50 - Reduction in neuroscience 1:23:03 - Micro and macro - emergence 1:29:35 - Agency and intelligence
Panelists:
This is the third in a series of panel discussions in collaboration with Neuromatch Academy, the online computational neuroscience summer school. In this episode, the panelists discuss their experiences with stochastic processes, including Bayes, decision-making, optimal control, reinforcement learning, and causality.
The other panels:
Panelists:
This is the second in a series of panel discussions in collaboration with Neuromatch Academy, the online computational neuroscience summer school. In this episode, the panelists discuss their experiences with linear systems, real neurons, and dynamic networks.
Other panels:
Panelists:
This is the first in a series of panel discussions in collaboration with Neuromatch Academy, the online computational neuroscience summer school. In this episode, the panelists discuss their experiences with model fitting, GLMs/machine learning, dimensionality reduction, and deep learning.
Other panels:
Catherine, Jess, and I use some of the ideas from their recent papers to discuss how different types of explanations in neuroscience and AI could be unified into explanations of intelligence, natural or artificial. Catherine has written about how models are related to the target system they are built to explain. She suggests both the model and the target system should be considered as instantiations of a specific kind of phenomenon, and explanation is a product of relating the model and the target system to that specific aspect they both share. Jess has suggested we shift our focus of explanation from objects - like a brain area or a deep learning model - to the shared class of phenomenon performed by those objects. Doing so may help bridge the gap between the different forms of explanation currently used in neuroscience and AI. We also discuss Henk de Regt's conception of scientific understanding and its relation to explanation (they're different!), and plenty more.
Timestamps: 0:00 - Intro 11:11 - Background and approaches 27:00 - Understanding distinct from explanation 36:00 - Explanations as programs (early explanation) 40:42 - Explaining classes of phenomena 52:05 - Constitutive (neuro) vs. etiological (AI) explanations 1:04:04 - Do nonphysical objects count for explanation? 1:10:51 - Advice for early philosopher/scientists
En liten tjänst av I'm With Friends. Finns även på engelska.