James Mac Shine, Professor of Systems Neuroscience, University of Sydney.
Philosopher Andrea Hiott talks with neuroscientist Mac Shine about how the brain works. Shine works in systems neuroscience, which tries to describe the brain at many scales at once instead of focusing only on single neurons or only on whole-brain images. He explains that the field does not yet have simple, agreed-upon explanations. Part of the reason is that researchers from different backgrounds (for example, physicists and biologists) often use the same words to mean different things.
Most of the conversation is about a paper Shine wrote with his father in 2014. The paper argues that an important change in human evolution was the ability to hand off, or “delegate,” repeated tasks to lower parts of the brain, especially the cerebellum. Learning to walk upright is used as the main example. Once a task becomes automatic, attention is freed up for other things.
Shine and Hiott then apply this idea to driving a car, playing chess, and playing sports. They argue that thinking and body movement are not two separate systems but one connected process. They also discuss what makes human expertise different from simple repetition, the limits of AI tools for real learning, and the value of curiosity, care, and teamwork in running a research lab.
Runtime for this show is about two hours.
Find Mac at https://macshine.github.io/publications/
Mac’s Lab: https://profiles.sydney.edu.au/mac.shine
The Brain Inspired conversation Andrea mentions.
Mac Shine on Systems Neuroscience, Automaticity, and the Cerebellum’s Role in Human CognitionUniversity of Sydney systems neuroscientist Mac Shine in conversation about systems neuroscience and the challenge of linking single-neuron explanations with whole-brain network accounts. Shine describes his path from evolutionary biology influences and medical training to a search for simplified, cross-scale principles in brain science, noting how terminology like “representation” or “integration” can hinder communication across disciplines. They discuss his 2014 paper “Delegation to Automaticity,” inspired by observing his son learn to walk, which argues that cerebellar architecture supports the delegation of difficult skills into automatic control, freeing resources for higher-level cognition and potentially shaping human cognitive evolution alongside bipedalism. Shine contrasts basal ganglia habit findings with cerebellar predictive control, uses driving and chess to illustrate deliberate versus automatic processes, and emphasizes embodiment, learning through struggle, and collaborative, caring lab culture.00:00 Welcome and Guest Intro00:27 Automaticity and Evolution Teaser02:48 Why This Conversation Matters05:02 What Is Systems Neuroscience10:24 Mac’s Path Into Systems Thinking18:42 Language Gaps Across Fields26:15 Are We Assuming Cognition33:35 Delegation to Automaticity Paper37:19 Cerebellum as Automation Engine46:09 Bipedalism and Coevolving Mind Body47:29 Bipedalism And Cognition50:05 Natural Selection Good Enough52:37 Frozen Accidents And Delegation55:06 Humans Adapt Not Excel58:33 Repetition To Automaticity01:00:19 Habit Versus Automatic01:08:18 Driving And Chess Layers01:16:18 Beyond Motor Cognition01:21:57 Flow States And Expertise01:27:26 Scientific Humility Ethos01:29:16 Curiosity Over Credit01:30:57 AI And The Value Of Being Wrong01:33:07 Learning Through Struggle01:37:59 Embodiment And Driving01:40:56 Cerebellum Body Brain Coupling01:48:14 Expertise Through Variation01:53:31 Lab Curiosity And Next Questions01:59:49 Care Love And Leadership02:04:44 Closing Reflections
Approximate chapters of video without A’s Intro
* 00:00 — Introductions; what “systems neuroscience” means
* 04:00 — Why the field is still young; the single-neuron vs. whole-brain problem
* 05:00 — Shine’s background: his father the evolutionary biologist, medical school, systems thinking
* 13:00 — Why brain words (”thought,” “representation,” “information”) cause confusion across fields
* 20:00 — The “cognitive-behaviorist” problem: naming a function, then hunting for it
* 28:00 — The main topic: the 2014 “Delegation to Automaticity” paper
* 31:00 — The origin story (his son Tyler learning to walk; his father’s question in the park)
* 33:00 — Why the basal ganglia story didn’t satisfy him; the cerebellum as a better fit
* 44:00 — Evolution as “good enough” solutions; the inefficient human eye; frozen accidents
* 1:03:00 — Learning to drive and playing chess as examples of automaticity
* 1:11:00 — Counting without moving: cognition without overt action
* 1:17:00 — Flow, expertise, and knowing when to switch between automatic and deliberate
* 1:22:00 — Running a lab; curiosity, humility, “wanting to be wrong”
* 1:26:00 — AI, large language models, and why process matters more than answers
* 1:33:00 — The body and brain as one coupled system (cerebellum experiments)
* 1:45:00 — Expertise as adaptability, not exact repetition (Schöllhorn’s training method)
* 1:54:00 — Care, love, and community in science
The central paper
“Delegation to automaticity: the driving force for cognitive evolution?” J. M. Shine & R. Shine (2014), Frontiers in Neuroscience 8:90.
* Open-access full text (PMC): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4010745/
* DOI: https://doi.org/10.3389/fnins.2014.00090
* PDF on the author’s site: https://macshine.github.io/publications/2014_delegation.pdf
Core idea: humans became distinctive not by being “better” than other animals but by being able to transfer well-practiced tasks from higher brain centers to lower ones (notably the cerebellum). The huge challenge of balancing on two legs may have favored brains good at this kind of “delegation,” which then freed up capacity for higher-level thought.
Mac’s work and his lab
* Mac Shine — ShineLab:
https://shine-lab.org/
* Lab members (”People”): https://shine-lab.org/people/
* Full publication list (with PDFs):
https://macshine.github.io/
Lab papers mentioned in the episode:
* The “multiple species” / renormalization paper (Brandon Munn’s project): Munn et al. (2024), “Multiscale organization of neuronal activity unifies scale-dependent theories of brain function,” Cell. Borrows the physics idea of renormalization and finds a consistent scaling pattern across five very different animals (zebrafish, C. elegans, fruit fly, mouse, macaque). https://www.cell.com/cell/fulltext/S0092-8674(24)01152-8
* Arousal / neuronal “burst cascades”: Munn et al. (2023), “Neuronal connected burst cascades bridge macroscale adaptive signatures across arousal states,” Nature Communications. https://www.nature.com/articles/s41467-023-42465-2
* The thalamus / arousal modelling line of work (Eli Müller): Müller, Munn & Shine (2020), “Diffuse neural coupling mediates complex network dynamics through the formation of quasi-critical brain states,” Nature Communications. https://www.nature.com/articles/s41467-020-19716-7 · Eli Müller’s full list (incl. the brain-temperature/metabolic work referenced): https://scholar.google.com/citations?user=h1gl_ckAAAAJ
* Neuromodulation and large-scale dynamics: Shine et al. (2021), “Computational models link cellular mechanisms of neuromodulation to large-scale neural dynamics,” Nature Neuroscience. https://www.nature.com/articles/s41593-021-00824-6
People, papers, and books discussed
* György Buzsáki — raised by Hiott regarding the “cognitive-behaviorist” habit of naming a mental function (like memory) and then searching the brain for it. See his book The Brain from Inside Out (Oxford University Press, 2019). https://global.oup.com/academic/product/the-brain-from-inside-out-9780190905385
* Germund Hesslow — his paper on doing “cognitive work” without any overt movement (e.g., a child learning to count and then counting silently). Hesslow, G. (2002), “Conscious thought as simulation of behaviour and perception,” Trends in Cognitive Sciences 6(6):242–247. https://doi.org/10.1016/S1364-6613(02)01913-7 · Full text: https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(02)01913-7
* Matthew Larkum — Shine’s collaborator (lab at Humboldt-Universität / Charité, Berlin) on how single cortical neurons integrate signals; the basis of the Dendritic Integration Theory of consciousness (Aru, Suzuki & Larkum, 2020). Overview: https://www.humanbrainproject.eu/en/follow-hbp/news/2023/01/13/redwoods-cortex/ · Theory paper: https://philosophymindscience.org/index.php/phimisci/article/view/8946
* Wolfgang Schöllhorn — the sports scientist whose Differential Learning method trains athletes with constant variation instead of identical repetition. Meta-analytic review: https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2021.533033/full
* Nuo Li — collaborator whose mouse experiments silenced the cerebellum during a delayed-response task and disrupted short-term memory. Accessible summary: https://blogs.bcm.edu/2018/11/15/the-cerebellum-more-than-just-muscle-control/
* Wade Regehr (Harvard) — collaborator whose cerebellar-inhibition experiments were paired with markerless video pose-estimation of body movement (mice on a “rotarod”). Regehr Lab: https://regehrlab.hms.harvard.edu/
* Richard (”Rick”) Shine — Mac’s father, an evolutionary biologist/herpetologist, now at Macquarie University; co-author of the delegation paper.
* Karl Friston — mentioned as an influence Hiott has also interviewed; known for the free-energy principle and predictive processing.
* Marc Tittgemeyer — the colleague who pointed Shine to the myth of Prometheus.
* Valentino Braitenberg — cited for the 1990 argument that the cerebral cortex is essentially an associative memory network (also known for Vehicles: Experiments in Synthetic Psychology).
* Jeremy Schmahmann — referenced for cognitive dysmetria and cognitive/affective symptoms after cerebellar damage.
* The Santa Fe Institute and the idea of a “frozen accident” (a term associated with Murray Gell-Mann). https://www.santafe.edu/
Concepts and ideas referenced
* Systems neuroscience — studying the brain across scales rather than one neuron or one brain image at a time.
* The scale problem — reconciling single-neuron recordings with whole-brain methods like EEG, MEG, and fMRI.
* Basal ganglia vs. cerebellum — the standard “habit” story centers on the basal ganglia; Shine argues the cerebellum fits automaticity better.
* Habitual vs. goal-directed and automatic vs. deliberate — two separate axes that can combine in different ways.
* Embodied, extended, and distributed cognition — thinking is not sealed inside the head; external “scaffolding” (language, tools, the car, the road) does some of the work.
* Ecological psychology / affordances — J. J. Gibson’s tradition.
* Representation-hungry problems — the philosophical worry (Clark & Toribio) about tasks like imagining or planning.
* System 1 / System 2 — Kahneman’s fast/slow framing (Thinking, Fast and Slow).
* Predictive processing — the brain constantly anticipates the next state and acts on prediction errors.
* Dynamical systems / basin of attraction — expertise as a wide, robust “river delta” rather than a single fragile point.
Also mentioned in passing: Spinoza (”God, or Nature”); Merleau-Ponty (phenomenology of the body); Kurt Vonnegut (the hippopotamus joke); J. R. R. Tolkien (The Hobbit, The Lord of the Rings, Tom Bombadil); Claude Code (as an AI tool that saves time but can make knowledge feel “flimsy”).
TRANSCRIPT
Mac Shine on Love and Philosophy: Brain Dynamics
Love and Philosophy — Beyond Dichotomy · Host: Andrea Hiott · Guest: James M. (Mac) Shine
Lightly edited for readability: timestamps, filler words, false starts, and backchannel interjections removed; wording otherwise preserved. Misheard proper names corrected against published sources.
Andrea Hiott: Okay. Hi, Mac. Welcome to Love and Philosophy. Thank you for joining us today.
Mac Shine: Hi, Andrea. Thanks so much for having me.
Andrea Hiott: So I’ve been following your work for a while and, one thing I guess I wanna introduce to people who might not know your work is the idea of systems neuroscience. If you just had to give a kinda general idea of what that is in relation to neuroscience what would you tell people who’ve maybe never heard of it before?
Mac Shine: Yeah, it’s a, it’s a tough one to define, succinctly and it’s a broad church as well. You often find quite different people calling themselves systems neuroscientists. I think to me what it means is that you’re making a commitment to try to look at the brain from a variety of different perspectives and really to try to capture as much of the brain in your explanations of behavior as possible. So one of the challenges we’ve faced as a field of neuroscience, which I think is really still quite a young field, no matter how you look at it, we feel like we’re old because there’s so much rich history so many Nobel Prizes that have been won but we’re still really even just figuring out all the parts of the brain. What are all the different neural types, what are all the different glial types, how are they connected together? We don’t really have a simplified story about how the system works. And part of the challenge, I think, has been that because it’s so difficult to record from a brain, you’ve got to really make a choice when you go down to understand, your question, your neuroscience question. And so you can start at the animal level and you can, basically for a long time you were forced to make a choice to study a particular neuron or neural population and let’s say stick a probe into the brain of some part, let’s say the hippocampus and just get all the neural spikes you could and do whatever you could to try to understand that spiking behavior. You’d throw out things that didn’t quite fit. You’d select neurons that responded to the stimulus or something and you’d end up with a smaller and smaller subset. And then at the end of it, you’d want to make this really strong claim about the behavior that you were interested in. Let’s say it’s navigation or something. At the end of the day, you’re really tempted to say, “This neuron is doing the navigation.” You want to make this link between this neuron and this behavior. But they’re really different scales of understanding of a system. If you in contrast, you’re a human imager. Let’s say you’ve got access to something like whole brain EEG or MEG or fMRI. Now what you’ve got is really pooled activity, or it might even just be blood flow or an indirect measure of that blood flow, something that’s really much harder to pin down to an individual cell. But now you’ve got the entire system and you can make that same Let’s say you have the same type of task, you get the same type of behavior, something navigation-wise. Now all of a sudden you’ve got this broad network of regions that are now responsible for that behavior. And it’s well, how on earth could both of these things be true at the same time? How could it be that one individual neuron is important but now a whole network and system of neurons is important for this same behavior? And I think this enduring mystery is staring us in the face as neuroscientists. We have to have explanations that accommodate both of those perspectives and understand how can we even speak about them In the right way. And to me, what systems neuroscience as a whole, as a field is doing is really like taking that challenge on, right? It’s trying to scale up the neural recording so it’s not just one neuron or 10 neurons but now thousands, hundreds of thousands of neurons with clarity, or taking the people that are working on the human imaging side of things and forcing biological constraints into their explanations. So it’s not just the default mode network did this, it’s the default mode network, which is comprised of these types of pyramidal neurons and those pyramidal neurons project down to the brainstem in this area and control the activity of this set of neurons that then change this set, right? It come it’s coming with these constraints. Pulling I like to think of these as pulling from both ends and forcing the answers onto the same playing field. And to me, that’s what I really admire about modern systems neuroscience, is it’s taking that on. So to me it’s, as you can see, it’s a broad church, right? It’s, it’s people from all different backgrounds trying to find that common language to describe how it is that a brain could work, taking all these recordings and all of the benefits that come from each of them but then trying to find a way to bring them together
Andrea Hiott: Wonderful. That’s exactly what I was hoping you’d say because a lot of this show is about those tensions, trying to hold what you just said, that it’s, it’s, it can be some very specific part of the thalamus or even some very specific set of neurons that matters, completely for wakefulness or whatever. But at the same time, it’s a whole system at you can’t just look there too and it’s very hard, I think, for us to hold that. And isn’t it also hard in academia to figure out how to do that? I mean, that you’re trying to do that and that there’s a lot of people trying to do that, feels really important to me because it’s more, efficient to focus on one thing, maybe the hippocampus or maybe the thalamus or and just know everything about that, which you still can’t, you can’t read every paper. But, trying to hold all this at once is not an easy thing to do. So I’m, I’m, I’m wondering how you came into wanting to do that, first of all and maybe we can get into how you’re doing it. But first of all, just how did you start want why was that the thing that you knew you wanted to do,
Mac Shine: Yeah. I think it comes with a large dose of foolishness a complete inability to see far in front of my face. Yeah, look, I think the seeds were planted for this quite early in my life. My father’s an evolutionary biologist but he works on what he would call evolutionary ecology. So taking principles of natural selection and using the constraints of natural selection as an explanatory principle to explain why you see the kinds of variation that you do around you. Why is it that this particular lizard has a very wide shaped head versus this other group of lizards that has a very narrow shaped head? What were the benefits of that in terms of its adaptation? And but using the explanation of the environment around it, its predators, the kinds of prey that it was trying to eat and the kinds of problems that it was trying to solve as your explanatory set. So I’ve, I was exposed to that for my upbringing. You know, I’m the son of an academic so we got to follow Dad around the world, to amazing places, New Caledonia, to Africa, to the Northern Territory. We lived up in the Northern Territory for about six months when I was about 12 years old and went to a little school there with about 12 other children, half of whom were Aboriginal, half of whom were, the sons and daughters of the local people that worked in the little station there, in an area called Middle Point in the Northern Territory. And just me and my brother were just inundated with this. It was just, around us all of the time and you couldn’t help but be curious about what Dad and his colleagues were doing. Oddly enough, my father, studies reptiles. That’s his main passion. So back then he was really interested in water pythons, Wow different types of monitor lizards and neither my brother or I have shared this interest. Anytime there’s a snake, that my kids see when they’re out on a walk, I call my dad and get him to come take care of it. I don’t, I don’t want anything to do with it. So it wasn’t the interest in reptiles that stuck but it was this interest in this really the beauty of the natural world and the complexity of the answers that you need to understand it. So I had that baked into me from a really early age and then went through a bunch of different things in undergraduate. I did a really In high school, I loved chemistry. In undergrad, I loved biochemistry. I really loved the molecular biology and thinking about proteins and RNA and DNA and all this, beautiful world you couldn’t see. But I was hopeless in the lab. And I also really liked thinking about psychology and and neuroscience but again, didn’t really wanna work as a psychologist. So I ended up going to medical school. And in medical school, in many ways it was an incredibly good training for someone with a interest in systems thinking, because what you’re doing in medical school is you’re, you’re forced to do this challenge of learning this major amount of information really quickly. And so you’re constantly having to take complex, detailed information that has all these different features to it. It has different scales of explanation, like a physiological level as well as a systems level but it’s also got, anatomy and it’s got pathology and it’s got all the different details of all the different systems. You’re constantly having to compress them down into a real gestalt understanding. You have to really get what’s going on. You have to think of the kidney as a filter, quite simply but you’ve also got to be able to understand how the countercurrent exchange mechanism works and the loop of Henle and all these fascinating things which I’ve long forgotten the details of. But the point is that I think that study in medicine like really invigorated that systems thinking that I was exposed to as a child and let me learn that quite systematically about the body But the funny thing is, when you then get up to above the neck and into the brain, a lot of the simplicity and elegance of the explanations that we had in other systems in the brain in the body, like in the cardiovascular system, in the lungs, in the kidneys, bones, the immune system, a lot of that explanation just flitted away. And all of a sudden, when you got to the realm of psychology, you started getting into this stuff about minds and, concepts like symbols and things like that. And these things just felt so foreign to me. I To me, I should have been able to start from where I was in medicine and just like extend a step, “Okay, we’ve got the cardiovascular system. It works like this.” Then the neuroscience is just like a, something like that. And you ended up just not seeing that anywhere. And so I think when I first started, I, I essentially had this really naive, expectation that I’ve kept with me ever since I started in research, that I should be able to find similar types of simplified, compressed, yet still poignant explanations about how the system works. And the real I mean, we can you can ask other questions if you like. I feel like I’m, I’m probably at the point of rambling now. But, the thing that’s really fascinating, I think, in neuroscience is that we don’t have those. They’re missing. There’s like this empty space in the textbook. And I think what I’ve been trying to do slowly over time and now I think much more rapidly now that I’m surrounded by this incredible team of young scientists that are working with me to push this idea forward, is to start to build those building blocks. Like what part what are the parts that we need? How do they work together? What can we do to scale them up? What types of analogies are really good analogies that don’t trade off all of the things that you really had to work hard to learn by studying an anatomy textbook, or really reading the modern literature on physiology, or thinking about things across scales? And I think in the process of developing that language and developing that conversation, the best part of it is that all the questions you can ask afterwards are like thousands of times better than the ones you could have asked beforehand, that you get this real benefit of the nuanced compression for that next step. And that’s what we’re really excited about at the moment and we’re starting to try to build that program out, organically. But yeah. So I, Sorry if that’s a longwinded answer to your question but it’s, it’s been a long time growing, I think, for me.
Andrea Hiott: No, it’s wonderful. I mean, a lot of things come up but first it’s funny because, often when I would try to go to your website I would get your dad’s with, the lizard? Yeah. And it’s so awesome that the both Shine Labs. But so I was thinking a lot about-
Mac Shine: You, you say that but, Dad was at the University of Sydney until I came back from the States and I started my lab and then he immediately moved to Macquarie University across town. So he’s running away from me and my neuroscience. Oh,
Andrea Hiott: He’s trying to get away. So not two Shine Labs in one place. But it’s fun. I mean, that speaks to it though, that same tension, because it’s a really interesting tension. It’s the same one you brought up, I think, that I want to try to understand. It’s hard to talk about Mm but, something like what your dad’s doing, it’s, it’s biology, it’s physical. And then what you did in medical school, it’s physical. But then like you said, you get to the mental and you’re still dealing with the brain. It’s very physical. I mean, you know that when you do the medical science of it. But at the same time, there’s that weird gap in the way that we consider it or even the language we use in our papers but also, just everyday life. And is that what you’re talking about when you say that space, or how can I better understand from your perspective? Because as a philosopher, for me, I see that as not a real gap but more a, way that we switch, into a different mode, of even speaking about all of these things that can be very hard to somehow connect in a way that we were just talking about, where you’re actually looking at the larger picture. Maybe you can tell me a little more about what you meant by that. Like, what exactly are you clarifying in your lab in the way that you just expressed relative to that?
Mac Shine: Yeah. Great question and I wish I had a really lovely, succinct answer for you. I think One of the things that’s really stood out to my group over the last few years and we come back to this again and again and again, is that the words that we use, we’re often fumbling for them in different subfields to “Oh, I don’t quite know what I mean so until I’ve got that word, I’m gonna use thought,” Or, “Until I’ve got that word, I’m gonna use representation,” Or, “Until I’ve got that word, I’m gonna use information or something like integration,” These kinds of really loaded words. We don’t begrudge anyone for using them, right? We appreciate that it’s really difficult to put your finger quite on what you mean. And if you hold anyone accountable in any one of those different terms, they would almost certainly, unless they’re completely heuristic, be willing to back down from the strongest version of it, right? The I don’t think if you talk to people that think symbolically about intelligence, that they really believe that there’s a little icon like something on our computer desktop sitting in the brain somewhere or in this mystical mist that’s our mind. But they just mean it in a way that helps you talk about it rather than not talk at all. So I understand that a lot of these things are grist for the mill for discussion But one of the problems that comes with this is that these words get used oftentimes in completely different fields, in completely different contexts, such that you’ve now got no ability to actually communicate with one another meaningfully. And we found this out the hard way and we’re still finding it out the hard way, by populating the lab with a group of physicists that are trained in the, statistical physics, in first principles, oversimplification, take all the details and throw them out and I wanna be left with a little process, some little compressed equation that hopefully will generalize and extend to everything in the whole known universe. And then on the other side, we have people that I’ll call biologists for now but they’re all types. We have philosophers in this group. We have people that are medical scientists. But I think the biologists’ core principle at the, at, at its, at its like nugget, nuggety, center is this notion of the idiosyncratic detail mattering. Like, to my father, it matters so much that this lizard is not the same as this lizard, right? It’s I wanna characterize them as different things and they’re basically the same to any Like, you and me wouldn’t be able to tell the difference. But my father would say, “Oh, actually, the dorsal stripes on this one are there’s three for every fourth scale and we This one’s three for every third scale or something.” And we’d be “Oh, okay, fine.” But maybe that matters to them deeply because this one likes to eat, this type of bug and this one likes to eat that type of bug and it matters for some ecological question. But so the biologist’s answer is something about the idiosyncratic detail that they wanna pull apart and the physicist’s answer is, “Don’t tell me about the details. I just wanna extend. I just wanna go I wanna know F equals MA. Does that count everywhere or are there places where it breaks?” And The tension between those two fields is not even just a terminological one. It’s, it’s almost like I don’t, I don’t even think the ontology is the right word. It’s like the flavor of the ontology is different for them. The kinds of answers they like are so different that there’s this weird tension between them. Yet, I think as neuroscientists, we’re right at that coal face because what we’ve got is we’ve got this big, complicated, more complicated than we could imagine system of a bunch of different eukaryotic cells working together and being indelibly marked by our experience, like shaped over billions of years of phylogeny, played out over the course of our lifespan, of our massively complicated social world in front of us and we’re trying to understand how that thing works in a detailed way. But yet we also want principles, like how is it that brain networks interact in a particular way? How is that different than what a large language model does? ‘Cause it’s really different but how is it different? How can we put our finger on it? And so we need these really different levels of explanation that are like almost to me written in different languages, not even in terms of the words we use but in terms of the types of answers we’re looking for. And I think that this tension point doesn’t get enough airspace. It doesn’t get enough discussion as a, as a whole for young scientists, young philosophers, young people interested in this space. Because without that discussion, I think we’re just We oftentimes just like hide in the area that we like. The mathematicians hide out here and the biologists hide out here and we just do our work. But what we need is answers that communicate in between. And I, I promise you I’m answering your question but this preamble helps me. So Without that tension resolved, I don’t care what words we use, because how can I ever tell you whether or not your use of the term integration is equivalent to mine or is counter to it? Because now I can’t cash them out because I don’t even know if you care about the answer that I care about and the flavor that I care about it. And so I think there’s massive growth for us to make as just a field. What answers are we looking for, right? What do we, what do we want this to be giving us, right? Do we At the end of the day, are we gonna be happy if we can say this sensory neuron is pain? That’s not gonna be what the answer’s gonna look right? It’s, it’s The answer’s gonna be sometimes in some context that neuron can end up being involved in a big complicated cascade that ends up in the person, you, feeling something that you would call pain because it’s like these other things that you’ve had in the past. Like, these longwinded answers, they don’t really satisfy. And so I don’t really think we’ve developed that part of this enough yet to really be able to get at the kinds of beautiful questions that you can ask that you, that you wanna get towards as a philosopher. Like, why is it that I feel this way and you feel that way? Or why, is this feeling that I had now comparable to the one that I had 20 minutes ago or two years ago? Or is it the same as the one that my friend’s having right now across the world or something? These are really beautiful, deep questions but they require to me, if we wanna use neuroscience to help them, a maturation of our ability to know what we’re even dealing with or how to frame it properly
Andrea Hiott: Thank you. I agree. A lot of what I was hoping to talk about and I know, György Buzsáki and his work and, -
Andrea Hiott: This whole cognitive behaviorist problem of the way we name something like memory and then we look for it. And I think that’s similar to trying to find the cells and name them for what they do, I understand all of that and probably a lot of good has come from that. But it does feel like we’re somewhere different now, where we have to hold this tension in a different way
Andrea Hiott: Ask those questions, like you said, of what are we actually trying to look for? Because just finding, as you said, the neuron for pain or memory or whatever, doesn’t actually help us with the, what we’re actually trying to do, which might be, help those who have cognitive issues live better or, -
Mac Shine: Mm
Andrea Hiott: Help, social groups know how to do something different that doesn’t require violence. Or, I mean, you could think of a million ways that these have very real practical implications that we lose somehow. I mean, or memory loss, all of these things that a lot of people are making real progress for. But what I wanna get at a little bit with you is are we assuming in that cognitive behaviorist way what we’re looking for? Is that part of the problem here? Is it that we are almost assuming a, That we already know what we’re looking for because it’s such a personal experience for us.
Mac Shine: Hmm.
Andrea Hiott: I know that seems like a really basic wrong assumption but I feel like it’s a really hard thing to actually notice that we’re doing even in our experiments, even in the way we set up the experiments. Yeah. Does that make any sense to you, or is that just too abstract?
Mac Shine: No, absolutely. I think that, yeah, every neuroscience experiment, I think in a way, has an, indelible elements of the experiment are baked into them. So I, I’m probably a bad example here because, my group mainly does analysis on existing data. We’re like recyclers of data. We work with lots of really brilliant collaborators that are extremely talented and put in the hard work to collect these incredible data sets and we like to try to find ways to ask questions that could then be using those data sets to help us refine our understanding of how the system could function. But so I don’t really design experiments the same way that someone would say, “Oh, I wonder if, I notice that I can multitask sometimes but I can’t multitask another. How could I design an experiment to test that?” Mm-hmm. I don’t really do that type of science. But I think that when people do that, they’re often tapping into something that they’ve seen as a a real thing. They’ll say to themselves, “Oh, I notice when I’m driving home and there’s lots of distracting lights, it’s much harder for me to notice the stoplight. Maybe having lots of, distracting things will make my stop signal performance degrade. Now I can build a little task that does that or something.” And so I think a lot of this stuff comes from your natural experience in the world. And, there’s no reason to deny that as a good starting point. There is a variety of experiences and it’s, a good enough place to start and say, “I wonder if this experience is dissociable from this other one,” And now I can go and compare the brain patterns and things like that. I think it’s a perfectly reasonable way to start. And, as a son of a, evolutionary biologist, ethology’s an extremely rich and important field and one of which I think we forget about to our detriment. We should not pretend like mice are little mini humans. We should ask what do mice like to do and what are the kinds of things that they’re really good at and what are the things they’re bad at. And, same for monkeys and same for, different types of monkeys. Some, some monkeys are, carnivores and some are herbivores and they have very different types of prey predator behavior. And all these kinds of things are really rich. And so I think I deeply care about ethology and taking that very seriously. But I also really admire where Yuri’s coming from in saying that if we take this terminology that was defined back, by William James and contemporaries and try to see if we can hunt for the patterns that best map onto that, you can see that’s not really leading us to a place that I think we’ll feel like we can really build something from it. It’ll have a You can almost then imagine there’d be like a bunch of little lockers and the lockers could have a bunch of fact stuff in them. You know, the working memory locker’s got a bunch of things in it. It’s got dorsolateral prefrontal cortex and the pinging model and it’s got dopamine and it’s got all these other things. But what I’m really interested in and I think this is where the medical physiological training really helps, is that, the heart is really fascinating but it’s just a pump. And, it’s a weird complicated pump and it’s got four different chambers and the inflow and the outflow look different and the outflow’s got little muscles on it and the inflow doesn’t and there’s all these different kinds of weird things with the pressure of the lungs versus the pressure of the head and the body and the legs and things like that come along and mess with the pump. But at the end of the day, you’ve got to understand that the pump’s also made of eukaryotic cells that are specialized for that function, right? Some of them zap along, once every second or so and then other ones are really good at conducting that zap. And then other ones are just really good at clenching really, effectively and causing force. And other ones are just structural. Some of them you need just to like keep the thing held up so that it makes the shape of a pump. But all together, all of those specialist cells make a pump And being able to understand both of those perspectives at the same time, right? That you’ve got this simplified story of a pump but also the detailed information about all the different cell types and that they’re not It’s not either/or, right? It’s both at the same time. Being able to understand that type of a, that, the dichotomy I think is so deeply important. And I think if we don’t have those types of explanations, then how are we ever gonna find the right kinds of terms that are gonna come up, that are gonna naturally, define the natural kind Mm, emergent properties of our of the system of a brain or a brain embedded in a body in an environment, what have you. I just think we’re gonna really struggle to find those just by guessing. So I kinda see it from both sides, that I think we need, It’s almost like we need to I like your term before tension. It’s almost like we need there to be an ethological tension that deeply respects behavior and emergent behavior in all of its gory details and like tries to be as precise as it can about it. But also on the other side, a like system first perspective that says, what are all the different ways that this thing can express its degrees of freedom? What is it allowed to do? And then we’ve gotta find the that tension and like hold it tight and then figure out afterwards, okay, now what happens if I break this part? What happens? Or if I like turn this part off, let’s say I have some, dementia syndrome, now I can’t, remember where I am or something. Oh, it, that’s likely to be these three things and not these other 17 things. And I think until we have that tense match, we’re, we’re, we’re just gonna be, a young a young field. That’s our real next step of maturity, I think.
Andrea Hiott: Yes and it is very hard. And in a way I think you’re saying this but I was thinking we have to get much more specific in the sense of we need to take it feels impossible because we need to take all this context into consideration in every situation so for this specific lizard in this specific moment and the whole replication thing falls apart a little bit if you really start to do that because you’re gonna have a different experiment every time you do an experiment. There’s that. That’s very hard. We have to get so specific and maybe we can now with the technology. If this technology’s good for anything, maybe it can help us do that. But at the same time, I feel like we have to get more abstract in the sense that we notice that we’ve been assuming we what cognition is, that it’s something in the brain or something we’re looking for and maybe open that up a little bit too of what are we actually noticing in our experiments. And to that point and too, you brought your dad up, there’s a paper I really love that you wrote with your dad and it’s called What is it called? Delegation to Automaticity. Is that the right the driving Here it I know I have it here somewhere. I have it. Yeah. Delegation to Auto Automaticity: The Driving Force Behind Human Cognitive Evolution. Now, I love this paper. I don’t know Oh, thank you how you feel about it, ‘cause you wrote it over 10 years ago and I know, sometimes we write papers and 10 years goes by and they’re really different. But what I love about this paper is you’re talking about all these things we’ve been talking about together and you’re holding the tension. I mean, first of all, becoming bipedal is a tension, right? Mm. Mm. Mm. It’s almost impossible. And the same feeling that we’re having here. And it does make a strain on the body that’s, a constraint that seems to be creative in a similar way to what we’re talking about here. But I love this paper because it actually starts with what, why are we doing this? What does it matter? I mean, we’re trying to understand what’s different about the human in a way, right? We wanna know how can we be sitting here having a conversation now, thinking about all this stuff and why Like, my dog is here asleep, that She doesn’t care about reading books and thinking about what neurons are and So this paper is saying, okay, what’s different about humans that it is all I mean, for me, this is why we do neuroscience. We wanna understand that because it causes Mm so much joy and wonder and it’s incredible, almost hard to take but it’s also very hard and painful and we want to do it better. So you start out with this in the paper and actually, the wonderful answer, or not answer but you say, okay, well, maybe the difference is delegation.
Mac Shine: Hmm.
Andrea Hiott: Maybe that’s the thing that’s actually happening differently in a human body, in a context, in an environment, as a creature that started to walk upright. Maybe there’s something different with the delegation that’s going on. For me, that feels really exciting and grounding instead of doing something like looking for, which I do, by the way but looking for memory in the hippocampus or something like that.
Mac Shine: Hmm. Hmm.
Andrea Hiott: So I wanna talk about that paper a little bit. Maybe you wanna set it up better than I did and maybe, is it okay we talk about that paper 12 Oh, - 14 years after you wrote it?
Mac Shine: Yeah. So, first off, let me say thank you so much. It’s actually, stood up as one of my favorite papers, for a few different reasons. So, there’s a lot and there’s a lot to talk about with it so I’ll, I’ll do my best to touch on points but we can come back to the details. But so the story of that paper, was actually, a really, a really fun, experience. It’s part one of the reasons it’s so close to my heart, is that, it actually really sprung out of a question that my father actually asked me while we were walking around a park near my parents’ house, in Sydney as my son, Tyler, who’s now 14 years old but at the time he was just over a year old, was walking around in the park stumbling around. And, my father as he’s, as he’s want to do, asked a very blunt question, which was, “Mac, about a couple months ago, Tyler was stumbling around. He looked like he was a bit drunk. He was hopeless and tottering around and now all of a sudden he’s fine. He can run around the world like nothing and having a conversation with us and he’s not looking where he’s going. He’s looking off to the side and, playing a game. He’s, imagining that he’s some, imaginary character while we’re playing a game.” He’s “Whoa, what a major change in his, whole capability of, dealing with this complicated world around him.” He’s “What changed? Like, what a, how did that happen so rapidly?”, he’s “You know, what part of the brain has, has developed and matured in this, in this in this little window?” And if you go and look in the literature, for, what has changed in, as we develop what people would call habits, as we become more effortless at performing these different types of tasks, you typically find, the answer and it’s almost exclusively 99% of the literature is focused on an area of the brain called the basal ganglia. And this is a set of different nuclei. They’re deep to the cerebral cortex. They sit at this way station between the outer layer of the cerebral cortex and the thalamus and brainstem between them. And they’re, they’re notorious in medical textbooks because they’re extremely annoying to learn because the basal ganglia are a collection of nuclei that all inhibit one another in a complicated circuit. So you get these weird, this one inhibits this one but that one also inhibits this other one but then goes back and inhibits this one. You’re “.” Like, why can’t you just like sit down and just become like a little feedforward circuit?
Andrea Hiott: I thought I was gonna go crazy actually trying Yeah do those tests. Seriously. Like, that was really hard to learn. It’s
Mac Shine: Brutal, right?
Andrea Hiott: Yeah.
Mac Shine: Yeah. Look, in part I feel it’s a little bit I don’t know if you’re a parent but for any of the parents out there, they’ll know the time when your child’s annoying you so you give them a really hard task so that they’ll go and focus on it for a little while. I feel like the, medical tutors were “Go figure this circuit out and see if you can come back to me.” But one of the things that, I so I studied the basal ganglia really closely in my PhD. I worked on Parkinson’s, Parkinson’s disease. This is a disorder where you lose a lot of the neurochemicals in the brainstem, like dopamine, that project into the basal ganglia and disrupt its internal function. And so I’d learn a ton about the basal ganglia. And one of the things that really stands out when you look at it, is that the basal ganglia is this has this massive collapse in the number of neurons that are at every single station along the, along the path. So imagine I’m like in the outer surface of the brain, the cortex. I’ve got a, let’s say hundreds of thousands of neurons. They project into about 1,000 or so neurons and then in the striatum and then those inhibit a much smaller set, let’s say something like on the order of 50 or 40. And then those go down to an even smaller set in the globus pallidus internus, let’s say 10, which then go and inhibit the thalamus or the brainstem or some other structure, which then has to come back and complete a loop But to me, that collapsing is about exactly the opposite of the circuit that you’d want to build if you wanted something to run relatively autonomously, right? If you wanted to have something that you’d learnt a habit be played out by a system, the last thing you’d want to do is blend it in with all the other things and smoosh it down and smoosh it down again. It’s like taking a JPEG copy of something and then another JPEG and another JPEG. You’re removing all this information as you go. And good luck when you go back to expand it. I don’t know if anyone’s ever tried this before but if you take like a JPEG copy of a high resolution image and it’s really small, like a little thumbnail and then you try to blow it up, it’s like what even, what even was this, right? You can’t even tell it’s a Rorschach test. So to me, this There was just this fundamental, incompatibility with this story, right? I was my son Tyler can now do this thing that he couldn’t do before and I really don’t think it’s because he’s like smooshed it through this compressor three times and then spread it back out again. So I had to go looking for explanations. And if you go looking around the brain, you think, well, the cerebral cortex would be a great place to look. But the cortex is, as Valentino Braitenberg, really beautifully said in a book chapter in 1990, it would’ve been really great if the cerebral cortex was the answer for how well brains were able to process information the way that we, that we do, that represent or like cognition and understand this complex information processing. But really it’s an associative memory network, right? The It There’s a lot of Hebbian plasticity. The rule is that if you’re a local neuron, you pretty much connect to almost every other neuron around you. Now, there’s exceptions to this. Some neurons prefer to project to here rather than there but generally speaking, there’s lots and lots of connectivity. There’s not a lot of selectivity so it’s not a really good place either But if you go looking down at the base of the brain in the cerebellum, now all of a sudden you find an architecture that is almost tailor-made for this function. You’ve got, inputs that are being received by the cerebellum from every single motor output that comes from the cerebral cortex, these big layer five cells. They come into these, this pontine nuclei, this big bulge that’s expanded in humans, spread out into the granule cell population, which is about half of the neurons in your brain. Those granule cells then communicate with via parallel fibers with these things called Purkinje cells. There’s these big, beautiful cells that project down to the cerebellar nuclei that then send that message back up into the thalamus, then back up into the cortex. They can also go down into the brainstem. And to me, I didn’t know much about the cerebellum then. I knew, some surface level details about, just for example, the fact that you had half of the cells in your brain sitting in the granule cell population. And I thought, whoa, here’s the answer, right? What’s happened is that my son Tyler, as he stood up, has gone, this is really bloody hard. You know, toddle around trying to move his big gigantic head forward on his tiny little shoulders and hips. He’s tried to navigate around this world that’s not flat, right? There’s bumps and and hills and everything everywhere and that’s a difficult problem. But what he’s learned is every single time he’s done it, his cerebellum’s gone, “Oh, you tried that. It didn’t go like that. Maybe try this next time.” And then he’s tried this next time, whatever and as he’s done that enough, he’s learned that the cerebellum can now anticipate what will happen in the circuit. So now as he takes the step forward, rather than him having to think, oh crap, there’s a little hill there or a little hill, like I’ve got to like walk this way, his cerebellum just knows how to handle that And it occurred to me, it hit me like a ton of bricks, if I had that in my system and I didn’t have to worry about all those little details, that’s exactly why I could look up to my grandfather and wave to him or pretend like I was Buzz Lightyear or whatever Tyler was doing. And it’s funny, sometimes this thing these things happen when you’re, when you’re working in science and I think if you remain open to the serendipity of being naively wrong about things, right? I didn’t have any preconceived notion about what the answer was. I looked into the literature. I wasn’t satisfied with the kind of mechanistic explanation that came from the basal ganglia story. I looked around a few, dead ends, then stumbled onto the cerebellum and thought, “Oh my Lord, I was looking for something with this shape and here’s something with this shape.” Bang. And immediately you can then see the, a bunch of steps into the future. If that’s the case, what happens if we go back and look in phylogenetic history and asked, how long since we started walking on two legs do you start to see an advancement of the cerebellar, volume in the brain and then social and cognitive, expansion and they lined up really beautifully. You can start to look into the literature of what happens if people have cerebellar damage. There’s these massive, cognitive sequelae. There’s this really beautiful work by Jeremy Schmahmann. His idea is a cognitive dysmetria. So there’s this really long list of phenomena symptoms that happen with people that have cerebellar damage, like strokes or tumors. There’s also this really beautiful literature showing that, if you scan people’s brains while they perform tasks where they’re learning them to become automatic, you can actually see different parts of the cerebellum taking over, like areas like the cerebellar nuclei become much more active and require more blood flow as you become better and better at performing the task. And there’s some really awesome studies in animals like monkeys and mice that really started to drill down into this, particularly recently. So this story, started 15 years ago with my son standing up, with my father taking an interest in, my son and the neuroscience of it and us having this conversation. But what it really led to was this platform to start thinking about a different way to think about what it is about our cognitive lives that makes us unique. So yeah, it’s just been such a fun one to have, both because of the fact that it was a part of, my family life and my being a father and everything but also because it gave me this new set of lenses to put onto all these other problems that I think have led to really fascinating next questions that I think we’re only really just starting to scratch the surface of.
Andrea Hiott: Yeah, I saw in the acknowledgments that you, Tyler gets a thank you in the paper. It’s very cool. No, I love that but I love this paper so much and actually I think you were already saying a lot of what is now coming out in different studies, maybe even relative to the hippocampus or to the way that we move through the world, the way that we study that. And it’s interesting, that you, when you were telling the story, you’re talking about your son learning how to walk. So this is a very hard thing to do actually, to stand up from all fours and to walk. It’s hard for the body. It’s almost like Oh, yeah you wouldn’t think a body would try that but of course life tries everything and somehow this has been good. This has done something good. And we usually try to explain it the other way around but maybe it’s doing something, as you suggest in the paper. It’s causing some tension that then causes this expansion, collapsing, architecture to do something different,
Andrea Hiott: Which, because as you were telling the story, you’re also saying, “Oh but he’s also starting to imagine and think about things and “ So as you told the story, you were already holding what we usually think of as mental and physical as coevolving together.
Mac Shine: Yeah. Yeah.
Andrea Hiott: So actually, those, that’s not really a separate thing that was happening. Those were happening at the same time and you were noticing that. So it wasn’t like cognition was changing and movement was changing. We were noticing, assessing it as cognition and as movement. Mm. But actually it was the same ongoing change in the body. I think that’s a really hard nuance to hold. In the paper you say that. You, I mean, you correct me. I’ll be really messy and too over general but it seems there’s an idea that it’s very rare for mammals to try to walk on two legs. Mm. That this happened obviously would’ve changed the way the body is encountering the world, the way that it’s sensing and dealing with that sensation. Let’s just be really like general. And of course the brain is a huge part of this. And it’s not that the brain suddenly grew a cortex, which is what a lot of people would think. They’d think, oh, you’re a lizard and you have a lizard brain and then when you stand up you grow a cortex. That’s actually not at all, I don’t think, the way that evolutionary science tells the story but somehow we get that story. But instead what was happening was something different was happening with the function of what was already there. I think what you’re saying is, Because we were moving differently, our cognition was different. Not that they’re separate in the way that your son was literally experiencing the world differently in a way we would assess his cognition as he was moving differently. Those, we have to assess those differently because we can’t hold it all together. But the body itself was doing the same process maybe. Is that at all close to what you were saying? Or maybe you can-
Mac Shine: Yeah. No, there’s lots of threads to pull on there. I think the way that I like to think about it is that there’s, there’s some, there’s times when you think of something as a impediment but it opens you up to a new set of opportunities afterwards just because you’ve now learnt that trick. And looking back at the phylogenetic history, of all the species around us, God, there’s some weird looking things out there, right? Like, do you ever It’s amazing stop and think about how strange it is to see a bloody hippopotamus? Like, it’s, it can, that thing can motor, man. Like, it goes but it’s huge. Like, it’s, like I used to So I, I’m, I’m not a religious person. Like, as I said, I grew up as a, in an evolutionary biology household so to us, the natural world was curious enough. We didn’t need, to have other explanation. It’s like Spinoza’s
Andrea Hiott: God, right? Nature.
Mac Shine: Yes, exactly. And but, there’s a, there’s a lovely line from Kurt Vonnegut that I’ll completely butcher where he just laughs at how silly it is. Like, how funny would it be? Like, God must have a sense of humor if he created something like the hippopotamus and the platypus and all these other funny things. But so one of the funny things about, natural selection, I think we oftentimes think of natural selection as finding, this perfect answer. Like, it’s gonna find the ultimate solution. Yeah. And I think a much better way to think about it is that it finds a good enough answer. It finds a good enough solution that works to give you a little bit of a benefit over, your contemporary, right? And it’s, it loves frugality. If you ever want to look at, in, if the, look under the definition of inefficiency in the dictionary, you could put the human retina there, right? The light sensitive cells in our retinal ganglia are sitting behind something like three or four layers of other cells so all the light has to go through the lens and all this, murky humor before it hits a bunch of cells and then finally gets back to the one that, responds to the photon. So this is, deeply inefficient and if you were gonna start from scratch and build a light sensor, the last thing you’d do is put it where you put it in the, in the human eye. So when, if you read books that would put the, human eye as this pinnacle of there’s no way that biology could have created this, must be created by some allknowing God. I would say natural selection did it and man, they had a sense of humor when they made it, ‘cause, who on earth would put it, the light sensitive cells behind all the other ones?
Andrea Hiott: Yeah, the blind spot and everything. Everything’s being tried, it’s like everything is being tried. There’Yes there’s not one line, or at least how, as I see it. I don’t know.
Mac Shine: Yeah. Look, I like to think that, Evolu-, natural selection likes to try lots of things and there are certain things that if it tried with them, it would be in big trouble if it changed it. Like, you don’t wanna go ahead and change your immune system. Those genes should probably stay about the same. And you don’t wanna change your cardiovascular system. It’d be a really bad idea to maybe see what it’s like if you put your heart down in your knee. Like, that’s, let’s not mess around Yeah with that infrastructure. But there are certain things that you can get away with trying. Like, you can make people a little bit taller or a little bit shorter, or you can make people a little bit more muscly or a little bit more lean, or you can make their hips a bit wider or a little bit thinner, right? All these different axes of variation now don’t cause necessarily a short-term problem,
Andrea Hiott: That’s a really good point a different advantage. Yeah Because it’s, it’s not about trying everything, I guess I mean in the way it’s always embedded in a historical movement. So you’Absolutely. You’re trying whatever you can from where you are, which But it can get away with I mean, we wouldn’t suddenly try to go from where we are to the putting the heart in the knee. So it’s always constrained by everything that’s come before. Yes, absolutely. I’m trying to get towards the repetition in the paper, too, ‘cause I think this connects but
Mac Shine: Yes. No, absolutely. And so there’s a lovely word from this, that, was I think popularized at the Santa Fe Institute called a frozen accident and so the idea was there were lots of different ways that things could be, right? There’s a big axis of variation and then one of those little parts of that space worked really well. So let’s say it’s, something like the if we keep it on the delegation story, as you stood, as the, humanoids stood up for whatever reason, it may have been to peer over the savanna, it may have been to throw rocks at things, it could’ve been so that they could highfive one another. Who knows, right? We weren’t back there. We can’t tell. But one of the things that was beneficial for them is once they figured that part out, once they got good at it, their brain had now learnt, or the people that had a slightly bigger cortico-cerebellar system to handle more complexity in that domain, all of a sudden could do that a little bit quicker and now they didn’t have to worry about that little detail of walking. They could now free up their conscious mind to start focusing on a higher level detail, “Oh, I wonder if that broken twig over there means that there’s a buffalo on the left rather than the right,” Or, “I wonder if that smell that’s on the breeze and the breeze just changed five minutes ago means that we should actually go right rather than left,” Right? You could imagine that all those higher order variables, which are extremely important if you wanna actually eat tonight, ‘cause you can then actually hunt your prey, would’ve been completely lost to the person who’s stumbling around trying to figure out how they keep their feet up because it’s so bloody hard to take this huge head and balance it on these, on these unstable hips. So the idea was and this is where really my father’s thinking was so crucial because he is so welltrained in adaptionist thinking and, thinking about the kinds of small incremental changes that can actually make a big difference for species, was you could think about those little tiny differences in that circuit having these amazing nonlinear benefits, right? A tiny More improvement in your ability to delegate meant you can free up your conscious mind for so much more stuff and that stuff doesn’t have to be predefined, right? It could be that as I stand up and I don’t have to think about it, now if I’m hungry, I can find the food. If I wanna find a place to sleep, I can look at all the different places I could sleep. If I wanna chase down that person that called me a, a funny name, I can chase them down. I can do whatever it is I want. I’ve freed up this capacity for cognitive variance that was just wasn’t available to me before. And I think this is a really beautiful notion and idea that really helps us to understand what it is about us that makes us special. Because, there’s this really, lovely idea, which a good friend of mine, Marc Tittgemeyer, told me actually dates back to the, myth of Prometheus. I’m not that much of a Greek scholar. I didn’t know this ahead of time. But the idea is that humans really aren’t all that special at good at almost anything that is around us in the natural world, right? We can’t swim as well as sharks. We can’t fly at all. We can’t see as well as eagles. We can’t run as fast as cheetahs. We can’t jump as high as frogs. Like, we can’t, breathe through our skin. You know, we’re completely dependent on the external world around us for food and all these things that just make us completely hopeless. So you can’t go looking for this like human exceptionalism and find evidence for it almost anywhere. But one thing that we are really good at that I think we do much better than most species is we adapt to new situations. We’re able to take information in and make it useful to the system but then spot new patterns of information and then take advantage of those Whether that’s related to our brain’s anatomy and the weird quirks of having a big cerebellum and a big mega dorsal and thalamus and pulvinar and big dorsolateral prefrontal cortex, what have you, or whether or not it’s related to having really nuanced, rich social environments that were really catalysts in the early days to create proto-language that then augmented those, features of our neurobiology, or whether it’s the fact that we had all this beautiful bountiful earth to work into that other animals didn’t ‘cause we were able to walk around and move and be nomadic or something like that, right? There’s Who knows what the actual answer was as to what the critical feature was. I think what I’m more interested in is that now we have a language that we could start to talk about what makes humans different rather than better, right? It’s not that being able to delegate makes us better than a hippopotamus or a platypus. It just makes us different in an interesting way that we can start to actually put our finger on and start to ask really interesting questions about, what have I knocked out this part of the brain? Maybe if I try to understand someone with Parkinson’s and their deficits as not being able to tap into this more automatic mode of thinking and they had to become much more deliberate, you’re “Oh, wow, that actually helps to explain why certain types of tasks are really difficult for someone with Parkinson’s.” And before I thought about the brain that way, I thought about it as this deliberate system. You know, I push a button here and I move my left arm and I push a button there and I see something and I push a button there and I remember a face and I push a button there and, I, smell a smell. Now I can think about that as a layer but now this more delegated layer as what I wanna tap into to actually execute and now you’ve got these systems working together. Sometimes I’m reliant on one, sometimes on another and now we have this richer language to explain this really nuanced set of phenomena out there in the world that I think have so far resisted really good description.
Andrea Hiott: Yeah. It’s putting everything on the same track and without needing to call it-
Mac Shine: Some new thing, yeah.
Andrea Hiott: Yeah. And or even just the way we study I think I’m going a little farther than even you might go. But even when we’re studying, how we move through the world and how we think, those don’t have to be on different tracks. They could be part of I mean, you say that in the paper that, that it could I mean, you say it before. A lot of people who are now saying it said it. That, which I was surprised to read it again but you say something like you describe how, which we should get into, how when we repeat things, like learning how to walk, let’s go to Tyler. Mm. The more you repeat it, then the more when we look at this, system that you describe between, for those who don’t know, just, this one part of the brain and another part of the brain. There’s three parts really. You could maybe describe it better. But something is happening with this, we call it a loop but it’s really a lot of loops and a lot of movement in different directions. But somehow when you’re repeating it, when you’re learning how to walk Once you get it, once it freezes in the right way, you don’t need to give it so much energy, right? And something shifts in this, cycle or this loop such that it’s like delegating it, right? So then as you were saying, then you have more energy, let’s just say, or more, bodily sensual attention to give to other things. So then your son can He’s, he’s learned how to walk so he doesn’t have to think about, “Oh my gosh “ He, it’s not even that he’s thinking about it. It’s more like his body doesn’t have to when encountering the world, be so awkward. It becomes habitual is probably a word people would understand better. And then other things can start becoming habitual and you can actually see this happening in the architecture of the body and the brain. And I guess what I’m saying is That, in your paper that you wrote with your dad, it already feels like a continuity that would then also become ways we move through conversations or ways we move through, whatever we happen to be moving through. Because for the body, it’s still that same system. It’s still you’re still taking in the same regularities and you’re trying to make habits of them so that you can take in more. Does that make sense at all?
Mac Shine: No, totally. Totally. So one quick, point of, terminology we should be careful about. People oftentimes don’t like using the word habit in this context because habit often means something very specific to a rewarded behavior. So you can imagine, if you gave an animal, a particular, food pellet in this part of their cage rather than that part of their cage, you’ve If they go back to that part of their cage, it’s a habitual behavior because they expect that there’ll be a reward there. But it’s not necessarily automatic. They could do it deliberately. They could say You know, you could imagine a little mouse going, “I wonder where I should go today. Oh, I’ll go to the part that’s where I’ll get all the food.” And so the habit and the deliberate part can be connected together. It may not be, we don’t know. But they can connect together in a complicated way.
Andrea Hiott: Yeah.
Mac Shine: So I prefer the term automatic.
Andrea Hiott: It’s such a complicated word. Also in philosophy super
Mac Shine: Complicated
Andrea Hiott: I mean, I’m thinking about, like Merleau-Ponty and stuff. No, it’s But how else would you I mean, there’s something about the body Well,
Mac Shine: So I like the term automatic
Andrea Hiott: Yeah. What would you say I guess?
Mac Shine: So I I like to have, habitual and goaldirected as on one set that you could use and I like deliberate and automatic on the other. And you could imagine a deliberate habitual or a deliberate goaldirected, or you could imagine automatic habitual, automatic, goaldirected. These things can all be, cross-matched. And oftentimes they’ll relate but they can be dissociated. And this is where, I don’t wanna belabor terminology too much but it’s just important to click on that as like a-
Andrea Hiott: No, it’s really important
Mac Shine: As a point of interest. -
Andrea Hiott: Actually, I would like to try to understand it better. So if the body is just moving through the world, locomotion, right? Learning how to walk. Yeah. You would say it becomes automatic.
Mac Shine: Yeah.
Andrea Hiott: That this thing. And then-
Mac Shine: And you could do it deliberately too, right? If you were walking on a bunch of Lego pieces and you wanted to walk on the least sharp ones, you could deliberately place your feet down, right? And you’re now not just going through the motions and walking. But if See, yeah weren’t thinking about where I’m gonna place each of my feet, you’d now just, automatically walk on the sharp Lego pieces and probably Well, who knows? Maybe you’d cause yourself more pain or not, I don’t quite know which one works better. I try to avoid Lego pieces. But the habitual versus the goaldirected is something that you can really dissociate quite cleanly in the literature by lesioning different parts of the brain. There’s a beautiful literature on this where they can overfeed animals and then they don’t want food anymore. And then if you lesion a part of their basal ganglia, now all of a sudden they wanna eat more. And so you’ve removed their habitual drive to not over, to not overeat ‘cause they’re full. And so there’s all these kinds of really clever experiments you can do and I think a lot of the reason that we thought that the basal ganglia was where habits are is that it probably is a is really crucial for that whole process of, taking a action in the world and translating it into behavior that is then repeated. There’s no doubt about that whatsoever. I wanna make it clear that I’m not, trying to go against that entire literature. It’s more that I don’t think that’s the same thing as my son’s ability to walk without thinking about it so that he could then free up his, other resources in his brain to have a conversation Yeah with me and my father. Can you say why Can I say why?
Andrea Hiott: Yeah, why is it So why is it different Yeah. What’s the auto what makes that different? Just to be really clear ‘cause-,
Mac Shine: I think the way that the terminology has been used in the field has been really based on the operationalization of the terms that were in the context of the experiments that could be run in rodents. So in rodents you could, through a lot of really hard work and ingenuity and, it’s This is like decades of really clever neuroscience that’s been done here. They would essentially train animals, in a, in a cage. They would train them to they like this type of food pellet, they don’t like that one. They can get them to make that choice. They can then overfeed them. They can then go in and put, do a very targeted lesion in some part of their basal ganglia and now the animal will keep eating. They’ll overeat. In fact, they shouldn’t eat ‘cause they’re satiated. They should stop eating and now they’ve got, essentially a completely habitual response. Yeah. I know
Andrea Hiott: Those studies but I don’t Because, I mean, the rat doesn’t learn how to walk in the same way as a human in the bipedal way. So I find that a little hard to I see what you’re saying and I know it’s really important and I’m glad you brought it up. And I know we don’t need to dwell on this but I do think it’s actually very important in terms of what I was trying to say about putting it on the same track. Okay. Because the What you’re saying there seems to assume that the cognitive side is somehow not on the same track. I don’t think you’re saying that but what I mean? Mm. Because you could-,
Mac Shine: I’m not trying to say that so maybe you can clarify. Yeah.
Andrea Hiott: So if, I mean, when you damage this part of the brain, you can also have movement issues. So the same Yes in a human, the same things that would’ve been automatic when you damage that part of the brain in the human would no longer be automatic and maybe they would actually compensate or overcompensate in the same way that the rat would be eating. Yeah. So I see in the literature why those words are different but if we really were taking this very, radical thing Yeah that I was saying before seriously, you would have to look at that very differently in the human, I guess.
Mac Shine: Okay. Yeah, no, I take
Andrea Hiott: Your point. Because of the way we learn how to move through the world versus a rat.
Mac Shine: Yes. I think there’s tons of differences in the way that a rodent engages in the world versus a human so you’re right, that’s absolutely got to be there at the start of any conversation. The second thing to say is that even though there are, equivalent regions in the same parts of the basal ganglia that if they lesioned them in the area that subserves a motor domain, that it would cause an impairment in execution in terms of motor execution. It’s not exactly the same regions that they, impair when they’re actually trying to alter the decision-making of the animal. So it’s, the nucleus accumbens in the decision-making task and it’ll be something more like the globus pallidus internus, or areas like that, or the parts of the stri like the putamen versus the, the, nucleus accumbens and the and the ventral stratum in, in the decision-making task. So there are some similarities and differences there. And I think, the, topological argument that I made before about the compression of information as opposed to the expansion, I think is a really important feature of the kinds of, explanatory factors that we’re looking for when we want to understand how is it that particular, movement could have been executed without there being, a supervisory role played by some cortico-, thalamic loop on top. So I think there’s a number of arguments there that dissociate those but I certainly don’t want to making a claim that there are divisions in this space. To me, cognition is really best thought of as an embodied, or a factor that, results through the embodied interaction, of the agent in the world as we learn and navigate with all this complex opportunities around us for action. We take for granted the complexity of the natural world around us with all of the different weird things that are there, gravity, cause and effect, all these other kinds of strange phenomena around us. We learn to anticipate though the, sequelae that show up again and again and again. And my argument and the argument that we make in that 2014 paper, is that if you have a system that is capable of taking those situations and anticipating what the likely sequelae is so that you don’t have to deliberately think about them and calculate them but you can just take them for granted, that frees up, resources to do other kinds of things that wouldn’t have been available to you. So maybe give a concrete example, because I think walking is a tricky one, right? Unless, some of the listeners here have had a stroke recently or a loved one that’s had some form of neurological impairment, they probably haven’t had to learn to walk at a time when they’re actually selfreflective and thinking about it. But one thing that we all probably can remember is learning how to drive. And if you remember learning how to drive a car, it’s an extremely overwhelming experience. You have sat in many cars before that point. You’ve probably not paid very much attention other than the fact to know that sometimes your parent or the person driving honks the horn and says some rude word to someone who drove in a weird way around them, or screeches to a halt or sees a light or something. But you’ve not thought about it. And now all of a sudden, you’re thrust into this space where you have to hold the steering wheel, push different things on different sides, look at a bunch of mirrors that you didn’t even notice before, push this pedal but not that pedal at exactly this right time. If you are unlucky like me to have to learn on a manual car, you also had to worry about gears and gear sticks and clutches and all the timing of that. And
Andrea Hiott: When you first Is that normal in Australia? The gear Sorry? Is that a norm I’m, I’m just Because in America, we never have to learn on the manual, in Europe It
Mac Shine: Used
Andrea Hiott: To be normal everyone has to. So it’s a, it’s a real And how is Yes is it there, this-
Mac Shine: It used to be normal. It’s now no longer normal. And-
Mac Shine: The newest car that we just got, you push a button to start Oh, okay. Well you flick a little thing to go into different gears and I find it I still find it so odd that I, flick a button to change gears. It’s so weird. I used to I used to really like changing gears.
Andrea Hiott: Yeah but this is a great example though. Yeah. So it’s very hard to learn how to drive no matter what but especially when it’s normal
Mac Shine: No matter what, right? It’s a very difficult information processing challenge. But if you then forward to the last time that you drove a car, I can almost guarantee that you spent almost no time thinking about what you were doing. You didn’t think about where your feet were going. You didn’t think about where your hands were going. You just flicked the different things at the right time. And this is a, is a reality when you switch cars and all of a sudden you drive a European car where the indicator’s on the left and the windscreen wipers are on the, windscreen wipers are on the right. Because what will happen is as you go to turn you’ll hit the windscreen wipers and go, “, shit.” And I and you mess it up all the time because you’ve got all of these inbuilt automatic responses that you don’t even think about that’ll let you do other things. And the point that I wanna make here is that when you get in a car to drive, you don’t think push pedal down, release gas, brake, whatever. You think, “I wanna drive to the store right now and buy the milk that I need for my son’s cereal in the morning,” Or, “I can’t wait to go to work and have that meeting ‘cause we’re doing this really cool project,” Or some other thing. Or, “I don’t wanna be late for the train.” So you’ve got all these now higher order cognitive tasks that you’re performing that I would argue if you didn’t have the capability to actually drive your car automatically, you’d have no hope in doing. In fact, if you asked someone asked you when you were learning how to drive to actually do some form of extra task on top, you probably would’ve smacked them and said, “No, I’m trying to learn how to drive.” This is an auto-, already an overwhelming, complicated thing to do. So to me, this is a much more prescient example because it’s something that we all know and we all experience. But I don’t think that One thing I don’t like about this example is that driving feels very sensory motor-based, right? It’s like an input comes in and then I push the things in the right way and then the car drives. But I would actually like to go much further than that and say that, very cognitive tasks, things that we would think of as purely mindbased, something like the game of chess, has exactly this same type of structure to it. If you ever Have you ever played chess? And the first time you played chess, you look at these pieces and you’re “What are these things?” There’s these weird lumpy objects and there’s two horses and two things that look like castles and weird pointy ones and one’s got a little, cross on its head and these little small ones. But you so you first learn the names of all the pieces and then you kinda learn what the rules are of how they move and there’s still, you sorta think about, you’re “Wait a second. Can the horse go up two and across one but it can also go across two and up one but it can’t go up three? That’s not allowed.” So it’s gotta have these very particular rules, right? But then as soon as you start to play a few games of chess and you start to get that under your belt and you start to learn those rules, now all of a sudden you can start to think, “Oh, my opponent’s trying to take my pieces. I really would like to take theirs instead. I wonder if I could line it up so that I can get my bishop to roll down and get that piece that Get that horse of theirs. That’d be great.” But then as soon as you start to get that down, now you start to realize, “But wait, my opponent can see me doing that. They can see me setting up the board to take their piece. Now I’ve got to be subtle about it and now I’ve got to send this horse out to look like it’s attacking before I send out the bishop and now I can take their piece because they were so focused on my, on my knight.” And so you can see now you’re going up now in layers of abstraction, layers of complexity in this very cognitive game. And I’m willing to bet that the experts that play at the highest levels of chess have got so much automatized knowledge of the different ways that games can play, that they’re not sitting there thinking, “Oh, gee, I wonder if a pawn could move one or two rows.” No, of course not. They’re thinking, “Oh, I’ve seen thousands of games like this and when they start out like that, they’re almost always trying to do this thing. So I now I’ll get around that and I’ll do this other thing but I’ll do it in a way they can’t see me do.” And so you’ve got this like layers upon layers of complexity to the task that I would argue cannot be achieved if you don’t already have that automatic knowledge built up to actually get you to that higher order. So I think this is just a different way to start thinking about framing the cognitive capacities that we take for granted. But I think if you put, any cognitive capacity through this lens, it could be playing chess, it could be mind wandering, it could be composing a symphony, it could be trying to struggle for that right word in the next sentence. All of those, I think fall prey to the exact same type of an architecture where you can rely sometimes on automatic fill in the gap type responses and sometimes on more deliberate, effortful, detailed extraction of information. And it’s the interface between those that really defines our cognitive capacities and what makes us really interesting as a species.
Andrea Hiott: Great explanation. And you said you don’t like the driving you like the driving example but you don’t like it because it’s sensory motor. So I think-
Mac Shine: I just Yeah, it’s just not, it’s not like purely cognitive in the way that I think we’d often think of chess as like a really cognitive game. Yeah.
Andrea Hiott: Yeah but I think that’s what I’m trying to explode a bit here and what I’m I’m being annoying a little bit but I actually already mentioned No, please do in your paper because I don’t I mean, that’s what I was trying to get at first, even bringing up the cognitive behaviorist thing and the way Mm take things for granted. Like, why should we imagine that the sensory motor action is somehow different than the chess action? Because even when we’re playing chess, we’re literally moving and so our bodies are also habituating to regularities. When I move it this way, or even when I hit the button, ‘cause a lot of people play online now but then actually, it’s, it doesn’t feel different to me. It feels, that we are what we’re basically If we, if I make it really general, is the body is encountering certain regularities and the more often it encounters them, the less it has to focus on them because it just There’s a habit I mean, sorry, an, there’s an automaticity that just develops in the interaction between the body and, whatever it’s encountering. And that doesn’t matter if it’s a car or chess or learning what, how the basal ganglia works? Yeah. I mean, for me, that was about going over and over the material of how, of the anatomy of the brain in a similar way. But do you see a break somewhere where you’re, you’re learning how to drive a car and then you’re learning, how to play chess?
Mac Shine: Not necessarily. I think one of the reasons that I want to make sure that we can put examples into this space that don’t rely on sensory motor interaction is that I think that there are skeptics out there that may be skeptics that really fell in love with the idea of a symbolic language of cognition, this idea of, the human, the pinnacle of evolution, the human mind somehow finding a way to run computer programs, in our wetware, is that I think you can come up with quite trivial examples where you don’t need to have any form of interaction with the world and you can still do cognitive work. So, Germund Hesslow has a really beautiful paper, in Trends, from back in the day that I can dig up if you’re interested, where he argues that, one of the really, prescient, examples in cognition is learning how to inhibit otherwise prepotent motor responses. And so he gives the example of a child learning how to count and when you learn how to count, you see five things in front of you and you go one, two, three, four, five and then maybe you put a different finger on each of them. And then you go one, two, three, four, five. And then I, let’s say I remove the sounds And now I can just instead of flicking my hands, I can just nod at them And then I can do all of that without having to do any of those movements, right? And so what I’ve done is I’ve systematically now pulled away each of the different motor responses. First, overt in the world. Secondly, my fingers. Thirdly, naming them, removing that and then actually moving the fingers at all. And yet I’m still doing something, right? I’m still counting. And we’re all capable of doing this with this like inhibited potent motor response in the world. Now, whether or not physiologically in our brains there’s still something that is doing exactly all the same parts Imagine we can, like Let’s say we could, lay out in a row all the different elements that were important of that sequence of five all the way from over here to being out in the world and all the way in here to being that part in the brain that’s stuck around at the end. I’m slowly paring these away but these parts in my brain are still probably doing something pulsatile and going, one, two, three, four, five, probably. We don’t know how to test this exactly yet with any precision. But we can do it without there being an overt movement in the world. And I think a lot of times when you talk to people that are interested in understanding cognitive phenomena, they’re not so interested in whether or not we can pick up a bottle and move it around. They’re interested in whether or not without even moving it I could imagine what it would be like without even doing that. They want an answer that has no motor response at all, no engagement in the world at all. And so I think as someone that’s really interested in those deeper answers and how we come up with a constructive, process model that fits all these things, I think we need to be able to make contact with them. But I don’t think for a second that, you can get to those same levels of cognitive complexity without that natural interaction in the world. At least, that’s my best guess as to how our intelligence is developed, is through that interaction, failure, development over time. But I’m not a developmental, neuroscientist so I don’t want to speak too closely to a lot of that nuance.
Andrea Hiott: No but it I think it’s important in so many ways what you just said. I mean, for a philosopher, it reminds me of, the representation hungry arguments, where, okay, you can explain everything through embodied movement in the world but then when you get to something like where you’re just sitting in your room imagining and planning, how do you explain that there’s no representations in the brain or something? Mm-hmm. I think these are really important questions but for me, there’s a a thing that gets, ignored there, even when you’re accounting and so on and it has to do with delegation, I think, in the same way that it is in your paper. Which is that we as humans, one way we’ve learned how to do more of exactly what you explain in the paper is that we create other, external, scaffoldings or, we can create language, for example, or images-
Mac Shine: Mm-hmm
Andrea Hiott: Which does some of the delegation in a different way. So we have these external representations. We have this situated, all these words that get used but basically what we’ve done is create some scaffold that’s external that can do a lot of the same delegation work that maybe is happening in the body, or at least can cue it or something, in a way that’s less, takes less of our attention. So when we when we speak, we’ve done that And then it would be, okay, then I don’t need to move my fingers maybe, because for the body itself, I’ve, I’ve tagged it to a external or a sensory representation. The weird thing is, how can I turn my own attention on my own, patterns? Yeah. And that’s where it all gets very crazy. But maybe not, right? Like, we can turn our attention to walking. Yep. We can turn our attention to driving and that doesn’t feel weird. So why is it I guess what I’m really saying is if you really took it literally, it actually wouldn’t be that weird. It would just be that humans, that the same thing is happening, it’s just that Mm-hmm from a our own position trying to understand it, we’re figuring out how to do that but-
Andrea Hiott: Does that make any sense to you at all? Do you see that No,
Mac Shine: No. No, absolutely
Andrea Hiott: Like a shift in that -
Mac Shine: No, absolutely. I was I was just thinking that, a lot of the, a lot of my intuition comes from sports and playing sports for a lot of my life. And, one of the things that we like to talk about a lot in the lab is that sometimes that, reflective attention as a sports person can be really helpful. Like, if you’re in flow and you’re playing in a game and you have set a defender up three or four times with one thing, you could think, “Now’s the time to get them with that other thing,” And you can deploy it. But other times it can be quite counterproductive and if you’re not in the right brain state, if you think, “Oh crap, when I shoot a free throw, do I do I put my elbow out or in? Like, where, what’s the exact angle that I put before I shoot?” All of a sudden, now all that automatic architecture just disrupt It’s like the house of cards crumbles and you’re “What?” And you just miss the shot. And it’s, it’s really, I think, a an art form to knowing when to deploy deliberate versus trust the automatic. And I think, just watching, elite athletes and I think musicians are very similar and I’m sure it’s the same in a lot of vocations, being able to trust that the system will do what you want it to do when you want it to do it exactly the way you want it done is such a, it’s such a beautiful thing to think about, that, experts have like found that stability, that calmness, they can kinda keep themselves there but then they know how to like tweak it a little bit and, they can put the little tremulo on the guitar solo or they can, tweak the ball just so it avoids the Like, I just love watching these things of like in slow motion the defender’s hand comes and the ball goes like that and you’re “Oh, how did they know how to do this?” And there’s just this elegance to learning how to deploy a nervous system at its highest level. And, I’m really, excited by the ability to the fact that we can like move towards better understandings of how that works. And I’m with you. I don’t think there’s a strong fine line between things that are cognitive or abstract and things that are concrete. I think it’s a, it’s a a line that can stretch in many different dimensions and it really comes down to the kinds of questions you wanna ask and what you’re satisfied by in terms of an answer.
Andrea Hiott: And also to what you were saying at the beginning about being able to really think, More specifically about, holding the tension, I guess, is what I was saying too and not Some of the, what you just described reminds me of that. I mean, even, I love sports too and there’s a I mean, it can be music, it can be writing books, whatever, where you do get into a flow state, right? And you don’t want to get out of that. You don’t wanna, be disturbed. But at the same time you can also watch yourself in that state and there’s a different steering that happens. Is maybe that what you’re talking about too? Yeah.
Mac Shine: Yeah.
Andrea Hiott: Which feels to me like you’re able to hold the tension and you’re not trying to resolve something. Does that?
Mac Shine: Yeah, no, absolutely. And I think the real virtuosos out there, have figured out not just how to automatize all of the details and also, if you’re a sports person, you have to be, the right height and have the right body mass and the right kinds of, kinds of, physiology to even be able to compete and I’m sure the same for a lot of different, areas. Like, great musicians have to have the right ear and the right kinds of, dexter dexterous control and things. So there’s, a bunch of other factors that come into it. But there’s something that I think that we, if we think of, our favorite, expe experts in all the different domains that we were interested in thinking about, there’s something else, I think, that they are learning how to wield, which is exactly this capacity to know when to deploy which resource. The, there’s like a taste to it? Like, when will I go all out versus when do I hold back? And, when do I get aggressive versus when do I go passive or. And there’s there’s real, there’s like an art to this that I think is really quite beautiful. And probably what draws us into it as, spectators, is this yeah, it’s hard to put your finger on it but when you see it, you can see it across all the different domains, this ability to deploy the ballots in a way that, yeah, is beautiful.
Andrea Hiott: It reminds me, actually and not to try to flatter you but when you were talking about, I’ve heard you say it a few times, where you have these moments of things click.
Mac Shine: Mm.
Andrea Hiott: To me, like with Tyler, I mean and your dad saying that there’s a state you’re in, right? Where you’re-
Andrea Hiott: Open and you’re, attentive and you’re actually I mean, a lot of people might be stressed and not listening to their father, or maybe they would feel, as we do, like we need to prove to our father our knowledge. So you basically just start giving him a You know, the knowledge and here’s the answer. It’s the cerebral cortical loop and the basal ganglia is doing this. And even, -
Mac Shine: His eyes glaze over,
Andrea Hiott: Yeah. Yeah. But like those could have been the responses. So I just there seems to be something with your stance in the world or your way of being that you’re able to take in a lot. Does that make sense? I mean, it’s even what you’re doing with the systems neuroscience, or what we started with in a way but do you see what I mean? Like, there’s a way of being where you’re, you’re ready To know that you need to flip the ball in this way. I don’t know if that makes sense.
Mac Shine: Well, I mean, that’s very lovely of you to say. I think one of the things that I’ve really loved about this recent phase of my scientific career and there’s many different phases of careers as you go, is that when I first started out, I was so young, right? Learning and didn’t know how to write, didn’t know what I was doing, didn’t know statistics, didn’t know, linear algebra, didn’t know fMRI, SPM, all these other You know, I saw you interviewed Karl Friston, right? I’d read so many Karl Friston papers, technical, philosophical, right? And just, how on earth can we fit all this in together? And you just feel like this, little pebble in this, canyon. And as you go along, you don’t I don’t think you grow much bigger in terms of being a pebble. What you end up doing is finding other pebbles that are a bit like you and you coalesce together. And I think one of the things that I’ve really loved about the last, few years is that by being By giving into that, right? And by trusting that those kinds of things you’re talking about, that susceptibility, that humility, that curiosity and really trying to essentially just check your ego and just say, “You know what? I’m gonna be wrong so much more than my students. Like, I want to be the wrongest. I want to demonstrate to them that I’m the most ignorant. I want to ask the dumbest questions, because my thinking is if I ask the dumb question now, the next question I ask is gonna be better. That’s my, that’s my hedge.”, we end up naturally bringing in more and more people that share that same type of a, ethos and it’s not really explicit. We don’t put in our, in our papers, “Please reach out if you’re like this.” But what we’ve found is that more and more students that have that same passion, that same curiosity and are just not They just don’t care whether or not you say that they’re the one that came up with the answer. What they want is to be part of a group that is trying to figure it out. Man, it’s so It’s amazing. Like, I mean, the people that I get to work with now are just brilliant and hardworking and enthusiastic and just they challenge me all the time. Like, I’m their student, right? They’ll They know about the thing that they’re studying so much more than I do and I get to have a new PhD where I get to learn about nonlinear dynamics and, mean field modeling and I get to learn about the arousal system and the cortico-thalamic system and the cerebellum and all these other details that, I wasn’t trained as an anatomist and I could never do that in a million years. I couldn’t do physiology. These are all things that are, way beyond my ken. But by being open to that And be susceptible to it. Now you get to work with people that have all those levels of expertise so that now working together you get to compress that information down into a ball and carry it forward in the hope that it helps you ask a better question next. And so I feel so fortunate. I honestly pinch myself that I get to be in the situation that I am working with these beautiful people and just learning from them every day. And I think it’s a lesson for students out there that if you have something like that in you and you’re curious about it, just lean into it and try to find those elements that you find selfrewarding and just go with it. Trust your, trust your gut. And you’re gonna find way more fun looking for ways that you’re wrong than you are by trying to reinforce that you’re right, ‘cause odds are you’re wrong. It’s a pretty big complicated world out there and, there’s, biology and nature has got, lots and lots of, answers in store for you that you never would’ve guessed ahead of time
Andrea Hiott: It’s almost becoming exciting to be wrong or, I think I really appreciate what you just said and I actually think it’s more the future than we realize because, I mean, everyone’s very worried about AI and all of this. But actually-
Andrea Hiott: I think it’s actually showing us that’s always been what’s important. And just to this narrowing and delegation, thinking of the, what you were saying about even the basal ganglia or this condensing thing, that’s an important thing to do. But we can now have other ways of, technology and stuff can help us do that in a way but nothing can really help us open, to what we don’t already notice? Yeah. You, the, anytime you’re using technology, it’s narrowing you based on everything that’s come before. But only when you’re in situations like you described where you’re literally with people that are challenging you and you’re learning how to sit with that tension and be challenged, does this magical stuff happen. Like, we don’t even have a clue what’s possible, actually, with the body and the brain. As you said, it’s so young and it’s so exciting to think about that. Mm-hmm. Not to try to find the right answers through old frameworks, which is pretty much what we would be doing if we just think AI’s gonna answer everything. It can help, of course. Yeah. I know it can be creative. It’s not I don’t wanna get into a fight about that with anyone, listeners. But, the thing I’m trying to say here is, just to I don’t know if we can relate it to your paper but I think it is related in a way to the delegation. And, what’s the point of that? The point seems to be to have ways to expand what you can be attentive to. -
Mac Shine: Yeah. No, I love that. That’s a really lovely way to put it. We’ve, we talk a lot about, what people call AI in, in the lab, in our discussions. And I sometimes, I, we think about this practically as well with we’ve got, two teenage, well, a teenage boy and a another son who’s about to become a teenager. Mm-hmm. And, language models and the like are just ubiquitous challenges for them to grapple with. I almost, I almost begrudge the world a little bit for laying that at their feet, ‘cause, it’s, it’s hard being wrong. It’s hard to gain knowledge. It’s hard to read difficult texts. It was hard to read Lord of the Rings when I first read it. It was long and boring in lots of parts and, I couldn’t quite get through it.
Andrea Hiott: Oh.
Mac Shine: But when you I
Andrea Hiott: Started with The Hobbit and it was so great
Mac Shine: No, The Hobbit was great but The Lord of the Rings, right? Lord of the Rings is hard getting to Tom Bombadil was like a slog, right? But when you get through the end, you’ve, you’ve you’ve gone through this process in the, in the literature and nowadays I feel like the equivalent of that with the way that we’ve been processing and treating information is to go, what was the point of Lord of the Rings? And you could just like get the summary and then you could be, you could be disparaging of it. You’d be “, Frodo was grumpy the whole time,” Or something. You’re “That was the point. You know, you missed the point. You didn’t have to be a hero to be a hero,” Right? And so there’s all these things that come from the process of like grappling with these things over time. I think we’ve gotten to this position, in large parts ‘cause it probably makes a bit of money for whoever was trying to build the thing, where we just want the answer all the time. Answer, answer. And I don’t think I’m alone in saying this. I think there’s a lot to be gained by the process, by not necessarily getting the answer but having to go through the process of trying to discover it I, I tell this to my students all the time, people don’t wanna hire you when you get out of this PhD program because everything there is to know about, automaticity in functional MRI cognitive tasks. They’re, they’re looking at you as someone who’s dove into a really difficult problem, read a ton of stuff, made a bunch of mistakes, learned a bunch of things and then come out the other side with some form of, understanding of what they could do with knowledge. Like, you’ve learnt how to grapple with this complexity of the world and I think that’s really important and I think you can’t fake that, right? Like, you can’t, you can’t just, like If I just woke up tomorrow and downloaded phd.pdf or something, it’s not like I’ve, got all the things that I would get if I do a PhD. There’s something about walking the walk and going through the steps that matters. And I think at the moment we’re really obsessed with, outcome, outcome, outcome. And I just think that’s selflimiting and I love lots of parts of what we call AI. It’s really fun to be able to ask questions of things and get things done really quickly. Claude Code is incredible to use and, saves me a lot of time. But it also feels like it’s making my knowledge really flimsy. It’s like pointing every putting everything I know now on, little sticks. Whereas if I went through the process of learning, I could build these kinda, mud brick houses that are just, stuck on hills of stuff and if you try to knock it over, it would just stay stuck on the hill. Whereas if you take my knowledge that I learnt through a large language model and smacked it would just fall over. Like, I don’t know how to defend anything around it. And I think that’s a really big problem for us to figure out how to solve and I don’t think it’s easy ‘cause it’s not gonna be economical to do it. It’s not gonna be a It’s not gonna be, a satisfying, quick answer for people. You can’t just tap in some tokens and then have the answer come back. It’s more like if you really wanna be, the best violin player in the world, you can’t download the answer on YouTube. You’ve gotta go and do the practice and I think we haven’t, yet grappled with how to deal with that. Yeah, it’s hard though.
Andrea Hiott: Yeah, I think, I really feel all you’re saying. I actually had thought about I wonder if your kids are going through that, ‘cause they I imagine they are right at that age where it’s just, insane. I mean, nobody knows what’s going on. It’s ubiquitous, yeah. There’s something that strikes me too when you’re saying this, is that I think for a long time we all were being told that if you had the answers, you would get hired. And it was all about having the answers. Yeah. And I think there’s a hangover right now where we had I think, I mean, I even remember just, 10 years ago or whatever, you really think I mean, I remember being in school and you really think, especially as a philosophy, you just need to know everything. You need to have at least You need to have read everything and the only way you’re taken seriously is if someone brings up a term and you already know it? Yeah. And so now I feel like there’s a hangover where that’s getting satiated, where everyone can just be relieved because all of that stuff is just right there. But after this hangover of that, maybe we realize what actually mattered was something else. -
Andrea Hiott: Maybe the friction and the tension and-
Mac Shine: Yep
Andrea Hiott: All that stuff we started with, which is really feels like what makes us human and what is life. And, maybe we can actually notice that differently. I, I wonder how we do that but I guess before we go, I just wanted to bring up, the bodily aspect, because-
Andrea Hiott: You’re a neuroscientist and we didn’t talk about it that much but you really know a lot about the parts of the brain. But also this paper I can
Mac Shine: List them for you if you’d like. Yeah.
Andrea Hiott: Yeah. I know you can. It is amazing. But the, this paper’s about the body in a way. It’s about a changes in the body that then, And I just, I know it can be annoying a little bit, to think, oh, it’s embodied cognition and all of this but And you said before you think it’s relative to the body. But even right now in light of all that we’re going through-
Andrea Hiott: And in your work, as you’ve continued this, ‘cause I know you’re still, you’re probably doing much more with this, these ideas now. Do you try to consider, the body too? Or is it’s just too much for you in the lab or in your-
Mac Shine: No, look, it’s, we definitely do. And, I think bringing it back to the driving example maybe makes a interesting controversial claim but I think one that underlines exactly your point. The next time anyone is driving somewhere and they catch themselves, not thinking about what they’re doing, I would challenge you to pay attention to just how little you actually do with your body while you’re actually driving. And what you actually find when you’re driving your car is it’s actually the car that’s driving and you’re just doing little adjustments, like your hands will move a little bit like this or something, or you’re like You obviously have to flick the indicator but the indicator then has an entire machination inside that then turns a little light on and off at the right frequency that then people can see out the back behind you. You had nothing to do with that. And if you really think about it and you remove yourself from the piece, you couldn’t drive the car without the car being able to do all of that. And you actually probably also needed the road to be set up and smooth and have little lines on, done the right way and be a certain width and a certain, a rugosity and all these different kinds of things. And so your capacity to drive, you might say, “Oh, I’m good at driving a car,” But what you’re really good at is being in the right position to take advantage of the fact that the car can drive itself on the road that’s already ready for you. And I think this is just a silly example but it, I think it speaks to exactly what you’re talking about, that, we love to be able to pull out the driving, right? We love to pull out the variable cognition, like let’s say working memory or something and say, “Oh, here it is. I’ve found it. You know, here’s the brain bits that did it.” But none of that would have happened without the entire system of an agent in a world that had lived for however long they could do before they could do that task to actually perform the thing and your body is a huge part of that. And so for things that my group are really interested in About delegation and trying to understand deliberate behavior versus automatic behavior. We absolutely think about the body. Just recently in an experiment, we had access to neural recordings from, an animal from a collaborator, a mouse, a collaborator in the US, Nuo Li, who had essentially gone in and inhibited the output of the cerebellum while a mouse was performing a working memory task. And they actually show that the mouse can’t do that task particularly well. The task explicitly was they get their whiskers, tweaked and they have to wait a particular point in time and then they have to lick to the side that they got the whisker flicked on. So they have to It’s called a delayed match-tosample task. And if you inhibit the cerebellum during that delay period, it basically can’t do the task anymore, which is consistent with this idea that the cerebellum has to feed back the information that you need to actually do something protracted in time. But we also had data from a different group, Wade Regehr’s group, that had actually done a, inhibited a bunch of the cerebellum in mice but not recorded anything from their brain. They just take little videos of them while the mice try to walk along what’s called a rotarod. It’s this like small little rod that rolls out into the, into the distance and rotates around. So it’s a little bit of a tricky task for them. It’s a little bit like a jungle gym. They have to walk down the little rotarod. And what they found was that when they inhibit their cerebellum, all of a sudden, instead of the mouse, mice being able to like scramble across this rotarod, it starts to stumble along and its legs don’t move quite wrong and its tail’s wobbling all over the place and it basically stumbles along like it’s drunk. And what we were able to do was to take markerless pose estimation recordings of its body movements and we could do the same analysis that we ran on the brain to show that the brain activity was not predictive of the future state to do the same thing with the neural recordings. And you can actually find that very similar types of, explanations are possible from both. So essentially that both the body and the brain are evolving according to that same algorithm of trying to anticipate the next step and doing a poor job of it is both related to the body not doing a good job of predicting and feeding that information back to the brain, the brain processing it and feeding it forward to the body. But they’re really a coupled system that’s inextricably linked. And so I really personally think that thinking about them separately is, it’s possible and if you have a question that’s about, the meaning of justice and, what it’s like to do abstract reasoning or something, then yeah, you can probably get away without thinking about the body. But if you want a really deep, rich understanding of how it is that humans can do that or other animals can do it, you’re probably gonna need to start thinking about how the body’s involved in that process. Otherwise, you’re just gonna not have that same holistic understanding of how it actually works. You’re gonna have a like slice of the pie rather than the whole thing and for people like me, I really just like the answers that have that can touch everything and that’s just something that I really enjoy
Andrea Hiott: We didn’t, we don’t have time now. It’s already been a long time. But, I really wanted to ask you about this because I love what you do in your work when you’re talking about Actually, I think it’s probably the work you were doing with Matthew Larkum. I was at the Charité Oh, yeah by the way and I love Oh,
Mac Shine: Great
Andrea Hiott: What they’re doing. And, -
Mac Shine: Matthew’s amazing, yeah.
Andrea Hiott: Oh, my gosh. I mean, I could That’s another thing. But but the, the point is I’ll try to give, a really general example. In this way that the body is doing what we’ve been talking about, a lot of those predictive models would say, if everything is going along as normal, then you just go along as normal. But you’re saying, well, okay, if you’re, if Like, if you were gonna predict that, you’re a monster Not a monster. A lion is about to come eat you or something, then you need to act, right? You don’t just-
Andrea Hiott: Say, “Okay, I was right about my prediction.” Yes. So I guess what I’m saying is already the body and all that stuff we’re thinking of as predictive isn’t separate, because of course we move out of the way-
Mac Shine: Yeah, absolutely when
Andrea Hiott: We predict. And I think you’re showing that in a way with your work. I would almost call it, ecological or Gibsonian in a, in a, in a Not in the way that Gibson meant it but that In a more immediate sense of, what you were saying before, like the body is part of what’s becoming the delegated activity-
Andrea Hiott: Already. Like in the example with the driving even, I know that if I had to go drive a car standing up, I would feel very awkward. Or like you said when you go to another country and everything’s in a different way, you’re suddenly Oh, yeah I mean, I feel like totally weird. Like, oh. So actually even though you take it for granted, you’ve posit your body, bodily position is part of what’s-
Mac Shine: Oh, yeah
Andrea Hiott: Being regulated. I don’t, I know you’re very messy with it, I
Mac Shine: Learned how to drive in Texas when we were living there and it was so weird sitting on the other side of the car, right? Oh. Oh my God. And so you’ve got the wheel and the same indicators but now with your right hand rather than your left hand you have to move the gear stick and I That was so strange but the pedals are in the same order. It was so weird That’s so weird ‘cause I was like half of my body is doing one thing and the other half’s And I’m “What is going on?” Yeah. There And don’t
Andrea Hiott: You feel crazy then? I mean, you lose the, the things you’ve been taking for granted. I even have it happen sometimes I’m in so many different countries and like the computer keyboards or something, right? Oh, yeah. Are different in different countries and you’re used to where everything is and then suddenly everything’s in a different place and you’Typo’s
Mac Shine: Everywhere.
Andrea Hiott: And you and you can’t think properly.
Mac Shine: Yeah.
Andrea Hiott: You know?
Mac Shine: Yeah. Oh, I totally agree. Yeah. Yeah. Totally agree.
Andrea Hiott: But these are,
Mac Shine: The example I even have two different sized laptops. Yeah. Yeah.
Andrea Hiott: What, where, what do
Mac Shine: You mean? Sorry, I didn’t mean to interrupt. No. I have two different sized laptops. I have Oh, you do MacBook Air and a MacBook Yeah and they have different sized keyboards and so I actually realized I was really struggling to write on my MacBook Air and then I was “Oh, I got “ I one day I spent the, time on my MacBook Pro. I was “Oh, this feels so good,” And I was like just churning out writing. And I was “Oh, that’s what I’m used to writing on.” I’ve written on a MacBook Pro for 15 years or something. So yeah, it just takes time. I think so
Andrea Hiott: Much of this has such so much to do with even the problems that we study. I mean, this, these are trivial but if you think of it in terms of something like when you’re starting to have memory issues or just And the what’s actually happening and better than me, I mean, in the brain, right? Mm. When you’re doing that and you think of these, loops and what happens when they’re disrupted by the world changing?
Mac Shine: Mm. Mm.
Andrea Hiott: That’s a really interesting thing because especially if you start thinking about amnesia or, or when you think about memory loss or when you think about, Parkinson’s or something, we don’t really take it for gran we don’t really think about that bodily posture That part of it or poise almost. Yeah. -
Andrea Hiott: In terms of actually what the brain is used to. Do what I mean?
Mac Shine: Yeah. No, totally agree. And, -
Andrea Hiott: I don’t know how we do that but
Mac Shine: Yeah, I don’t know. I mean, we’ve got a lot of work to do, right? We’re still super young as a field. And there’s also other sides to this as well which, people say sometimes traveling can, knock you out of ruts. I think sometimes it can be these things where now all of a sudden you were doing things automatically for so long that now you’ve got to redo them, in this more deliberate way. And in doing so, there’s like this opportunity now for refinement and sculpting of this thing that was otherwise just, sitting there gathering dust in the back of your brain? This is how I drive and this is the way I do it. Maybe that was suboptimal. And, there’s a, there’s a great line in, an old paper from John Krakauer which I really love which, pushes back on automaticity as synonymous with expertise, which says, “If you automatize at a point which is not at expert performance, you are no long you’re not an expert,” Right? Like, the thing that we would say defines an expert is that they’ve not been satisfied with their first level of automaticity. They’ve sculpted and refined it. They’ve put the time in to make sure that their jump shot looks exactly the same every single time, even if they’ve been knocked over and they’re tired and they’ve got a bruise on this thigh and, there’s a person right here versus right there or something like that. It’s the people that can deploy that skill at so, at, such an automatic way and with such precision despite all of the other variables moving around them. That really defines their expertise. And I think this is part of why, traveling and trying new things can be really fun and c ‘cause it can kinda, force you to resculpt that whole weird, complicated brain environment thing that defines kinda who we are. Well,
Andrea Hiott: You have to switch your attention back to it, too, which is-
Andrea Hiott: Within itself already something. But actually to that point you were just saying, in the paper, you I got the feeling, the one with your dad that we’ve been talking about I got the feeling that, it’s not necessarily that you do the jump shot exactly the same way every time, it’s that you can almost do what you were talking about before where you’re somehow able to know how to adjust relative to what’Yeah changing. So it’s not that you’re doing exactly the same thing even when the conditions change, it’s that actually there is no expert condition? Because every new situation will be requiring a little bit different, posture or poise. And so there’s something about that actually that at least in when I was reading your paper, I felt like that’s more what you’re saying, which we don’t like to hear because we wanna say, “Okay, you do it this way and it’s the right way,” Or this-
Mac Shine: It’s the way it’s done. Yeah,
Andrea Hiott: Yeah but that’s not actually the case in any of these probably, situations. There’s,
Mac Shine: We probably don’t have, time to get into it in great detail but there’s, there’s this really incredible, sports trainer, Wolfgang Schöllhorn, who was involved in a bunch of Olympics training teams, back in the day and has all this incredible run of success training elite athletes. And he had this very different philosophy for training athletes. Rather than training them to do exactly the same thing every single time, pick up the javelin, throw it exactly the same way, what he would do is he would say, “Now I want you to take the javelin. I want you to aim Instead of aiming for the target, I want you to aim for that thing over there on the left. I want you to hit that, the side of the,” Like the little thing that they, they throw in. “And now I want you to do it off one foot. Now I want you to do it with your eyes closed. Now I want you to do it with, this really heavy weight on your shoulder,” Or something like that. And what he was hoping to do was to provide them with so much different grist for the learning mill that by the time that they’d learnt what it was that they’d learnt, they’d now learnt the robustness to all the different things that could change. And then the hope would be that on the day when they had to throw the javelin, it wouldn’t matter if there was wind. It wouldn’t matter if someone just came up and called them a dirty name. It wouldn’t matter if they had a bit of a sore hamstring. They’d already trained in so many different ways and so much variety that they could essentially execute under pressure. And I think this is a really beautiful idea. It If you wanna link this to dynamical systems thinking, it’s about shaping the basis, basin of attraction so that rather than there just being this like tiny little flimsy point, it’s now this big like river delta where no matter where you try to start, it always gets you to the right answer. And I also think it reflects on this idea that, an expert is not someone who can execute exactly the same thing every single time but someone who’s mastered the process and knows how to get from A to B, whether that’s getting the ball in the hoop, getting the sound to come out of the violin, whatever it is. But yeah, it’s a different way of thinking and, it’s, it’s fun to start thinking that way. If you start looking at the world through that lens, a lot of things look really different and different kinds of questions pop up to you afterwards.
Andrea Hiott: Yeah and just one thing I also wanted to say to your other points was, it can I think travel and all of this does open up our attention in the way that you said. But also it makes, I think it’s interesting to think about it through different diseases too, why some people need everything to always be exactly in the same place. And why those kinds of, shifts in what you have to do with your body every day can be so disturbing. There’s definitely so much more that we could explore there but I do feel like the work you’re doing is offering a different way into thinking about all of that. And so I just wanted to put it out there. But also to the point you just said about the expert, that reminds me of what I was trying to say about the, your approach, right? Not only the systems approach and dynamical systems but also the way you do your lab and the way that, you can talk about anything with anyone and you can find a path in. That feels like a similar, action that I was trying to describe before. I don’t know if it feels like that to you, if you see in your own lab that’s a, like you were saying, a, the pebbles, are attracting in that an environment. I don’t know. I just, I guess what I’m trying to get at is, since you wrote this paper with your dad and and you do have this very big scope of what that we’ve only barely touched on here Where has your attention, gone now that you’ve, delegated so much over such time? What what’s, what’s like the thing that you’re caring about right now? Is it, is it the people in the lab? Is it a certain idea? Is all of that already one thing, is what I’m trying to get at, I guess.
Mac Shine: Yeah. Great question. I don’t think there’s a unidimensional answer. A lot of it is, finding the curiosity and just trusting in it. And so a lot of that comes from people that join the lab and so they’ll have their own interests, their own fascinations, their own, confusions and really enjoying learning and learning with them. You know, try to find the thing that they find, driving them. One of my,
Andrea Hiott: You have so many cool papers, by the way. I’ll try to list them in the show notes. But like the one about temperature I wanted to talk to you about, I can’t remember which postdoc or-
Mac Shine: Oh,
Andrea Hiott: Sweet graduate.
Mac Shine: Oh,
Andrea Hiott: Yeah. Eli’s amazing And then, the recent one from last year Yeah, I was just about to mention him, yeah where there were like six different animals that you all, or s-
Mac Shine: Oh, yeah. That was
Andrea Hiott: Super fun. Oh, I mean, there are so many great papers That was Brandon. You have a lot of cool stuff. Oh, thank
Mac Shine: You.
Andrea Hiott: Yeah. Thank you. So I just, I just wanted to give your students some Or your team-
Mac Shine: Oh, yeah. So so the two that you just, mentioned so Eli and Brandon, were postdocs with me. They’re both now fellows. They’re running their own groups. Oh,
Andrea Hiott: Wow
Mac Shine: They both came from physics and they both took the, I think, quite brave choice to come with an early young PI, right? I, didn’t really have much experience when they joined me and, they were willing to come in. And, I didn’t have like a research program that I could point them toward and say, “Hey, we’re gonna do this, we’re gonna do this, we’re gonna do this.” In fact, we took the things that I’d been doing in my postdoc. I was working on functional brain imaging and networks and these kinds of things and these kinds of analyses and cognition and we pivoted more towards really baselevel questions why on earth would the thalamus be sitting there connected to the cerebral cortex? Like, what would that do to the system? And asked quite really baselevel questions. And I have to say, when If you’re lucky enough in your career to have young, intelligent, welltrained and egoless people trust you enough that they’ll meet you in the middle and ask a question with I mean, you just have to like throw your hands up and be “Let’s go. Let’s try to figure out where this leads us.” So where we’re at now is the questions that we’ve asked and the papers that we’ve been able to work on together that have been really fulfilling and interesting almost invariably to a paper and we almost define them according to this, ask more questions than they answer. And so then you get busy trying to answer those questions. And so, for the temperature one, you can ask, well, if the thalamus can do that, what about the arousal system? And if the arousal system’s different to the thalamus ‘cause it has chemicals and G protein-coupled receptors as opposed to glutamate as the main and ion channels as the main executor, what does that do? What is the difference between those? How would you control one versus the other? What does it mean to have a system that can control this one slowly and this one quickly? How do you combine them, right? You can ask lots of questions like that. For the paper you talked about with the different species, that was Brandon’s baby, taking a beautiful concept from physics called renormalization and applying it to neural data and just finding this consistent scaling pattern across all these different animals. You then ask, well, is that there in humans? Is it that there when we’re doing these different types of cognitive tasks? Is it there when you look specifically inside a task versus generally inside a task? You can ask all these different questions. Is it there when I give a person this drug and that drug? There’s all these things you can ask that are just natural questions and you don’t know where it’ll lead you. But I think we trust the process in that the kinds of questions we’re asking, feel like they’re leading towards more coherent stories. So to bring it back to the delegation, what we’re really fascinated by now is trying to put some of those ideas down into, like real concrete terms. So rather than just saying, “Look at this beautiful story of the corticocerebellar circuit,” And, it fits plausibly that it could do that, let’s go out there and try and test it. So I mentioned those experiments where you can lesion the cerebellum and see what happens to the brain, see what happens to the body. That’s one line. Another line is to try to actually like in silico build something like this, like a little toy model of it. It’s easier said than done because the cerebellum is an exquisitely complicated circuit. You can’t just like a You can’t take like a RNN style approach and just say, “Let’s have just a bunch of random nodes randomly connected to each other and let them all wire.” The cerebellum has so much broken symmetry in that this neuron projects to that one and not to any of the others and it does it in this way at this timescale. So there’s so much detail there but we’re trying to extract the main details that we think matter and we think we’ve made really good progress there. And so the hope is that if we can come up with a version of that we can actually use to play around with, now we can start putting it into contexts and testing some of these ideas and seeing if they actually play out the way we think they do. And, man, that’s so much fun and exciting. Like I really don’t know what we’re gonna find but I can’t wait to see.
Andrea Hiott: It is really exciting and I think it goes along with so many different things that a lot of people have been doing. You talk about system one and system two, for example. Yeah. And even though I, I wanna get past the dual duality of that and think in more multiplicity, I think that model and even thinking of the delegation paper that we’ve been talking about and what you just said, there’s so much richness there that I could imagine in terms of the way we look back at our data. And I don’t know if you care much about, the hippocampal literature but this, idea now that the way we move through the world and the way we think actually aren’t that separate, is there too very strongly. So it feels like we’re trying to see the puzzle and if anyone’s gonna see the puzzle from a wider perspective, it’s probably gonna be you. But the last thing, because I really need to let you go now but the last thing is just as you were describing your lab and what you’re doing, I was thinking about care because I do always try to bring up, care and love and these words that we’re not supposed to talk about really. But that I think is probably why education is still important and will become more important in a way that we can’t foresee because it has to do with that, what we were talking about with the sports and everything, that-
Mac Shine: Mm
Andrea Hiott: Really have this urge to to collaborate and communicate and you can’t really say exactly what it is that’s moving you but you feel so moved, right? And you’re part of something. And I know it’s not every day in the lab. These are hard stuff. But, I guess I wanna say thanks for, telling students and telling people that you, approach the world in this way because I think it’s important to hear that part of teaching and being a good leader is actually asking the questions and letting the questions be asked. And I, I see that will be a sign of strength in the future. But I wonder what you think about that and just in general if that word makes you uncomfortable to think about, caring and love about what you’re doing and about, what a lab is even? Why we’re doing all this.
Mac Shine: No, it doesn’t make me uncomfortable at all. I much of my intuition for how to navigate the responsibility of being a lab, leader or, someone who is a mentor in the lab, I think it’s really difficult in We have a very flat architecture in my group and try really hard to ensure that the different people get chances to play all the different roles. That’s very important to us. Well, it’s an example
Andrea Hiott: Of your work. You’re almost like a distributed cognition lab. Yes. I mean, you’re actually an example Yeah, of the stuff that you’re studying, which is super cool.
Mac Shine: Yeah. Well, yeah, art imitates life, right? But, in, in many ways that’s very deliberate. And it’s because we want to try to bring out the, unique expertise of each individual to make something that’s synergistically better than everything that we could have done on our own. That’s really important. But that intuition is really, built on, sports and my, my time in sports over the years. And I played a lot of sports over my life. I’ve almost never been the best player on a team. In fact, I tell my kids that the one team I was on when I was the best player back in high school, we were terrible and I think we only won one game or something. But on teams when I was, a good player but not one of the best, we won a bunch of different games, won a bunch of championships in different sports. And the to me, the thing that set us apart from the other teams that we beat wasn’t because we had the best, fastest, strongest players or whatever. It was a sense of community and care and, we wouldn’t use that term in sporting context but I think it is a version of love that we would essentially want to fight for one another. We want to make sure that my teammate doesn’t, suffer that indignity of loss or something like that. And I think that type of caring about your contemporaries and, banding together is something that we do really amazingly as a species but many other species do as well. And I think, not to get too, inside of the times, I think we’ve forgotten a lot of this, as a species and we’re very divisive and look for the worst in each other rather than trying to find the things that we can do to actually combine together. So the hope is that there’s gonna be ways forward through this and I share your intuition that it’s not gonna be through looking for who knows the answer but it’s gonna be about figuring out how to value the process and trying to find ways to bring in each of our special abilities and yeah, find out how to make the most of all of them together. But, easier said than done. We’re, we’re just, I’m just very I feel very fortunate to be able to do this job and, work with such an amazing group of people, collaborators, students, very, supportive wife who, lets me make lots of messes around the house with all my papers and things and and be absent-minded professor. But and very supportive, loving kids as well. But it’s, it’s a, it’s a great job if you can get it. Very lucky.
Andrea Hiott: Well, thanks a lot for what you do. There are a few people who everyone always says, “Gosh, they’re, they’re such great people and it would be so great to work with them.” And I know that’s flattering but I think it’s important for students and people to hear that matters. Because when you’re in an environment, it matters so much. Half the time you might not see it or hear it in the paper but there’s so much mental issues right Oh, yeah, for students and just someone who’s trying to create a caring or appreciating a caring atmosphere, I think matters so much. So anyway, Is there anything Thank you very much that we should talk about before we go? From you?
Mac Shine: No, that was fantastic. Thank you so much for taking the time to read it and to ask such interesting questions. And, it’s fun to have someone coming from the philosophical perspective and pushing on some of the things that were interesting. Like, that paper wasn’t really setting out to be philosophical but I think it does have a bunch of philosophical implications that are really hard to speak about. So yeah, I’m, I’m all ears. If there are specific things, I’d be, very happy to chat again and discuss with them.
Andrea Hiott: Awesome thanks for spending time with me today.
Mac Shine: No, thanks for having me, Andrea.
Andrea Hiott: All right. Bye.
