Support the show to get full episodes and join the Discord community.
Check out my free video series about what’s missing in AI and Neuroscience
Ken discusses the recent work in his lab that allows communication with subjects while they experience lucid dreams. This new paradigm opens many avenues to study the neuroscience and psychology of consciousness, sleep, dreams, memory, and learning, and to improve and optimize sleep for cognition. Ken and his team are developing a Lucid Dreaming App which is freely available via his lab. We also discuss much of his work on memory and learning in general and specifically related to sleep, like reactivating specific memories during sleep to improve learning.
- Ken’s Cognitive Neuroscience Laboratory.
- Twitter: @kap101.
- The Lucid Dreaming App.
- Related papers
Ken 00:00:03 And so here’s the interesting part that practice of bringing the material up and working with it again, I think happens sometimes when we’re not aware of happening it happening, maybe when we’re just mind wandering and maybe when we’re asleep. So as we’re asleep information that you acquired recently comes up again and doesn’t just sit there by itself, but connects to other things, Hardly anybody is studying dreams. I think in part, because the methods were so limited. If, if the only thing you can do is talk to somebody and say, so tell me about your dream. What did you dream about? And that’s the only data you get. It’s really limited and hard to make progress. And so we have people arranged so that they can move their eyes left right left, right. And they can tell us, I just realize I’m in a dream, they make a specific left right signals to tell us that. And then we know they’re in a dream.
Speaker 4 00:01:00 This is brain inspired.
Paul 00:01:13 Hello, everyone, it’s Paul. And that was the voice of Ken Powell who runs his cognitive neuroscience laboratory at Northwestern university and Ken and his team been in the popular science news. Uh, for the past couple years, based on an experimental setup, they developed to be able to communicate with people while they’re in a state of lucid dreaming. So lucid dreaming is when you’re in a dream and you’re aware of the dream and Ken’s team developed a method to increase this chances of inducing, elusive dream in people, but then to go on and be able to ask them certain questions, uh, right now like easy math problems and people can communicate using their eyes, the answers to those questions and this ability to communicate with people while they’re in their dreams has opened up the ability to ask a lot more questions, for example, conscious states and related brain activity while people are sleeping versus when they’re awake.
Paul 00:02:09 But this is really just the, the latest in a long line of research that Ken has conducted based on his interests in consciousness and memory and the functions of sleep and how all those things are related. So during the episode, we talk about his interests in different forms of memory and how he got interested, uh, in sleep research and then more recently dream research. So we cover a lot of ground here, including their lucid dreaming app that is publicly available through their website. And of course, I link to their website in the show notes at brain inspired.co/podcast/ 133 happy dreams out there. Hope you’re doing well, enjoy Ken. I know that a lot of your, uh, focus these days and for a long time has been on memory, but correct me if I’m wrong. Did you get into neuroscience because of your interest in consciousness?
Ken 00:03:00 That’s true. Certainly. Uh, I think part of the, uh, exposure I got to psychology as an undergraduate made me enthusiastic about a lot of things. And my professors at UCLA advised me that starting with neuroscience would be a better way to go rather than emphasizing the psych. If, if I’m interested in the intersection of the two, their advice was good advice to go get a neuroscience degree, and then you can, then you can, uh, move on to the more psychological aspects of, of neuroscience.
Paul 00:03:32 So you think that was the right advice?
Ken 00:03:35 I think it was. Yeah, I think it it’s a little harder to go the other way because you don’t quite have the foundation to build on in the same way. It’s possible funny people do that too, but, uh, that was the advice I got from, from folks at UCLA.
Paul 00:03:50 Well, we’re gonna end up talking, um, a little bit about some of the work you’ve done with lucid dreaming, um, which I know that you’ve garnered a lot of popular press about, but, uh, let’s go, let’s get there by way of your work in memory. So is your interest in consciousness? Is that, is that what led you into studying memory?
Ken 00:04:10 I suppose so, but I, I might put it a little differently. You know, I stumbled into, uh, a great graduate program at UC San Diego and then had some, you know, did rotations in labs had, uh, choices of what to, uh, be exposed to and then eventually pick for my dissertation topic. And so I ended up working in Larry Squire’s lab, which was a very exciting time in the eighties. There was new ideas about different sorts of memory, declarative and procedural was the name at the time. And then it, it metamorph folks into declarative, non declarative, which makes sense because the category of things that aren’t declarative memory is a diverse set of different things. And that got me very excited and connected to my interest in consciousness because this division between declared and non declarative is really cutting across something close, but not the same as conscious memory and unconscious memory.
Ken 00:05:11 In other words, the, the type of memory that amnesic patients have the most trouble with is defined, uh, in Aquire tradition as declarative memory, meaning their recall and recognition of facts episodes. So that’s a definition, it’s a behavioral definition. It’s not based on the person’s experience, but in fact, when you recall and recognize facts and episodes, you have this conscious experience of knowing that you’re remembering. And that’s something that you don’t get in all the types of non declared memory, which we’ve explored, uh, over the next many decades. And, you know, the early work with Hm, by Brenda Milner showed that well memory, isn’t all one thing, cuz Hm, is still able to learn motor skills, certain motor tasks that he could do normally. And then it gradually expanded on more and more things that amnesic seemed to have not just less trouble with, but seemed to be normal at seems that the damage in, in amnesia, the hippocampus and other structures disrupts this ability to do the recon rec of facts and episodes, but leaves other types of memory intact.
Ken 00:06:21 And so then at U C S D there were some exciting studies, uh, while it was there by Peter GRA who worked together with Larry Squire and George a Mandler and they studied amnesic patients and started laying the foundation for a other type of non declared memory called priming. Mm. So in priming, people will see a word or a face or, you know, any kind of stimulus. And then the next time they see it, they’ll be perhaps a little bit more efficient processing it. Sometimes we call that fluency and this fluency can be measured in a priming test. Like you are quicker to read the word or you read the word more efficiently and, uh, maybe don’t know that you’re doing so. And Peter graph in his study showed that the amnestic patients would be, uh, processing the words they saw previously differently. Although if you asked them to recall the words they do very poorly, if you ask them to recognize which words they saw, they do very poorly.
Ken 00:07:22 And yet they were quite normal at a priming test. The one they used first was a stem completion test. So you see a word like motel, and then later on you’re given a stem M O T and you’re asked mm-hmm, <affirmative> just tell me any word that pops some mind. Give me the first word you might say motor, you might say mother, you know, so there are different probabilities of the different words you might say, but it increases the likelihood that you’re gonna say motel, if you’ve seen it earlier, even if you’re amnesic. And even if you don’t remember having seen any words previously, but you’re still fluent at it. So that tell told us that this is another type of memory that doesn’t depend on the hippocampus, and doesn’t have this conscious experience that goes with it. It’s just something that happens as you work with the material.
Ken 00:08:12 And when I learned of that study and I got enthused because, well, this is cutting right at what consciousness gets us. In other words, the, the memory, the memory processing that you do to show priming seems not sufficient to give us the experience of, oh, I remember seeing the word motel earlier. So there’s something extra that we get. And so now we have a, a little foot in the door of saying, well, what is that extra stuff that leads to a conscious experience? And it’s, you know, it parallels perceptual, subliminal, Uhhuh experiments. So if, you know, if you, if you, if you’re shown the word motel so briefly that you don’t really know you saw it, it might still might influence you. It might have got in a little bit, but you’re not conscious of it if, if it’s not shown for long enough in the right circumstances.
Ken 00:09:01 And so that’s that also cuts it, but that’s, you know, that’s a little harder to study because it’s very, you know, cloudy kind of right at the border between seeing it and not seeing it. And this working at news really cut very cleanly that these patients could respond to words and faces and other stimuli differently. So they had stored some information, but not sufficient information to give them normal declarative memory. Right. So that was a little long-winded, but that’s the, that’s the background of how, how memory is memory research is really fundamentally right at this border between understanding what consciousness, you know, what is the, what are the requirements for a conscious experience versus not having one? And at the time studying consciousness, wasn’t really in Vogue, in fact, right. It was a little bit taboo. And yet memory researchers were right in there saying, well, this seems to be exactly what what’s going on.
Ken 00:09:58 And so other people call the distinction, explicit memory and implicit memory, Sam, these patients can’t do explicit memory. That’s kind of conscious memory. And I think we just use the term declarative memory, cuz I think it’s not the, the, the hippocampal contribution. Isn’t exactly mapping onto what you need to be conscious. It’s just the really necessary component of it that it allows for the conscious experience. So we don’t define that type of memory that way. And there’s some other border types of memory that fit on one side or the other. And aren’t really about whether you’re conscious of the material or not.
Paul 00:10:34 What’s your overall picture these days about just memory in general memories, right? The different lines of memories. It, it seems like every field of neuroscience that one goes into you start, you have, like, for instance, I had a naive idea about quote unquote consciousness, right? When I went into neuroscience and then you start digging around and it’s really not clear where lines are. And that’s kind of what you’re just talking about memory and that there are about 4,000 different types of memory. Um <laugh>. Is that, uh, frustrating to you? Or is it exciting? What’s what do you see as the current state of, of the study of memory? Do we, do we have the right ontological categories of memories now? Or is it still fuzzy?
Ken 00:11:15 I, I think it, you know, in science it’s gotta be a continued effort to work harder to see, do we have things right or not? So I think we made progress in looking back on the eighties. There were a lot of people even into the nineties that were arguing, you know, there’s really one thing called memory and we just have to understand the basic principles. And once we do that, you know, and maybe in conditioning paradigms, we’ll figure it all out. Those, those principles really apply to all more complex types of memory. You just have to understand the different complexity and that view got replaced, got replaced with this multiple memory systems view where we say, no, actually there are different brain mechanisms at work and you can disrupt one brain system and not the other. And, uh, it’s complicated because these brain systems and most of us work together and they interact and they’re not separate, but we can separate them out experimentally and understand a little bit about how they’re different types of memory.
Ken 00:12:11 And they work on different of principles. So conditioning principles, aren’t the same as skill learning. It’s just a little different than fact learning. And the priming I was talking about. So we, we kind of acknowledge that we need to work out the principles of memory separately in these different domains. And I would agree with you, maybe we have the domains, right? <laugh> maybe we don’t continue working on them. And as we get to a more and more fundamental understanding of a type of memory and the neuro circuitry involved, mm-hmm <affirmative>, then we can say, actually is this type of memory really different from the other type or, you know, is one happening in the auditory system and the other individual system, but really they’re the same sort of thing. So that’s not a fundamental distinction. Whereas another argument, you know, what’s happening in this item might be really, this is quite different in the way it works. So I think we start out with a behavioral understanding of memory, you know, what does it feel like to remember a fact or something else? And we build on that, but to have a more fundamental understanding, we need to know all the neuroscience too. And then we can ask, we can build or equipped to ask the question well, which are the fundamentally different types of memory and understanding the principles behind them. So that’s my position on that.
Paul 00:13:23 Uh, I mean, I, I was trying to make a slide, um, to kind of dissect and go through the different forms of memory. Right. And that’s just a frustrating endeavor these days because the, there are too many little bubbles on there with, with words in them. <laugh> right. So, so it’s, I don’t know, it’s kind of an everything happens that happens in science happens with this kind of explosion, right. Where we become finer and finer grained divisions between concepts.
Ken 00:13:48 Right. Well, for me, I think the fundamental distinction, you know, there may be many types of non declared memory mm-hmm <affirmative> and even types of declared to memory. But I think there’s pretty widespread agreement on that distinction that in memory research is all over the place, acknowledge that distinction, even if they give it different terms, it’s, you know, it’s, it’s cut so clearly in iMusic patients that mm-hmm <affirmative> we could bicker about how is perceptual probleming different from conceptual writing and things like that. And when do they overlap, but declarative, non declarative it’s, you know, there’s a clear border. And then there are little, little other types of memory that fit near the border, cuz maybe it’s implicit memory, but it depends on the hippocampus in certain ways. And so it, it doesn’t quite fit into the structure neatly, but I think that’s okay. Just means, well, there’s some plurality of types of memory that makes it all interesting, but there is this broad distinction that’s that we agree upon.
Paul 00:14:46 How much do you think? Um, you know, I’ve I, as I’ve learned more throughout my quote unquote career, uh, I’ve appreciated more, how much goes on under the hood without awareness, without our subjective awareness. You, you just mentioned what it feels like to have a memory, uh, how much goes on under the hood, in your current, how much learning and memory, right. Are we not aware of? And that it matter when we’re aware of it?
Ken 00:15:10 Yes. That’s, it’s a fundamental aspect of understanding the neuroscience of consciousness to say, what do we get by having these conscious experiences? And in the memory realm, we have been studying that comparing implicit memory to explicit memory, seeing how they work differently and how do they interact in certain circumstances. So they seem to be fundamentally different, uh, in terms of the neural structures required and also, you know, different in other ways. So, you know, your, your question is about is, is there something, I don’t know how to, how to put your question, you’re saying, okay, so explicit memory,
Paul 00:15:48 Well, we could talk about your targeted reactivation for example, right? So, um, the day before or someone sleeps, right? You, you teach them, let for as an example, you teach them to play guitar hero, right? Uh, and then you associate that with something else. Um, while they’re playing guitar hero and while they’re sleeping, let’s say you associate it with, uh, you know, a sound right, a car horn or something. And while they’re sleeping, you can play that car horn at a really level while they’re in deep sleep. And when they wake up, they’re better guitar hero players compared to people who didn’t get that targeted reactivation with the car horn. And of course that’s all taking place under the hood. And that’s a lot of what you’ve learned about is just how much we are continuing to consolidate what we’ve learned into our memories while we sleep right. But if I am learning a new vocabulary word and I think about it real hard and repeat it and repeat it somehow, and I’m aware of it somehow that also, uh, is consolidating and, and helping me learn. So there’s this fuzzy, well, it’s unclear to me and you, you have all the answers <laugh>, you know, how important awareness consciousness is to, um, <affirmative> uh, to making memories, to, to learning new things and making memories. I know it’s an impossible question.
Ken 00:17:01 Well, I think one of the things about your question is, is kind of jumped ahead, uh, to mix up some things. So if we, we should separate out at the time of retrieval when you’re retrieving a memory, sometimes you’re conscious of the fact that you’re retrieving a memory and other times you’re just engaging in some practice and maybe you’re better at it than you were before, but you don’t really know that that’s because of your practice. You’re just good at it. So there’s something different about the experience of demonstrating a memory when you also have that we could call it meta memory. Some no of the fact that you’re remembering at the time. And that’s what we think about as conscious memory. When I consciously remember what I did yesterday, or I consciously remember some facts that I know I can think about it in this more complex way, because I’m bringing a lot of additional knowledge to bear and not just engaging in some activity more efficiently <laugh>. So I think that’s fundamentally different at the time of retrieval, but your, your question also brought up, well, what happens before retrieval? So to actually form a memory and then maintain it for some of time and then retrieve it. A lot of that is happening under the hood, as, as you said, it’s it’s happening without your, uh, awareness of all the steps that got you to that point
Paul 00:18:20 Or consent <laugh>.
Ken 00:18:21 Yeah. So the UNC, the, the type of awareness you have at retrieval is one aspect of consciousness in, in its relat, a memory. And another aspect is all the stuff that happens in the intervening time, which we sometimes refer to as consolidation, which is essentially saying, well, memories can actually change over time. It’s not just formed at one point and static until you then retrieve it. But actually there’s a lot of change going on. And of course we know that change change happens. Some of it, we call forgetting <laugh> you, you forget details or you forget everything. That’s one of the changes that you can have, but sometimes the, uh, forgetting isn’t happening. And instead there’s something that you maintain and you’re able to remember it later. And consolidation is this process in the brain that we think is changing the way the memory stored.
Ken 00:19:11 And that’s again different for different types of memory, but for declarative memory, for facts and episodes, we think that it’s a change in which brain structures are playing critical roles. So the, the thinking is that we should talk about the cerebral cortex. It has all these different regions that have different sort of specialist and some areas get visual information about faces and can really help you discriminate different faces and others are processing color or shape or all these different aspects. But when you’re, uh, perceiving episode, you bring together all those things. There’s, it’s a very multimodal, all those different senses coming together are in different concepts. And you, you understand and experience. So you need all these cortical areas and to actually store that memory, you need to link those different parts together. So they’re in different parts of the cortex working away, and you kind of experience it all at one time, but then storing, it requires keeping all those connections and those conform in the cortex.
Ken 00:20:14 But it seems that if you don’t also have the hippocampal structures helping the cortex, then that memory isn’t standing much of a chance of staying around. So the hippocampus keeps it together. And then eventually the cortex seems to be really good at keeping it by itself, even if you’ve lost osteo, hippocampal function after some delay. So the consolidation period is the hippocampus and the cortex working together so that the cortex eventually has something that can stick with you for, you know, your whole life even. And that, that’s what we think about is the systems level consolidation of a declarative memory. And it happens without us knowing it <laugh> right. We don’t, we need to know about that. Uh, though, you, sometimes you do, like if you’re an actor, you’re learning some lines for your part and you have to study it and you have to practice it and you have to go over it and think about it and eventually you’ve got it and you can perform it.
Ken 00:21:10 And you, you kind of understand, well, I have to do the work for that memory to be there. We don’t always have to do so much work. If you, if you meet a person and you get their name, and then you wanna be able to call them by their name later, you have to have stored that effective. And how do you do that? Well, sometimes it happens and you’re lucky, but sometimes you put in the work and you think about their name and you, you think, okay, uh, let’s see now, how, how are you doing today, Paul? And you might use the name and right, <laugh> bring it up, you practice it. And that practice is part of the consolidation process. And so here’s the interesting part that practice of bring the material up and working with it again, I think happens sometimes when we’re not aware of happening it happening, maybe when we’re just mind wandering and maybe when we’re asleep.
Ken 00:21:58 So as we’re asleep information that you acquired recently comes up again and doesn’t just sit there by itself, but connects to other things and gets worked with, and that’s just like the, the practice that an actor might do when they’re learning their lines, the stuff comes up, it gets, uh, mixed about. And if it connects well with other things, it’s more likely to then have a solid foundation to stick with you. So that’s one of the implicit processing, uh, examples, uh, related to consolidation that the, the working with that information is gonna help you remember it later. And so I don’t think there are memories that just get formed and then left alone for, you know, days or months. And then you got it again. I, I don’t think that happens much. It’s just gonna fade if it’s not used, if it’s not integrated with other things, you know?
Ken 00:22:50 And so we can do that intentionally when we wanna, if you’re in a class, you wanna learn things you’re, you’re, you’re studying, you’re thinking about this stuff over and over again. Uh, and I think you’re doing that even when you’re not intending to do it as in your sleep. And that’s what the that’s exciting about this new dimension of memory research that I didn’t work on, you know, for the two decades of my work in memory research, but it was always lurking. There is pretty interesting, you know, what’s sleep about what it dreams about. You know, those were just mystery questions, but then they connected to say, oh, you know, people have theorized about that, but now we have actually good evidence that memories are reactivated during sleep. And that, that influences what you can to remember later,
Paul 00:23:32 Sleep has become a really popular, uh, topic. And, you know, as it, as it’s studied more, its importance is appreciated more and more, but of course in sleep, you have, you know, these different phases, right? And you know, like when you’re doing targeted re reactivation studies, you have to, uh, reactivate at certain phases of sleep to, to optimize the, uh, reactivated, uh, learned information. And of course, during wakefulness, we have different, uh, phases of wakefulness. I, I, I would say I don’t only are there different phases technically of wakefulness, like there are for sleep. I don’t even know.
Ken 00:24:09 Well, yeah, it’s absurd to think that wake is one thing and sleep is four different things that you have. It’s, it’s quite the reverse, you know, there, there may be four different types of broadly defined stages in sleep, and then there’s zillions of things you could be doing during wake. So I would agree on that and we’re still trying to figure out, you know, what are the critical features of sleep? So we’ve been given this idea of stages from the early sleep research many decades ago. And, and it just stuck with us that there’s, there’s, it’s hard to get rid of that Orthodox view of those of the stages, but everyone looks at sleep and they kind of know actually it’s more com you know, there’re more interesting things happening like that. And yes, there seems to be this interesting progression where you start off in what we call light sleep or non REM stage one, and you move to stage two and three.
Ken 00:25:00 And so you get what, what you might call is deeper sleep. And the development of slow waves happens then, and then you come out of that and you might have a REM period after that. And then you might repeat that cycle, you know, maybe going on another an hour, hour and a half. So those are the cycles of sleep stages that are not exactly that way, every cycle. And, you know, people go through them differently. There’s a whole, you know, plurality of waste things work, but the general principle still seems to be there that you have to go through the light sleep to get to the deep sleep. The deep sleep is very restorative. And so you don’t wanna miss out on that. And maybe there’s some penalties of missing out on the REM sleep, which you’ll get more in the morning, except if you get up too early and don’t get the REM sleep.
Ken 00:25:46 So the, these D and stages, uh, have this, you know, nice sequence and presumably some, uh, functional reasons why it’s laid out that way, something that’s adaptive about doing it in that particular way, that hasn’t quite been figured out yet in all dimensions. So what, what exactly is, uh, the best sort of sleep? So talking about sleep hygiene, uh, connects to the idea of just getting good quality sleep. So what is good quality sleep well for one you wake up and the next day you’re not sleepy <laugh> so that’s a good, you must have had good sleep. Uh, and so we could say is that seven hours is that eight hours. It varies from person to person what might do it. And of course, if you’re just getting caffeine during the day that interacts with what’s happening,
Paul 00:26:35 Uh, it has to be at 90 minute intervals, right? You have to wake up at a, at the 90 minute, you have, uh, packets of 90 minutes. Isn’t that the
Ken 00:26:42 Rule that’s, that seems to be the norm. It’s not there, there are lots of exceptions to that. Sure. And, and I, I also was getting at the idea that, um, if we wanna think of what good quality sleep is, we should really think even more thoroughly. So the, it, you know, it might be measuring how many hours of sleep you got, cuz you can do that, but really it’s well, how, how much slow wave sleep did you got? Did you get, and, and how big are your slow waves? So we can, we have those measures. And I think we need to even move beyond that because we wanna really say from my point of view, what kind of memory processing did you do overnight and was that effective for you? And maybe we should look at that to say, well, if you didn’t really have good sleep, if you were spending your night going over and over this negative thing that happened to you <laugh> and just working in, you know, in a depressing way, thinking about these depressing thoughts, that might not be good quality sleep.
Ken 00:27:35 And yet we don’t get a measure of that. If we just say, how many hours of sleep you had. So I think our more sophisticated future of sleep quality, our, our definition will get better and will understand. Well, it’s, it’s, it’s about all the things that your brain is doing during sleep that can help you. And, you know, your example of wearing a sleep mask is a nice one. And there new studies coming out saying how, yeah, getting exposed to light, really disrupts things and a sleep mask can help you with that. There, you know, we can think about the physi of that in terms of, you know, the cardiac issues that come up and as well as, as well as, you know, just whether you’re actually getting the sleep you need, because the, the light is, is sort of signaling your body that it’s time to wake up or that it’s not time to go to sleep yet.
Paul 00:28:20 And of course, throughout the lifetime, I’m sure. I mean, these are all gradients that, you know, the, the cool quality of your sleep changes and what you need and et cetera. Wh when is that future that you alluded to? When we will understand the I’m gonna use the word optimal again, the optimal sleep wake cycle, read toll story, right before sleep and do math problems, right? When cup, et cetera, when are, how close are we to that? And where do you see this field in terms of the end goal? Right.
Ken 00:28:54 Well, I think it’s an, it’s an exciting time to just be exploring those questions more than we have in the past and bringing it in. And, you know, I, I think an important, um, Princip of sleep is that there, isn’t just one reason why we sleep. And, and that’s, that’s been a mistake that I’ve seen some sleep researchers take, as they say, well, what’s the reason we sleep. And they look at different animals and they say, well, it’s really to, to keep out of dangerous situations for safety. And, you know, maybe that’s true for some animal. Wow. But you know, not of look at the biology of, you know, our, our whole body, there doesn’t need to be one reason for, for a thing. There can be many reasons like, you know, we have respiration to help us get oxygen, but look, it helps us communicate too.
Ken 00:29:42 <laugh> we have this we app, we co-op the same apparatus to, for communication. So I think sleep is beneficial for many reasons, not just one. And I’m focusing on the memory benefits of sleep, which had been controversial, cuz other people were saying, well, no, that’s not the main thing. The main thing is something else. And you, you know, it, it took a long time for sleepers to be convinced that, oh, actually memory is something that’s, uh, connected to sleep and that there’s improvements as a function of, of getting sleep. And part of that was the methodologies we had. Weren’t so good. For example, if you just had sleep deprivation of your method, you’d say, well, either people sleep or you make them stay up all night and what’s the problem of a person that’s been staying up all night, you know, they’re, they’re not paying attention.
Ken 00:30:29 Well, they’re not, you know, remembering well, all sorts of things. It’s very crude way to ask the question and even just comparing eight hours of sleep and eight hours of wake is messy. Kind of, for reason you said, it depends on what’s happening during that eight hours of week, it’s all these different things. So we have now more sophisticated ways of looking at sleep and actually connecting it with the neurophysiology, which is also very exciting that, uh, you know, I’ve been looking at EEG for a long time. I did that also in grad school with Steve Hillard and Marta Kutis and learn and all about E R P methods and how to use EEG. And you know, we’d see these signals in the, in the brainwaves and try to attribute them to some psychological process mm-hmm <affirmative> and now we do the same thing in sleep, in a sense.
Ken 00:31:15 So we see the slow waves and we try to say, what’s going on there? Is it just a, a, you know, a distant reflection of something? And now it seems that no, it’s actually part of the mechanism that the slow waves are actually indexing a neural firing that’s connected together that lots of neurons are firing at the same time at the upstage of a slow wave. And that that’s, uh, part of the mechanism that helps the other aspects of brain waves, such as the sleep spindles and the ripples and you know, these cute names for different types of signals that we can record that are giving us, uh, some indication of what’s happening in the brain. And so now we know that, oh, those particular measures correlate with whether memory gets improved. So we, we can actually tag those measures and say, well, they’re part of the physiology.
Ken 00:32:07 We don’t know the full physiology. Of course, we just have these hints, but those hints say there’s some important physiology happening during sleep where memories are processed. And to the extent that that happens, your memory is different often better when you wake up for those specific memories. So we, we can, we don’t just have to say, well, is sleep better for memory than wake is cuz wake is good for memory too. It turns out. Right, right. But we’re actually saying well, but we’re sleeping every night. And the sleep is contributing to our memory abilities. Cuz we can see these relationships of the sleep binds and the slow waves indexing part of the critical processing that we can now like measure and connect that to the behavioral improvements in memory that we seen. And so that’s the, the targeted memory reactivation method that you mentioned, one of the ways we do that, cuz we can ask, well, if we pro, if we kind of provoke you to reactivate a memory during your sleep, what are the repercussions of that?
Ken 00:33:08 And we’ve seen that in a lot of examples that it changes how well you can remember later, in other words, not surprisingly there’s neuroplasticity <laugh> and when you, you reactivate a memory, the, the neural circuits that are involved, don’t just statically do that and stay exactly as they were, but they change each time. You use them a basic principle of neuroplasticity and it happens during sleep when you reactivate memories and then see the benefits later. So we have all these different angles now of understanding is sleep important and, and allowing us to dig in and try to understand the neural mechanisms better and therefore understand well how what’s a good night’s sleep all about <laugh>.
Paul 00:33:51 Uh, so you mentioned, you know, the different EEG signatures that you can, uh, observe and tell when someone is in different stages of sleep. Although, uh, you also mentioned that it’s more of a gradient than, you know, these, these clear lines and you also mentioned like sleep spends, right. Which are associated with, uh, replay events in the brain where, uh, which are thought to, you know, where you’re replaying your memories forwards and backwards. And that’s thought to help consolidate those memories within the, you know, cortex, but you know, at least sleep well, at least replay. And I assume sleep spends happen when you’re awake. Also, is there something that happens during sleep that is fundamentally, uh, new or different than, or, or something that happens in sleep that does not happen during wakefulness? Or is it all just gradient that do things in sleep get ramped up and some other things get ramped down or, or is there a nice bright line where we can say this happens during sleep and it doesn’t happen during wakefulness?
Ken 00:34:55 Well, you know, there’s some of both interestingly, and if you take an example, like the hippocampal ripples, those can be seen both during sleep and wake and lots of studies are getting at what the functional relationships are between those and memory processing. Uh, and then we can study, well, what happens when you practice some information when you do rehearsal while you’re awake and what are the neuro chorals of that? And you know, lots of work. That’s what I worked on a lot before I even studied sleep. Uh, lots of work on that, but as you’re looking at the EEG, the EEG is really quite different during sleep. So this occurrence of slow waves and the, the timing of those with the spindles, that’s not something that you see during wake. So that’s, you know, the, I think we could ask the question, you know, more broadly speaking, what’s the difference between rehearsing information when you’re awake and consolidation moving forward and rehearsing information when you’re asleep also moving consolidation forward.
Ken 00:35:52 And that’s still an open question and there, there may be many similarities, but one suggestion that makes you think they’re different is this occurrence of slow waves and spindles during sleep. And it’s not the pattern of a, you know, a person studying information and, you know, reading their textbook and trying to learn. You don’t see those patterns of brain activity. You see the waking EEG, which is quite different than the stage three, so called slow wave sleep Sage. So just at a broad level of brain activity, it looks quite different. And there, again, I’m open to the idea. There might be some commonalities between what’s happening. There should be some, but there also seem to be some differences. And, and that’s kind of something to dig into as saying, what is special about sleep and memory process thing. And, and of course, one thing that’s different about sleep is you’re not being bombarded by so much sensory information.
Ken 00:36:44 You’re not walking around. You’re not getting a lot of information. Your eyes are closed. And you know, that kind of connects with it, the old way of thinking, which was that when you’re sleeping, you’re pretty much blocking out sensory information. And then I allow your brain to just work on things that it’s busy working on. And that, uh, was, was a Orthodox view that we kind of defied in our research because we went ahead anyway and said, well, let’s present sounds to people. And the old thinking was, well, if you present sounds or, you know, words or anything to people, right? Yeah. It it’s gonna wake them up. And if it doesn’t wake them up, it’s just because it’s being blocked out and nothing’s getting in. So you shouldn’t see anything working at all. In fact, I have a nice little video clip of Matt Walker saying that in one of his lectures about how, you know, it’s all blocked.
Ken 00:37:33 He shouldn’t expect anything to happen except maybe some old factory information could get in. Uh, but we, you know, we, didn’t not a sleep researcher for a long time. We went ahead in my lab and did a targeted memory reactivation study in, in 2009. And it was kind of modeled after a prior study that was done by, uh, Bjorn rationalist colleagues in Germany published in 2007. And they had presented odors during sleep mm-hmm <affirmative> and the odors were presented. Both people learned some spatial information prior to sleep and, and it was a rose odor. And that rose odor was presented again during slow way sleep or during REM. And then when people woke up, they got tested on the spatial knowledge and it turns out if they had been presented both during learning and during slow way sleep memory was improved for that information. If it was presented during learning and during REM, they didn’t see that improvement.
Ken 00:38:27 So, um, that result pushed people away from focusing on REM because in the prior decades, if people theorized about memory and sleep, they always said it must be REM sleep and dreaming and all that. And here wait, all of a sudden, slowly sleep gave them the result. And importantly, they didn’t get the result if you didn’t present it during learning. So if you only got the rose Oder during slowly sleep, no benefit, it’s only that it connects. So the idea was that the rose odor reminds you of your learning. And then you start thinking about those objects in their S spatial locations. And remember that information better. So that’s where you call it targeted memory react, activation. The odor made you reactivate those memories. And in my lab, we went ahead and do the, the same sort of experiment, except we used sounds instead, not believing, not supposed to do the Orthodox, not supposed to do that.
Ken 00:39:18 <laugh> it was taboo. In fact, you know, I think because if you look at it from the 1950s, people started doing some experiments like that, and they were thinking, you know, we can do things to make people better at their sleep and learn stuff even. And some papers came out saying, well, actually all that finding is flawed because they didn’t measure of the EEG. And if you measure it, you can show that if people learn anything during sleep, it’s because they woke up for a moment. Right. And they learned it and it’s all an artifact. And so decade after a decade, sleepers stayed away from that because it was just, you know, if they did studies like that, they just kept them in their lab and didn’t publish ’em because they were ashamed of, of doing such things. And then it turns out, you know, that the views kind of gradually changed.
Ken 00:40:00 And now we can see no, actually some information gets in odors, particularly because bur and rash thought, well, let’s use odors because that’s not gonna wake people up. They didn’t want to use sounds. Oh yeah, we use sounds, but presented them very softly, just, you know, whispers. And we found that the memories that we reactivated for spatial information, again, it was, they get words, you know, or sounds like a sound of Meow. Uh, when they see the cat and they have to learn the location of the cat, they get the same sound and another 25 sounds during sleep. And then their memory for locations is improved for specific memories that we reactivated with those specific sounds. And it has to be that the, you know, the, that behavioral result says, those sounds must have done something. They must have been processed by the brain. Furthermore, the sound was a Meow and your memory performance wasn’t tell me what a cat makes <laugh> what’s outta cat makes.
Ken 00:40:57 It was well, the Meow made you think of the cat in a sense while you were asleep implicitly, perhaps, and then that made you think of the location of the cat and your memory performance was your location of the cat is improved. As you try to remember, where was that cat? Where did I see it last? And you identify that location more accurately because you a sense reactivated that memory during your slow wave sleep. So, uh, and now there are many, many studies showing that, that in fact sounds aren’t necessarily blocked. So that’s getting back to our starting point. Not a lot of information might come in while you’re asleep, cuz you’re not walking around, but it doesn’t mean your brain is off. And it doesn’t mean your sensory pathways aren’t working. So sounds can come in. And that influences what memories you’re processing. But normally, maybe there aren’t that many sounds and memories are being processed every night, uh, in your night, peace and quiet, uh, of your sleep because you’re not doing anything else.
Ken 00:41:56 And your brain is busy working on that stuff for your benefit. And actually I think not randomly, but I think if you have part, and this is an idea from, from Winston and cart rate, for example, they talked about how maybe during sleep, you’re bringing up memories that you recently acquired, but also your goals and what’s going on in your life and what are the challenges you’re facing and what are the problems that you need to solve in your life. So bringing up the recent information on having this trouble with my friend and maybe bringing up old information that connects to that same issue and how you’re struggling with it, and maybe something that’s gonna help you get a solution. So I think we bring up information that’s on our mind and that’s, that leads from one thing to another thing to another thing. And that’s the memory processing that happens during sleep, not just anything coming up, but things that connect to things that are on your mind the same way.
Ken 00:42:52 If I just tell you to, you know, your eyes and, and, and kick back for a few minutes, your mind’s gonna start wandering, where does it wander? Well, seemingly random, but with some structure to it, you know, what’s on your mind, what’s been bugging you, you know, the issue that you suppressed and now it comes back up and you’re thinking, how do I gonna solve that problem? And, oh, I had that same problem when I was 14. And I remember this thing that, you know, so it can connect to memories and that’s all part of connecting memories together and having consolidation move forward, not just so you have memories, but also so that you can deal with the struggles you’re gonna deal with the next day and potentially have ideas about how to cope with what’s going, what’s gonna happen, the problem you might need to solve.
Paul 00:43:36 Yeah. Well, my dad used to, you know, go to sleep with a problem and wake up with a solution that doesn’t happen to me. Um, maybe not as bright does that happen to you?
Ken 00:43:44 You know, we study that in the lab and I think that’s one of our really exciting studies with, uh, mark Beman. Who’s a colleague of mine at Northwestern who studies problem solving and creativity. Yeah, yeah. And so he and his student, uh, Kristen Sanders got together this set of puzzles. So, you know, they’re not solving, uh, world hunger or something, but little puzzles that they would, they would give to people. And, and sometimes you could solve it right away. But a lot of times they couldn’t come up with solution. So everybody got six puzzles that they failed to solve in the lab. They went home and had a good night’s sleep, uh, wearing some portable, uh, sleep monitoring equipment. We gave them and the equipment registered when slow wave heat sleep happened. And then presented three sounds that had been related to three of the six, three of the six problems that they had.
Ken 00:44:39 So basically we, we reminded them of three problems and then came, they came to back to the lab. The next day we gave them the six problems again and measured how often they solved them and their ability to solve the problem went right up quite a bit for the ones we were activated during the sleep. So that’s the problem that you haven’t done that your father was good at is you need to have a method to get the problem to come back to you while you’re asleep. Maybe you’re just thinking about something else during sleep, but we showed that if we can provoke you to think about a specific problem, you’re more likely to come up with a solution later. And so that, you know, connects with the old RIS about, you know, sleeping on it, thinking it over, making a better decision. I think your brain would be doing that.
Ken 00:45:25 And we just manipulated it experimentally to try to get a handle on it. And to be able to study that, you know, it’s, it’s kind of a, a thing thing happens, but most of the evidence was anecdotal of just, you know, Paul McCartney woke up with his song in his mind and somehow it came to him in his sleep, but we don’t have, you know, evidence of that happening. Our study kind of moves us forward as saying, no, actually we can, we can actually provoke it happen. And then study, what is, what of the brain mechanisms that helped you get to the solution? So I think your dad was certainly on to something
Paul 00:45:58 You, you know, that’s slightly apocryphal about Paul McCartney. He woke up with, I think one or two chords in his head, but not like the, I think it was yesterday, right? The song, it wasn’t a complete song by any means. He had to like work it out in the morning. So anyway, <laugh>,
Ken 00:46:12 And, and there, there are lots of other examples like that too, of people coming up with things. And, and there’s a cool study from my colleague, a Delphine Odette in Paris, where you could, um, just have that happen at the, at the sleep onset period. So stage one, just as you’re falling asleep can be a, a, a huge time of creativity. And you can show that if people, each that stage, uh, and, and have a little time in it, but don’t go fully to sleep instead, wake up, which was Thomas Edison’s method. You wake up, I’m
Paul 00:46:42 About ask you
Ken 00:46:42 About that. And you’re more likely to come up with a solution not immediately. So interestingly, almost like the example you gave with long Heartney, the, the results were that after some minutes they would eventually get to the, but they would get to it more often if they had that period of this sweet spot of just a little bit of, of this, this one period of sleep. And of course we think REM sleep might be like that too. That’s later in the night. So, you know, sleep has all this different power and different different components, uh, to look at.
Paul 00:47:14 Is that true? Um, sorry to be all over the place here is that first of all, is that true about Thomas Edison? So the story is he held like metal balls and would sit in a chair and start to doze off. And the purpose of holding the balls was that when he dozed off, he would drop them and it would make a clatter and wake ’em up. And that you think that’s true.
Ken 00:47:32 <laugh>, that’s exactly the story. I mean, he described it. We, no one was recording his EEG at the time. So that’s, you know, this study in Paris is so much better cuz they could actually say here’s what’s happening and they didn’t right. You know, people can hold a ball, but they could monitor their brainwaves and do it as well. Salor Dolly also did something like that, that he ed in a, in a crazy book about, you know, how to, how to get his, uh, creativity going, uh, by making use of this period. And, and there’s something, uh, something else that, that folks around the same time, uh, uh, did in sort of having this, this kind of sleep session of trying to go to sleep, but use their sleep to provoke their creativity and, and having kind of a, a in between stage of not quite asleep and, and not quite awake and, and again, all anecdotal stories, but we think there’s really something to it that you, we can go back now.
Ken 00:48:26 And it’s just like the sleep learning stories from the 19 and fifties of, yeah, it’s not amazing things don’t happen where you just acquire new knowledge while you’re asleep. But interesting things are happening in the brain. If we go back now and look at it, we can actually, you know, discover a lot more about what’s happening in the mind that I think has been ignored, uh, by a lot of sleep research and, and, and think of, think of how important dreams are for us that, you know, all, all kinds of people have dreams and like to talk about their dreams. And <laugh> how, you know, incredible creativity. And yet, if you, if you look in psychology departments across the country or elsewhere in the world, hardly anybody is studying dreams. It just didn’t take on as, you know, a scientific indepth ever that that seemed to be paying off for people, because basically it was so limited and just kind of isolated, you know, separate from sleep researchers that seemed to, you know, you can go to a sleep conference and you hardly see anything about dreams at all.
Ken 00:49:25 So it’s kind of just pushed away. I think in part, because the methods were so limited. If, if the only thing you do is talk to somebody and say, uh, so tell me about your dream. What did you dream about <laugh>? And that’s the only data you get, it’s really limited and hard to make progress. And now if we, we, we kind of shake on the neuroscience of dreaming. We can add a lot more to it and say, well, what’s happening in the brain? How does that connect with the stories people have? And how does that re relate to other phenomenon of sleep such as memory processing and the things we’re looking at? So I think bringing it all together is gonna enrich our understanding of all these different components of sleep.
Paul 00:50:06 My wife hates it. When I tell her about my dreams, she says, it’s the most boring thing in the world to hear about other people’s dreams. But I imagine these days you hear about a lot of people’s dreams, given the, the lucid dreaming work, especially that you’ve, you’ve been working on.
Ken 00:50:20 And don’t you like talking about your sleep, your dreams
Paul 00:50:23 It’s love, love it. It’s
Ken 00:50:24 It’s, it’s this thing happened to you this in your own, the original virtual reality, this, this, this amazing thing happened. And of course you want to talk to people about it.
Paul 00:50:35 Yeah. But I wanna talk about, uh, your lucid dreaming work here, because I, I wanna make sure that we don’t, um, skip it, but, um, I wanna back up also and ask you, uh, kind of a broad about memory. So, you know, talking about sleep and how that, uh, helps solve problems, essentially bringing new things together in a creative mode. I used to think of memory as the storage of past events, but as I’ve come to appreciate more, the notion that our brains are for movement and in action, I’ve come to think of memories, imagination as possible future actions or planning, or as future oriented, essentially. Do, do you think of memory as future oriented or do you think of it as a store of our past? Or how do you think about that?
Ken 00:51:22 I, I think you’re, you’re right on when you, when you say that, you know, we can use our memories to reminisce at, and, you know, we can enjoy doing that and thinking about the things that we’ve done, but it’s most beneficial for us when it guides our decisions about what to do in the future. And we can say, what, how should I behave in this circumstance? Remembering things that have happened in the past and have that knowledge and use it accordingly. So of course, memories are most important for how they guide our big behavior, how they guide our decisions.
Paul 00:51:50 Okay, good. I’m glad we’re, we’re in agreement on that. So, but
Ken 00:51:53 Some people just like to remember, it’s, it’s fun to remember things and think over it the same way. It’s fun to remember your dreams. So you can kind of ask the same question. Well, what is my dream useful? Or is it just, you know, some imaginary thing that happened and again, and also you can think about dreams as how incredible real they seem. So when you’re in a dream, you think that it’s reality. You think that you’re awake in other words, and that you’re wandering around and it’s only at the moment you wake up, usually where you say, wait a minute. None of that actually happened. That was all in my, in my mind is just something I dreamed. And, and, and you can think of imagination similarly, right? Like I can imagine something happened. And when you do that, you know, it’s, it’s not quite as real.
Ken 00:52:38 You’re not really immersed in it the same way you aren’t a dream. So a dream is, you know, really supercharged imagination where you’re actually there and you think it’s real. And that’s, that tells us something pretty interesting that the reality we have during the day, it’s the same process. It’s our mind can make a reality and it has, you know, all the features of our world in it. And the very same thing happens when we’re dreaming. So we actually don’t need the sensory input to have an experience of being here right. In this moment. It’s something that our brain’s equipped to do. In fact, it’s what our brain does to make us think. We’re in a world and, uh, that sort of puts another light on, you know, what is our reality about if it’s something that our brain can, can make so real, um, just in our dreams.
Ken 00:53:30 And then it connects that that should connect to what you mentioned, lucid dreaming. So lucid dreaming is when you’re in a dream, but at the time, you know that that’s what’s happening. You realize you’re in a dream and the dream carries on. And then that has a different character cuz amazed that look how real everything looks. And, but yet it’s, it’s just a dream. And then you can go sometimes one step further and say, I wonder what’s gonna happen next. You know, maybe, you know, my old friend is in the next room, let’s go and see. And you know, you might be able to meet someone that you wanted to talk to. You can in fact, change the dream a little bit here and there to do what you want to go, where you want to make things happen the way you might want to.
Ken 00:54:09 And so that’s how lucid dreaming can be so exciting for people that like to engage in it. And yet it’s a rare thing. So most people don’t have very many lucid dreams and the people that are good at it, maybe they have them more often, but not always on command. And so lucid dreaming has been challenging and there are lots of books written about how to have lucid dreams more often. And our approach to that is thinking about targeted memory react activation. So what we do is we take our same method presenting, you know, not the Meow sound connected to the cat, but now we want to connect to what you just learned before sleep. That is about thinking about your current experience. So am I asleep right now? You could ask or am I awake? And you can think about that carefully and try to think, well, this does everything seem like my waking world or is anything weird and you can get in a mode of checking that.
Ken 00:55:05 And we have people do that before they go to sleep connected to a sound. The sound might be a little violin hitting a few notes or, you know, whatever we want to use, we connect that experience. And when you hear that sound you’re summary, oh yeah, let me check. Am I awake now? Or am I asleep? And then our method of course is when you’re asleep, we’re monitoring the EEG and we can see what stage you’re in and we can present that sound again and remind you to get in that mode of carefully checking your experience and that can provoke often people to go into elusive dream and understand, oh yes, I’m now an elusive dream. Uh, let’s explore it. And furthermore, they know the experimenters are recording my brain activity now as well as my eye movements. And I can talk to them with signals because I can’t talk to them with my mouth because in your sleep you’re laying in bed.
Ken 00:55:57 You’re not moving, you’re not speaking, but your eyes can move. And so we have people arranged so that they can move their eyes left right left. Right? And we can see the electro activity that that gets generated from our, and they can tell us, I just realize I’m in a dream, they make a specific left right signal to tell us that. And then we know they’re in a dream and we can continue to, uh, monitor their brain activity or also provoke them to have a particular type of dream. So our first experiments to try to show if we could communicate with people that we published last year were to present them questions that they could answer. And we decided on math problems, because then we would know what the right answer was. And we could tell, did they hear the problem correctly? Cuz we weren’t sure there would.
Ken 00:56:43 And if they heard it correctly, could they actually think logically to compute the answer and give us the answer? And we found that in fact, yes, repeatedly, we could get people to hear our question, think about the answer, signal the answer back with their eye movements or now we use sniffs. So you can also sniff with your nose go and you can make a little code. And, and we can measure that because we can measure the air going in and out of the nostrils with a little tube and they can answer our questions. And then we can ask them, uh, in the future studies, what are, what are they dreaming about right now? Are you, you know, and understand more about dreams in real time getting the information. But our first study was just demonstrating that we can in fact, communicate in real time with people, we can have a dialogue, we can ask questions, they can answer the questions. And that, as you mentioned, got a lot of interest in the press because it’s sort of a, an exciting discovery that sleep clear, something we didn’t know about sleep. We didn’t know would be possible to get information in and get information back out from people so that we can communicate with them and learn about their dream. In, in real time with this method,
Paul 00:57:52 This is a crude question, but why do we care about communicating with dreamers?
Ken 00:57:57 Well, I think it going back to what I was saying before, if you wanna study dreams and you’ve only got what they said, when they woke up, you don’t even know exactly when that happened and yeah, we wanna have a neuroscience of it. So we wanna measure the brain activity, understand what is that brain activity telling us and connect that to, and what’s their experience at that moment. And we can’t ask them about their experience when they wake up because they can’t really connect or they forgot a lot of it. Or in fact, sometimes it’s started and our studies also show sometimes communication happens and there’s a real event we know happened and they wake up and they don’t remember it. <laugh> yeah. So, you know, we forget so much, you know, so another aspect of dream research is that you’re just measuring a little teeny bit of this whole, you know, it’s having lots of dreams every night and you’re remembering a teeny little fraction of it.
Ken 00:58:50 So we want to get better. We want to really probe what’s happening during sleep and ask some questions about, uh, what’s what the process of the dream is and connecting it back to the other research of, and how does that relate to memory consolidation mm-hmm <affirmative> and how does that relate to problem solving and the creativity of your dream, which is amazing. And how does that relate to what, what is the brain, why is the brain doing that in the first place? Cuz I think that’s a wide open question that we don’t really understand. Why do we, why do we have these dreams? Why do we have slow wave, sleep interspersed with REM sleep and going back and forth, there must be some good reason for that. And we’re still struggling to put together stories about how that makes sense. And I think we have to think about, you know, not just asking about, uh, one aspect of sleep, but think about the whole combination of sleep and how does it promote, uh, better health? How does it regulate our cardiovascular system and how does it regulate our, uh, psychological, um, endeavors thinking about the problems we’re having and so forth. So trying to put that whole package together and keeping dreams on the table too. <laugh> as part of it, not just, uh, uh, one we wanna push aside because it’s so messy to study.
Paul 01:00:02 Did you, did you ever think that you’d be studying dreams when you first got into neuroscience?
Ken 01:00:06 Well, I think for, I think I always thought dreams were fascinating, but looking into literature, it seemed like a big mess and you know, you didn’t want to get in there because there wasn’t, you know, didn’t seem like much progress was being made. And so now I’m a lot more and I have these, uh, wonderful students and postdocs that are working on these issues and are gonna take it to the next level and, and move dream research to a better place. So yeah, I’ve shied away from it because of the, the history of, you know yeah. Uh, theories that just weren’t test about dreams. Cause that’s not really good science and now I think no we’re gonna have a better set of theories and a better ways to test them. And you know, more things to measure that can allow us to then make some progress and really understand more about it.
Ken 01:00:51 And, and then circle back and sort of try to ask the question well and how is this useful people for people? How, how can we be sure that sleep is, uh, functioning optimally, if it can, or, you know, it’s part of a lot of problems that people have. If you look at PTSD or depression, there are components of sleep that aren’t working well. And you can think of that as a symptom of these disorders, but perhaps it’s also really part of the cause is part of the problem is it’s contributing. What’s, what’s not working right in sleep is attributing to the problems people have during the day. So I think correcting those problems, uh, might help us all sleep better if we can sort of figure out how to, you know, what are the, what, what are the ways in which sleep is helping us and what our mental activity during life is giving to us. And to what extent is it hindering us or hurting us, uh, if it’s not working well.
Paul 01:01:44 So I, I guess like any experimental setup, you are essentially perturbing someone’s ongoing cognition, right? They’re otherwise natural, ongoing cognition in the world by your experimental variable. And of course when you induce halluc dream in someone, um, part of the, part of the reason is like you said, you can, you can change their dreams right. Intentionally, um, and, and have them well, the, the, the question is, um, how confounding is it scientifically and how much do you worry about how you’re affecting a, their sleep? You know, I know some people wake up when they’re cued, but you know, know you throw those studies out, right. That they’re not valid. Um, but just interfering with their sleep in any way. Does that worry you, that it’s affecting the interpretation of the results?
Ken 01:02:30 Yeah, I think that’s a fine point. And, and the best answer to that is, well, we, we need very multiple methods to use. We need to attack, we need attack these problems in many different ways and manipulating what’s happening during sleep is one of the methods. And as we do that, we’re seeing more and more of the physiological signals that we could measure out of the EEG and try to understand them. And then we can go back and other studies don’t present sounds during sleep, but still measure those signals and see how that relate. So that sort of gives us a potential of convergence between the different methods so that we can, again, exclude this idea that the, the, uh, uh, manipulations of sleep, isn’t all just an artifact and not what’s really happening. If we can go back to undisturbed, sleep where there aren’t noises and see the same sorts of things happening and the relationships, the spins and the slow waves and the later memory performance.
Ken 01:03:23 So I think what we do have is actually convergence from different methodologies, different ways to study, sleep that, uh, point in the same directions. And, and yes, we can be more convinced by the ideas that get demonstrated in different ways. For example, this, this <affirmative> the importance of slow waves. There’s studies that show that you can induce more slow waves with, uh, entrainment methods that are, could be electrical or auditory and entrain, more slow waves. And the result of that is often improved memory. And so that’s a correlation between memory and slow waves that’s demonstrated there, but you can also see that in other studies where you’re not perturbing the slow ways and also see the relationships and in our studies where we’re presenting sounds during sleep, and how does that connect to, well, what phase of the slow ways is it happening and actually, does that make a difference? It seems to make a difference. So it also converges on the idea that something’s the important is happening in these with these slow waves and then bring in the spindles as another measure. So I think, uh, you’re right, that every method has its potential sure. Shortcomings, and we, we have to not sort of deny those, take those on board and then think about how do we make progress with other method that have different shortcomings and bring them all together in that sense. So that’s the typical method methodological approach,
Paul 01:04:49 Uh, yeah, sorry. I’m sure you get asked that all the time, uh, you were talking about future studies and sort of expanding the range of types of questions and the vision of someone being able to more, uh, thoroughly describe the content of their dreams and their conscious experiences within dreams. So immediately I thought, well, there’s, you know, originally you’re asking math questions, very simple math questions, because it depended on how many eye movements someone can make. So you can’t ask 97 times three because that’s too many eye movements, for example. So there seemed to be,
Ken 01:05:20 Yeah, no one subjects MIS misheard our question and they, they said, how am I gonna give you 22 <laugh> Cause they misheard what the problem was
Paul 01:05:28 After they woke up. They, they, uh,
Ken 01:05:30 They
Paul 01:05:30 Told us that’s what they, yeah, yeah. Um, anyway, but, but then I had, uh, just as, as we were talking, I had this flash of, you know, you could potentially hook someone one up, uh, to a brain machine interface, right. The EEG and decode the, their somehow decode their thoughts into a curse on a screen, right. Where they, or, or like a keyboard where they could type in their dream and just write out the contents, um, slowly, I suppose. But so it, it me start to think like, what are the current limitations or obstacles that you see many of which I’m sure there are, but what is like right now, grinding on you, what are you, what problems are you solving in your sleep BA you know, uh, to, um, to make these future studies plausible and beneficial.
Ken 01:06:18 Uh, so are you asking specifically within studies of lucid dreaming, how to, how to, how to make the communication work better?
Paul 01:06:25 Well, uh, it’s all connected. I know. So, um, it could be that, but you might have some other pressing problems just related to cognition and memory and, you know, what’s, what do, what do you feel like you’re up against now that you need to get past, to make real progress?
Ken 01:06:39 Yeah, the there’s always a string of such problems. Uh, here’s one, we we’ve been trying to use our methods, not only in the laboratory, but in people’s homes so that they can, as I mentioned in one of the prior studies take home some of the technology, some wearable equipment to monitor their sleep. And is this the app percent? Uh, it’s, it’s connected to that. So we we’ve, we’ve done some studies at home with different kinds of apps that can monitor, take the signals of, of we get from monitoring sleep either from a headband or from a Fitbit watch that measures heart rate to, and we can try to use that to present sounds to people, but we don’t know exactly how noisy their room is. And do they have a fan on, do they have, you know, music from their neighbor playing? So we’re, we’re, you know, it’s not as controlled as the laboratory, but yet we wanna make that work.
Ken 01:07:31 And one of the things we found that we didn’t expect to find was that actually the sounds presented during sleep can sometimes make memory worse and, uh, together with some other research, we have the idea now that if the sounds arouse you from sleep, then it’s detrimental to memory. In our laboratory studies. We are always very careful and we would monitor sleep. We’d be looking at the EG. And if the sounds were too loud and we saw potentially you were gonna wake up, we’d back off and stop sounds and then present them softer. And so we could do that online. We got good at that. And so our studies were working and memory improvements were seen with skill learning with spatial learning, lots of different types of memory, face name learning. And then we saw that, uh, uh, in the home studies, it wasn’t working so well.
Ken 01:08:20 And then we were able to look more carefully at the data and see that, well, actually, when it was working, when the sounds were too loud, it was actually going the other way. And the sounds had to be really down to a low level to not disrupt sleep. And so now we have that understanding that we’re gonna take forward and say, okay, we have to be really careful when we do this at home, because it’s not the same as being in the lab where we can adjust it moment by moment. But we have to kind of, uh, connect the, the software to actually be able to do that and to be monitoring moment by moment with measures and then change, uh, on the fly. And so that’s, that’s sort of an interesting, uh, aspect of it that we hadn’t realized at first, because we kept getting results that were showing memory improvements and not realizing that we actually had to, we really had to titrate that sound level quite carefully to be not too soft, not too loud.
Ken 01:09:16 And we, we, I guess we got lucky when we did it the first time, but we picked a good level. That was, that was in there. Cause you know, we were, we were worried about waking people up. We knew that if wake people up, that’s no good, cuz then we won’t know if it’s actually a sleep phenomenon or not. So we had to really be way careful on, on, uh, on the parameters to be sure we weren’t waking people up. And so now, uh, as you mentioned, we have a, another app we’ve made, made available publicly. And so, uh, the studies we’ve been doing with the apps have been Mo mostly focused on slow wave sleep and the memory improvements that we can get from reminding people of what they learned before sleep. And that’s exciting. Uh, but this other world is what about presenting sounds during REM sleep that can provoke them into having El Luci dream and then provoke them into a particular lucid dream that they maybe wanted to have, but didn’t remember that that’s what they wanted to do.
Ken 01:10:14 And so the first step is provoking the Luci dreams and with Karen Conley in my lab, we’ve developed the methods to, uh, and, and, and uh, Michelle Carr and others have worked on the methods to try to make that work with the training before sleep. And so now we’re doing that training on an app that then you can, uh, use in your own home, get the training, go to sleep and a sound again, when the, uh, technology thinks you’re in REM sleep. And so we’re doing that with a Fitbit that connects to a, uh, a smartphone that’s running our software and controlling the whole thing. And that’s what we’ve made available on the internet for people to download. They can get that software, but they have to have their own smart art phone. That’s an Android phone and they have to have their own Fitbit. That’s the kind that our controls heart rate, but if they, uh, measures heart rate, but if they have those two things, they can go ahead and, and run the software and try to have elusive dream. And we, you know, we’re not, we’re not marketing this as something that already works. Rather, we’re saying it works in our laboratory and we’re trying to get it to work remotely. And people are welcome to try it and help get us the data, you know, more citizen science that can help get us the data to show when does it work and help us perfect it and make it better. And so that’s, that’s where we are now. So at my website for my lab, which maybe you could give somewhere
Paul 01:11:42 I’ll link to it.
Ken 01:11:42 Yeah. Information. I I’d love people to go out there and use this again. If you have an Android phone and a Fitbit, you can download the software for free and give it a try and we’ll get the data and we’ll find out, well, how many lucid dreams do people have? And is it getting close to what we can do in our laboratory when we can get, I don’t know, sometimes 50% of people can get El lucid dream, uh, with one night of this thing. So we’d love to be able to make that work in the home. And again, it’s a challenge of presenting the sounds not too loud. <laugh> yeah. Not too soft and at the right time, cuz the sounds have to be there when they’re in REM sleep. Ideally, uh, at least that’s what works well in our laboratory. And, and yet we’re not measuring the EG, so we’re not actually getting the best measure.
Ken 01:12:33 One could get a REM sleep, but we’re getting an approximation, you know, the, the wearable technology, the Fitbit or the aura ring, they’re not the same as in the laboratory, but maybe it’s good enough for this purpose to actually hit sometimes, uh, and, and give people elusive dream. And then the next step after that will be, uh, you can perhaps, you know, imagining a future version of this software where you could put in, well, what, what do I wanna have my lucid dream about? Once I get into my lucid dream and you could have another sound that would cue you to remember what it is you wanted to dream about and then control the dream. Like let’s say you have some problem and you wanna talk to someone that could help your solution. Maybe your old dad or your grandfather could help you or maybe Einstein. And you could say, you know, I think Einstein might be in the next room. Let’s go check <laugh>,
Paul 01:13:25 But Einstein is created in your head. So yourself. Yeah.
Ken 01:13:29 Yeah. It’s all, it’s all you, but it’s unleashing a different kind of creativity than that. You might not so easily access in your waking state, but yeah, it’s not, it’s not actually Einstein <laugh>
Paul 01:13:42 That’s that’s presumptuous. I’m gonna go ask Einstein, AKA me about my problems. I would be amiss not to ask you the, you know, the, the podcast is mostly about neuroscience. It turns out, but I’m also interested in AI and the interface, artificial intelligence and neuroscience. And you know, through the years there have been sort of nods. One might say marketing nods towards sleep as being important. It’s known sleep is important for whatever reasons, like the reasons that we were talking about. And so you have things named like the sleep wake algorithm from a dozen or two years ago. And you know, even more modern artificial networks are doing things like adding sinusoidal noise at various times during training to sort of mimic some of the function of sleep. Uh, but really there, you know, I know that the science of sleep is an ever advancing frontier and you know, we’re not, we don’t have all the answers yet, but if I were an AI researcher and I asked you, what, what is it about sleep? Like what, what can I take from sleep to incorporate, to make my models better? What would you say?
Ken 01:14:52 Well, that’s a good question. I think, uh, all the things we’re learning about how the human brain works could potentially inform AI, I guess, and the questions are which, which of the nuggets that are gonna be most beneficial. So, you know, the goal of AI isn’t to recreate what the human brain does, I’m sure, but you’re, you’re saying, but maybe there’s some good hints there that could help us do it better. And so offline processing of memory is something that my computer doesn’t do and I’m quite happy with that. You know, I put a file on my computer and then I go get it later and it’s the same file and it hasn’t changed.
Paul 01:15:31 That’s good. Yeah.
Ken 01:15:32 But all the information we put in our brains doesn’t stay statically stored that way. And, and yes, we have forgetting, but we also have generalization. We have, you know, interesting things that happen as we consolidate and do this offline memory processing. So, uh, I think we haven’t understood it quite well enough to say exactly how to do it, but yes, offline memory prostate could be beneficial for a system to work through some of the information that it has. So if a, if a AI has, uh, gained a lot of knowledge, perhaps it should, uh, try to organize it in some ways. And, and this mechanism we have during sleep is kind of a self power reorganization. And like in one article I wrote for frontiers, for young minds with just a journal for kids, I wrote, uh, does, uh, uh, how did the title go do, do, do house cells clean your brain while you sleep?
Ken 01:16:29 Yeah. And so the idea here is that somebody’s in there organizing things and probing for connections that make things fit together better and working with the information. And so this offline memory processing could help organize things so you can find them better. And how do we find them? Well, we don’t find our memories by looking up the code of, of what, what bit stored it. We don’t register by the date and saying, tell me things that happened two weeks ago, last Wednesday, we don’t have, you know, that’s not the way memories are organized, they’re organized by the content. And so there is sort of content addressable that you have to say, well, what’s, what’s the related thing that happened two weeks ago and, and eventually get to the right memory and work your way there. And so that structure of organization means that you have to have all these connections to see how things are related to each other, to eventually find anything.
Ken 01:17:23 And I guess AI doesn’t necessarily work that way, that, you know, the information structured in a way that, that kind of, you know, where different things are put and in our brain, it’s all put together. And as we like to say, memories, aren’t stored in a vacuum, but they’re stored in this interactive way, which we think actually comes up during sleep. One of my postdocs latest papers is about how, when we activate memories during sleep, it’s not just a specific item that gets activated, but the context of that, and then the context, uh, of all the interrelationships that might connect are part of the reactivation, not just the single thing that you want to connect, but all the interconnections. And so that seems to be part of the way our brains are organized. And could there be some way to organize AI structures that would have the benefits of that <laugh> I guess, and the costs of that would come too. Yeah. So, so you’d have to decide whether that’s good for your structure.
Paul 01:18:21 Yeah. I mean, there’s also the different phases of sleep right. And different cognitive functions associated with those different phases and wakefulness as we were just, um, talking about earlier. So it’s, I mean, you know, in the one sense it can sound silly. Like, well, what can you, should you make a, a network that has different states of wakefulness, you know, and, and sleep. That sounds kind of silly. Right. But, but like you’re saying maybe there are some principles that we can incorporate, you know, like replay is a big one that, um, that has been built into, of, um, machine learning networks. And, you know, there are, uh, models out there that show that replay is important, likely for generalization, et cetera. But that also happens when you’re awake. So, you know, it’s not clear if there’s something definitive about sleep and dreams, uh, and, and different phases of sleep, et cetera, that would really benefit, um, artificial intelligence, but, you know, your overall picture, do you think that it would be important to include some principles, whatever those principles might be like you were just talking about? Or do we want to these static entities, right. Oh, oh. He, he just rolled his eyes folks, you know, it’s an audio podcast so that they can’t see that, but it was a kind eye roll though. Go ahead.
Ken 01:19:32 Yeah. I don’t know. And it’s just a question for the AI experts and I guess, yeah, the field is fun, cuz lots of people can try different things and you’ll find out what works sort of that way by, by people playing around and, and seeing, seeing if their potential, uh, benefits from doing such things. So, so it’s hard to say, I, from my perspective, I couldn’t tell what’s, what’s gonna work better in these different circumstances, but because I’m more interested in, well, let’s get something that does what the human brain does as a vehicle for understanding ourselves and our brains better, which is not the goal of AI <laugh> but it’s
Paul 01:20:07 Goal a small portion.
Ken 01:20:09 Yeah. It’s Alli there cuz it’s like, it’s, it’s gonna help us understand ourselves. And it, and then it’s part of the science of, of, uh, understanding the brain.
Paul 01:20:18 Speaking of playing around, I, I wanna close on this because, uh, I, I opened with your interest in consciousness and how that kind of set you, you know, along your path. And now how you, I don’t know if you’ve really come back because to the study of consciousness, because you never really left it in the sense that you were studying memory and declarative memories and uh, conscious types of memories. But I’m wondering if you have advice to people based on your own experience, you know, about how to go about, you know, for example, you were given advice to go to neuroscience rather than psychology, right. To study consciousness and you thought that that was good advice, but in some sense, you’ve kind of, you’ve kind of come full circle. And I don’t know if you’re, I don’t know if this has been rewarding, this latest like Luci dreaming and consciousness and dreams, if that’s been rewarding based on your early interest and consciousness, but how do you see your own path? Um, you know, not, not necessarily that these things happen intentionally, right? I think of my own path as a big scribble of, uh, unintentional things where you have to make decisions where to go next. But do you have advice for the, you know, people who are interested in getting into neuroscience or psychology based on their interest on consciousness or loose to dreaming, for example, is there a right way to do it
Ken 01:21:31 Well to your question about whether it’s rewarding? Yes, it’s, it’s great fun. I feel very, uh, fortunate to be able to play around in science and get to come up with ideas and, and do different research to have that flexibility, to have my own lab where I can just get an idea and go off and do something it’s great. And there’s so many more things to study that are interesting. Uh, and I guess I would advise my students and others usually that you find something you’re passionate about and dig into that passion and see where it goes. And you guess you need some building blocks. So you, you basically, you have limited opportunities as student. You don’t have your own lab, so you can’t do whatever you want. You find some other place to work where you can gain the skills that you wanna have so that eventually you can use them to all the other things you might be interested in.
Ken 01:22:22 For example, one of my current postdocs worked heavily in cognitive neuroscience to study working memory and just some basic memory stuff because he wanted to then have those skills. So he’d go off and study lucid dreaming and that’s, that’s great. He’s got this whole Armar of, of, you know, abilities now. And he had to put in some work and study, you know, uh, all these basic things to have the tool tools to then go on and do these other things. So a it’s great if you have a passion for a particular thing, but sometimes you have to put that on hold as you, you know, gather the expertise that you have that takes years to gather as you know, how do you do science? How do you play the whole game and do all the different parts of science to be able to get to the point of, of having the luxury of then going off and running with new ideas.
Ken 01:23:13 And I think, you know, academia is great in that sense that it gives a lot of people, this ability to play it and go, go try different things, but you, you have to figure out well, how to make it work. If you need to get funding, how do you get funding that allow you to do the things you want to do? So you have to convince other people that it’s worthwhile doing and figure out how to, how to do that successfully. So that, that takes a lot of time to work within the system. And it’d be, it’d be fun if you could just open your own, you know, I guess if you’re incredibly wealthy, you could just go in your bay and do all kinds of stuff like that without, you know, that would be the old fashioned, uh, way that these things were done.
Ken 01:23:52 Uh, but these days, you know, you, you could work as part of these teams. And I think, you know, even in industry, there are lots of exciting things that are going on. And if you can find the right people to work with in industry, you can do some wonderful of things and develop your ideas and, and new technology, new directions, new applications. So there, there’s all sorts of, uh, different ways to, to, uh, to envision how this is, how this could work for you. But, but I think, you know, just understanding science in school, it’s, it’s so different when you learn science in school, you do just learning, well, what are the facts that people already gathered and a little bit about how they got ’em, but it really, how does it work now? And you don’t get to do this other exciting thing of, we don’t actually know something.
Ken 01:24:38 What’s a question I have that I want to get the answer to. And how do you design an experiment? That’s actually gonna move you ahead and understanding this thing and kind of try, try to solve a mystery that hasn’t been solved yet. And that’s, you know, that’s different than reading textbooks about stuff it’s going off on some adventure, uh, story and trying to, to get some new pieces to put it together. And that’s, that’s, what’s really so, so much fun and, and, and great to do collaboratively with a bunch of people, cuz you bring in different ideas and move things forward in, in, in exciting ways where you, you can’t actually see where it’s gonna go and it has to develop on it on its own. So it’s wonderful for us to be training more scientists, to do all these sorts of things that are gonna be the future ideas.
Ken 01:25:23 I remember, I think when I was young, I didn’t know I wanted to be a scientist, but I kind of thought maybe being an adventure would be fun, uh, invent stuff. And then I, and then I thought, well, but all the cool things have already been invented. Of course, I guess maybe that’s not a good idea after <laugh>. And so, uh, I think science is so open ended cuz we can, we can kind of see the more we learn, the more new questions come up and we can kind of say all the, there are all these things to have yet to be figured out that the students of the future will get to tackle in. And I think, um, these explorations of memory research and how it intersects sleep research and this, all these, uh, progressions are opening up new opportunities that we hadn’t envisioned before about research directions.
Ken 01:26:08 And I guess that’s always the way it works, that there’s more and more openings coming up that, that, uh, gets you excited about the future. And then of course we, we also wanna say, and how is this useful? So it’s, it’s great to do science for science sake and just, just explore, getting new knowledge. But then it’s nice to also have this other dimension of well, can this be useful to help people, uh, that have certain disorders or can it be useful to help people develop empathy? If you’re trying to work to be a more compassionate person, how do you, how do you build that to your learning and how does memory consolidation make that work better? We’re exploring some ideas like that, that I think are, you know, gonna be more important and, and can connect to, you know, what’s happening in the world and how do we make the society work better?
Ken 01:26:58 And you know, how do we, how do we help deal with global warming? You know, oh, problems. Cause we have to think, well, how do people make decisions and, and how come we’re so shortsighted and selfish and how do, how do we get past that to, to do better because the world needs that. And so, you know, it’s, it’s fun to do science and make little progress and it’s even wonder we’re wonderful if we can make that use for, for our society and, and you see that happening in so many different ways. Uh, so I think that makes us be optimistic about the future that we can think of more, more and more ways that, uh, our work can be useful.
Paul 01:27:38 Well, Ken continued success and luck in your future future. Not that you need it, but, um, and, and thanks for taking the time to speak to me. This is, you know, I have about as always, I have about a thousand more questions to ask you, but, um, but we covered a lot of ground and I really appreciate the time you spent here. So thanks.
Ken 01:27:54 Oh, it’s great to be with you. And I, I appreciated getting to chat. A lot of fun.
Paul 01:28:03 Brain inspired is a production of, and you, and don’t do advertisements. You can support the show through Patreon for a trifling amount and get access to the full versions of all the episodes. Plus bonus episodes that focus more on the cultural side, but still have science go to brand inspired.co and find the red Patreon button there to get in touch with me, email Paul brain inspired.co. The music you hear is by the new year, find email@example.com. Thank you for your support. See you next time.
0:00 – Intro
2:48 – Background and types of memory
14:44 -Consciousness and memory
23:32 – Phases and sleep and wakefulness
28:19 – Sleep, memory, and learning
33:50 – Targeted memory reactivation
48:34 – Problem solving during sleep
51:50 – 2-way communication with lucid dreamers
1:01:43 – Confounds to the paradigm
1:04:50 – Limitations and future studies
1:09:35 – Lucid dreaming app
1:13:47 – How sleep can inform AI
1:20:18 – Advice for students