 Boom, what's up everyone? Welcome to Simulation. I'm your host Alan Sakyan. We are still in Cambridge in Massachusetts We are now going to be talking about the neural circuitry of emotion. We have Dr. Johann John joining us on the show. Hello. How's it going? Thank you so much for coming on greatly appreciate it. Pleasure to be here I'm super excited. Super grateful to Alex K. Chen for introducing us. This conversation is going to be lit I'm very excited Johann's background. He's a research scientist at Boston University focused on the neurobiology of cognitive emotional interaction. He develops computational models that investigate emotion related disorders and you can find the links below to Neurologism.com as well as Quora.com his profile there and on his Twitter profile as well and you've been very active on Quora. Your research is very interesting. I'm super pumped to talk about the amygdala interacting with other brain regions, computational psychiatry disorders. We're going to be jumping into all of that. Let's start things off Johann with what we love asking our guests about their perspective on the current state of humanity. So yeah this is a hard question but at the same time I guess the first answer I would give is probably what a lot of people have said to you. There's a mix. It's almost like we live in a mix of a dystopia and a utopia. When you look at the technology it seems like we're in some kind of Star Trek future already but at the same time when you read the news if you're punishing yourself by reading the news then it does seem quite dystopic so I guess the term is heterotopia. The future is here it's just not distributed evenly so that is kind of an opportunity like there's a sense in which things are going very badly in many parts of the world and also sort of on the inside there's a lot of anxiety and depression and you know worry of various sorts. So you could even say that the world is in a state of anxiety and based on maybe some research that we can talk about you could say that anxiety at least some versions of it involve being aware of problems and aware of possible actions that could solve them but not decisively taking action on any of them because you're never forced to make a decision. Humanity seems to be in this state. There's looming problems but do we have to decide now? Can we wait? Global warming being the cartoon example people are like well maybe we can let the next generation deal with it seems to be the answer of a lot of people and I think across the board there's this like there's all these opportunities to do things but should we do them or just have fun. Wow what a powerful synthesis yeah the future is not evenly distributed but it's here there's this utopic technological haven that we live in at the same time a dystopic so many looming problems that we're kind of pushing off to see how long we can keep it rolling until we have to actually problem take care of it. I like that idea of the earth being in a state of like kind of like anxiety or in a sense it's also we have a lot of transgenerational traumas that have built up as well that we need to you know be honest about have discourse about reparations were needed integrate that trauma into us and move forward and in many ways we haven't done that for so many things so that was that was very well synthesized and okay let's do the journey of how Johan John became who he is today you're born in India yes I was and then you ended up coming to Boston and pursuing your PhDs here so tell us about this where you're born and this transition to Boston. So I grew up in India but I also grew up in the US so my parents were in in the US right before I was born they moved here in 1980 I was born in 83 and my mother was back in India when I was born and then for the first nine years of my life I was in the US first in New Jersey and then in Columbia South Carolina and then my family moved back to India so I was there until 2006 so in college I studied physics and I also did a masters in physics and at the time I hadn't really heard of neuroscience I didn't know that that was a word I knew about neurology and I figured that that was a medical field and those days neuroscience wasn't in the newspapers like it is now every day almost and during my masters program I was looking at a list of possible summer projects because the standard thing to do was do a summer project get recommendation letters get a little research experience and I always liked using mathematics to solve problems but I didn't have a strong passion within physics I liked just using the tools my favorite field for a while was just mathematical physics and so I was looking at this long list and there were these all these topics they all sounded okay and then there was a topic neural networks and I didn't know physicists studied the brain so I asked a professor that I was close to what is and he says oh it's a really interesting topic physicists do a lot in there and he had a friend in Madras Chennai who was studying neural networks so he said do a summer project with him and it was a very interesting summer project I mostly just read papers and I did a little bit of coding but I was quite inspired by this field it seemed to tie so many things together I could use math and coding which I just like doing and you get to study humans and animals and also you know deal with philosophical questions without having to get bogged down so I thought well this is the best thing I could possibly do and I don't have to compete with any of my classmates for applications to go to school because no one else was doing your science those days so so yeah I applied I I'm just by fluke I met someone in my pro who is taking a course with me who heard I was applying for grad school in computational neuroscience and he said oh there's this department in Boston University which I hadn't it wasn't considering at the time cognitive and neural systems so I and he said I'll be perfect for you and it worked out really well I applied there and got in and it was a very unfortunately the department doesn't exist anymore but it was a unique department it was a standalone department devoted to neural modeling so it is a building full of theoreticians and modelers and drawn from all kinds of backgrounds there were people from physics and math computer science engineering biology psychology there was one student who had a theology background so it was great interdisciplinary work it can be frustrating but it's also sometimes the friction produces to more heat than light but when it produces light it's great so yeah so I started started in 2006 here at BU and my thesis was on timed behavior so how humans and animals can anticipate something that's going to happen in a minute or two minutes and sort of time their behavior accordingly and that involves and also how that kind of behavior can be modulated by drugs that influence dopamine and acetylcholine and you can model what's happening in the mind and the brain as they're predicting that in a minute something's gonna happen in the case it's a rat saying oh well in a minute the lever will give me food so there's no other queue in the environment so the rat has to produce an internal representation of the time what's going on in the rats mind is you know a philosophical question as I said but we can study the behavior and and look at which brain areas are involved in which ones this is so there's a region below the cortex called the basal ganglia which is a set of regions subcortical regions that's very involved in a lot of voluntary behavior and and learning and so the basal ganglia circuit seems to be crucial and also cortical input to that area so and dopamine can kind of speed up and slow down an internal clock in one way and acetylcholine drugs can do a similar thing in a slightly different way so my model that I developed over the course of my thesis looked into how that kind of modulation can happen interesting so was the neuro modulation potentially had an uptick as the reward was about to come and then it so there's that the there's definitely so you may have heard about how dopamine is involved in reward prediction and reward prediction errors so less known is that dopamine also has a kind of invigorating effect on just motor behavior you need a little bit of dopamine in order to do initiate action and that's why when certain dopamine cells are damaged Parkinson's disease arises which is often a difficulty to initiate and sustain movements so it seems as though this motor effective dopamine is why you have this speed up and slow down so basically dopamine agonists like amphetamine or methamphetamine or cocaine can temporarily speed up the internal clock so it's called speed for a reason and even for rats it's it's speed and so what you see is that temporarily the rats will respond as if time is going faster subjectively so they'll respond early but over time if they're kind of running multiple trials of this you know one minute later a reward arrives they learn to correct so they compensate for that shift and what's really fascinating is then if you take the animal off the drug you see a bounce back effect in the opposite direction so their clock slows down again temporarily and they compensate and get back to normal with the acetylcholine drugs you have a complementary effect you apply the drug initially nothing happens and then slowly over multiple trials you see a permanent shift in the clock time so there's a sort of stable delay which doesn't get compensated for and again it works in both directions slowing and speeding depending on the kind of drug so that's an interesting phenomenon call really calls for a model so that was you're doing a lot of modeling and when this is also I just wanted to also give a big you know shout-out to these these mentors or role models or that we have in our lives that kind of nudge us in a certain direction that basically open up so many doors and so much of our in that becomes our life interests thanks to the little nudges of yeah there is neuroscience go and check it out for a summer and then all of a sudden boom you you know explode and into it for yeah decade just the few months before I decided to take up neuroscience maybe my dad attended a talk by VS Ramachandran in India and so his his brother-in-law is a neurosurgeon and invited him why don't you come along for this talk and my dad was like what do I know but he went and and he was blown away by it and he was asked he asked me a first-year student in a master's program in physics is there any way that you could get involved in studying the brain and I was like no I don't think so it seems pretty far from my my world and there I am a few months later in a neuroscience project and reading VS Ramachandran's book phantoms in the brain okay alright so then that takes us to your thesis and PhD work and then that ended around 2011 2011 okay and then okay so now I want to hit on the the sort of interest in you know the neural systems lab and emotion and cognition and how what the other amygdala interfaces with the other brain region so take us down this this path so I work as a computational model of in a neuro anatomy lab and what we do is we my colleagues look at the connections between different brain areas in the recess macaque brain so so it's like a nice intermediary between a lot of studies involved rodents mice and rats and they're quite far away from humans and then you have human studies which are very useful but kind of much more fuzzy in resolution so studying in monkeys is important to kind of understand the structure in detail for a brain that's closer to a human brain and over the past few years with my lab's been focusing more and more on the limbic parts of the prefrontal cortex and the subcortical areas that interface with those areas so a lot of people think of the prefrontal cortex is this one monolithic new part of the brain but it's not really it's it's of quite of differentiated part of the brain so you have near the middle of the of the brain and on on the underside you have what might be described as simpler and potentially more ancient brain structures which have less well-defined lamina and then you have the other side for me lamina layers oh lamina layers yeah okay so you've heard that the brain is six layers the cortex but it's not exactly true you see simple brain areas where you can really only distinguish three or four layers and then you have classic neocortex with six but in v1 for instance you would almost delineate seven layers but so you have this gradation of structure where the areas like the DLPFC the dorsolateral prefrontal which is involved in a lot of higher cognitive tasks and working memory has very sharply defined layers whereas these simpler areas which interface with the the emotion related areas don't have that sort of sharp lamina structure not as sharp and so my lab started looking into the connections with the amygdala and now we're looking at the hippocampal connections and we're very interested in what that has to do with emotional disorders psychiatric disorders for instance we were looking into PTSD schizophrenia I recently published something on that and right now we're looking into the depression which also seems to involve these areas and okay so right before we get into the depression and schizophrenia PTSD which we're very excited to get in there so this is a neural anatomy lab yeah and so you you started kind of pointing out how different brain regions have different anatomies and so you're modeling those anatomies yeah and then you're building out this comprehensive model of neural anatomy and then also you guys couple that with technologies like EEG and fMRI and and monitor the way that neural activity works in with your neural anatomy so my lab doesn't do recording so what I try to do as kind of the resident theoretician is to tie some of the structural data with the findings from labs to do behavioral assessments or electrophysiology like recording from individual cells or from like EEG is kind of a more fuzzy measure so so one big task for computational people is to put together all these disparate pieces the puzzle pieces of of neuroscience and and try to tie them together it's still early days so I always say my models are toy models they're much more simplified than even what we know about the brain but in order to make things tractable you kind of have to simplify okay yeah correct definitely and that's as a synthesis I'm doing that all the time exactly yeah so so then you are yours you're synthesizing what a lot of the neural activity data is showing from the electrophysiology data that you're getting and then you're making also comparisons of that data to what you're seeing in behavior exactly okay cool cool all right all right so so walk us through if this was cool you know we've talked with you know Dr. Adam Gazzelli and other neuroscientists about top-down processing versus bottom-up processing which is a fascinating way of thinking about the way that we have our cognition occurring but I like the way that you said we want to have an amygdala that's working for us so a coordination versus a suppression excellent okay so yeah teach us about this so I can I can introduce it by talking about a model that we published a couple of years ago called the emotional gatekeeper so my lab a few years before I joined the lab discovered a connection of the amygdala that wasn't known in any species between the amygdala and a region called the thalamic reticular nucleus the TRN so deep in the middle of the brain there's a structure called a thalamus and sitting on top of it there's this inhibitory a thin layer of inhibitory neurons that's been described as an inhibitory veil it inhibits thalamus it receives excitation from thalamus and cortex and it regulates the input from thalamus going to cortex and Francis Crick for example was very interested in this structure and he described it as potentially an attentional search light so it seems to be a potent mechanism for filtering and the information that reaches cortex so what might a connection from the amygdala be doing our model tied together some data suggesting that this could be a way for emotionally salient important information both positive emotion related and negative to seize attention so it's a way for the for emotions to get access to to be a gate keeper so if the TRN is like a gate you want a bunch of different gatekeepers you know passport please papers please so you want important things to get through and reach the brain and because there's so much information in the environment you can't just let everything through as far as we know the brain really likes to keep to a minimum what is available to the decision-related areas in the brain so we can say right now we're taking in you know so much visual stimuli so much auditory stimuli even taste and touch stimuli exactly we're only because we're in conversation together we're really hyper focusing what we're exactly our conversation and but at the same time you want to be if someone screams for help we should probably interrupt the interview and go help them and and so a structure like so this projection from the amygdala can do could mediate that type of of interaction so that would be a thought what what in the model we call bottom-up emotional filter filtering so even if you're not expecting a particular emotional stimulus you need to be able to attend to them at the same time there's connections from the amygdala to cortex and cortex back to amygdala so here if you lift let's say you have a strong plan you could use this amygdala cortex amygdala trn projection to lock out irrelevant stimuli even if they are emotionally important so let's say you're really focusing on something important and you can't be too bothered by the dog barking or something like that like the elevator right now if you have a strong emotional association with elevators maybe so so so so that's an example of a top-down effect that it could be useful in some situations we're like quelling that exactly so even though it's emotionally arousing and is exciting the amygdala we're using the amygdala to suppress other competing emotional stimuli but it's the thalamus that's this that's this part of the decision as well well there's a lot of debate about what exactly the thalamus is doing it has for years been described as a relay which is kind of you know underselling what it does but we still don't have a good story but it's it's sort of like the entry point of a lot of information both from the outside world and from cortical processing itself to propagate to the cortex so this is a way almost like the reception center the the amygdala is one of the receptionists potentially allowing things in to the cortex okay okay and and then in a classic model of wanting to have better coordination with the prefrontal infrastructure with an amygdala something like a meditation as we were talking about before the interview that it almost really does feel like processes like meditation give one a greater coordination with the amygdala it feels like it feels like we regain control over what is normally just draws us to a reaction immediately or rather we can more easily slow down breathe cognize and move from there so so can you just give us a bit on what is that a is that a is that a conversation that's happening between amygdala and prefrontal is that a conversation that's happening between them that's a good way to put it the so the amygdala has bi-directional connections with a couple of different limbic prefrontal areas and so and there seems to be the potential for a kind of push pull in the sense that the amygdala can excite the cortex but the cortex and the cortex can excite the amygdala but the cortex can also inhibit so there's two classes of cells excitatory and inhibitory so the cortex so different courtesies have different relationships with the inhibitory neurons in the amygdala and this can be a powerful way to regulate how strongly the amygdala produces its signals for instance to the TR and elsewhere so a lot of the fear related reactions that we have like in rodents there's a freezing reaction just stop moving but humans don't always do that but but a lot of those reactions are generated by the amygdala going downstream to other parts of the of the of the nervous system and regulating when that happens is a good idea so you need the amygdala to talk filter salient things both good and bad dangerous and safe but at the same time you don't always want to immediately react to them so when you talk about coordination one thing that seems to be useful is use the amygdala to for the information involved in what's important and what's not in the environment but don't necessarily react immediately so you want the best of both worlds you want that filtering signal you want that alarm signal if you like but you don't want the knee-jerk reaction and here's where a combination of excitation and inhibition the right balance seems to be important and it can go both ways excitation and inhibition from yeah frontal cortex to amygdala yeah it's not always easy to map these directly onto this circuit but you could like for instance you have a reaction to something like pain or something so that's coming bottom up but you could also have a reaction to a sad story and and and that could come down to the amygdala and produce reactions that feel a lot like physical pain to some people your chest could start hurting breathing will get altered so we it's well known that for humans a lot of emotional pain resembles physical pain and at the same time let's say you're very angry let's say someone cuts you off in traffic and you're kind of angry if for instance you find out that there was a pregnant woman in the back and they were rushing to the hospital a lot of people their anger will dissipate fairly quickly so that there's an example of you know cognition and emotion having that kind of more damping down effective fairly rapidly in some cases yes yes so many interesting examples about the way that the amygdala interplays communicates with other brain regions and okay so then that's on a lot of the the amygdala side of things now I want to hear on the computational psychiatry side of things on the hippocampal side of things on the you recently published on schizophrenia in December and also you're pursuing aspects of depression and disorders related to that so let's venture into this excellent so computational psychiatry is a newish subfield of neuroscience and similar to the story with neuroscience I hadn't really heard of it for a while and then in fact someone contacted me because of quora to ask me what computational neuroscience was and then I was like I don't really know what that is but I think I could guess and and so I met this this person and talked about it and then because I'm studying the limbic system the amygdala and all these emotional related areas it was a natural outgrowth of that to start talking about disorders in these models so it's not just about successful filtering but also well what what can go wrong there's many ways the things that could get can go wrong and that allows us to not just ask what are the correlates like which area is involved but why so so so in the case of schizophrenia it's again this circuit involving the TR and this thin layer of inhibit inhibitory neurons and the thalamus and the cortex so we will just focusing on what else we could say about this three area circuit it's a it's a generic circuit so for different systems you have the same kind of motif we call it a circuit motif and so schizophrenia is a very complicated disorder and no one has really found a definitive cause or even a definitive correlate but in recent years one line of research experimentalists have uncovered is dysfunctional inhibition so there's local inhibition in the cortex and tamping things down and there's the TR in itself which has been linked with schizophrenia so so we thought let's look at this circuit and see what it can do and what can go wrong so when you're making a computational model you need to fix a behavior that is tractable like modeling hallucinations would be cool but it's very difficult because you know we really don't understand how language is generated so getting to that is is too complicated so what we found I I was quite intrigued to learn that a few years ago that one of the most readily identifiable symptoms of schizophrenia is disordered eye movements so schizophrenia patients and sometimes their close relatives have an inability to fix their eyes on a target and for instance you could have a dot on a screen that's moving around and and healthy controls they can follow along that dot just fine like going in a figure of eight for instance whereas with a schizophrenia patient their their eyes will often be quite erratic and there was a paper showing that a machine learning method involving a couple of eye movement tasks just using those could tell schizophrenia patients and their close relatives from controls with something like above 95% accuracy so it's a diagnostic symptom a very useful for you know at-risk populations and for the patients themselves and so the question is well what's producing this this symptom and might that shed light on other symptoms so I set up a circuit involving a simplified version of the the three brain regions cortex TRN polymers and I made a simple eye tracking model that that uses the this kind of circuit to maintain a target at the center of attention and what we saw is that these disrupting a couple of different classes of inhibition could produce this erratic behavior and so so so that's something that experimentalist can now maybe work with we showed that different interrupt disruptions could look vaguely similar but also have differences so this ties into the idea that schizophrenia and potentially other disorders may not just be one disorder they may be a family of disorders with different causes that seem outwardly similar and I think this is very important going forward for the field and I think the field is more and more acknowledging this so apart from that diversity of causes story one idea that we elaborate in the discussion section is is the idea that going off track is maybe a feature of other symptoms of schizophrenia so one symptom is having disordered thoughts so it may be that if you can't track a dot some if that similar disruption is going on elsewhere it's you're not able to follow the context relevant set of thoughts so the same disruption going on in an analogous circuit might prevent someone from holding on to the line of thinking and sort of jumping from one thing to the other so it's an interesting way of looking at how a simple local circuit disruption can have these wide spread effects in different systems okay this example was now I'm thinking my imagination is roaring about all these other potential diagnostic methods that don't yet exist that we're good that we're going to be able to figure out and leverage machine learning to be able to more effectively figure it out right now and then also when you're when you're explaining this potential that that something as complex as a schizophrenia could potentially be a myriad of different involvement in different brain regions is a very complicated and this is all the way to when humans even in the womb and then they're born and like there's you know foundational infrastructures that get laid out at the during that time period and how that interplays with their adolescent and adult life then okay I have I have a question pre as we get into depression I want to ask about schizophrenia I've had several conversations with people about how we may all be the spectrum of schizophrenia potentially from a good schizophrenia to a bad potentially schizophrenia and how in what what we've created with hyper realities where it like a Santa Claus it's kind of like a hyper reality we've made up a story okay and we and we call it real and then we tell some people tell their kids that that's real yeah there's cookies and stuff and sliding down the chimney with presents and all this kind of stuff reindeer etc and and then it almost gets you know in a sense there's these two competing realities right are then occurring and so for and and some people even say one of the most potentially one of the greatest measurements of intelligence is one's ability to abstractly reason multiple perspectives multiple variables on a given issue and so when you're doing that what you're doing is you're trying to see how like a low socioeconomic status person sees the world how a rich person sees the world how someone from India how someone from Pakistan how someone from different countries also different genders different religions how people see the world right so if you're trying to balance all these perspectives and see them all so do you see kind of where I'm coming that in this term and how do you feel about all of that so I just started reading a book sort of from left field compared to the neuroscience side it's called the sublime object of psychiatry it's a kind of a more of humanity's approach to schizophrenia so it describes schizophrenia as a sublime object and what they mean by that is it schizophrenia has been defined by various fields including biological psychiatry and psychoanalysis as sort of beyond understanding meaning that you could maybe study it but you could never empathize with with the schizophrenia patient which isn't it's a very strange book but very interesting because I have and you know people can relate to this they're like yeah the person when the schizophrenia patient is describing their hallucinations or their overarching world theory you can kind of make sense of parts of it but the whole thing you just very hard to understand what's going on with that person so if what if I understood what you're saying correctly perhaps some of these people are kind of torn between so many perspectives that they they end up with this mosaic idea that that is a product of perspectives that don't even sit well together it's a fun speculation but you know when you're talking about disorders that you have to remember there's a lot of suffering involved in this and so you wouldn't wish this on anyone and especially the psychosis side of things in catatonia and then there's also a way to as one potentially experiences a psychosis from this mosaic of worldviews let's say there is a way to live with that in a gentle way in a way that creates almost a deeper degree of empathy for for different worldviews in that sense and so although it can at times feel like it's a big taking on the burden to be able to relate to all these different worldviews at the same time it it makes one super well-rounded and in that sense but so it's a good another way to view this is just the way that we use the social media culture because when you take a look at people's profiles it's it's it's only almost exclusively their best selves and so then that becomes you know what's real in this sense what is actually you because is you really only your photos of you looking your absolute best traveling having the best meals the best people is it real yeah so because you're definitely not posting when you feel anxiety or when you feel depressed or when you you know you're feeling a lack of love or you know so so here's again this this this idea of what is reality and how can how do hyper realities how do balancing different worldviews into a large mosaic how do these things relate to schizophrenia it's a fascinating question if we're in a simulation or not these are very very hard questions but what you said about empathy I think can help maybe ground this in something so I read a few years ago a fascinating paper on schizophrenia in different cultures so I believe it was called hearing voices in Africa in Ghana and China there was an African city and an Indian city excuse me and the I discovered it through a blog called neuro anthropology which so the point of the paper was that it's widely believed that schizophrenia is a fairly universal disorder affecting around 1% of people worldwide but the specific manifestations of it can be quite different so meaning the content of the hallucinations is different so in different parts of the world like for instance apparently in the West and America a lot of the voices are potentially giving sort of violent instructions of some sort whereas apparently in India some the voices sometimes are a motherly voice telling you to clean up your room so so there's that difference then there's also the the way in which society responds to the person having hallucinations so the paper talked about how in certain tribes they frame this having of hallucinatory voices in terms of possession so a spirit is possessing you and interestingly you might think wow that's unscientific it's nonsense but it this may actually help reintegrate people into society because when the person is not hearing voices are acting weird their belief is well they can come join work because the spirit isn't with them today and allowing people to be in society and be a productive member of society is really important and one of the things you you read for instance in that book I mentioned earlier is that schizophrenia sometimes defined in terms of inability to contribute to useful work so there's a capitalist element to who we decide is like incapable of being useful so the way the society responds in one sense because if you give them a big paint kit yeah yeah yeah you know I often wonder if you know in you know and this is a popular idea that in earlier generations people who are prophets or maybe shaman might have been what we would now call schizophrenia patients so so how the culture responds is important and also how they tell people so for a long time in America there was an instruction to patients to ignore the hallucinations whereas what you said about empathy is very relevant because some places they instruct patient patients to listen to their not act on everything that the voice tells them to do if they're telling to do things but you could in the spirit of empathy towards people outside these constructed people that may exist in you are in a sense perspectives they might not be correct or accurate yes but as with most people people just want to be heard and and given some kind of consideration you don't have to list act on every piece of advice you receive but giving them the time to be heard is a great way to diffuse frustration and tension rather than ignoring people yeah so so so empathy towards one self may be important too yeah or one's selves yeah yeah yeah yeah yeah this is this has been such a fascinating back and forth on on schizophrenia and I really am so interested in it okay now let's get into the aspects of depression and disorders with that mm-hmm yes so I've just recently started looking into this as a result of so my lab has been sort of expanding out towards this understanding of the limbic circuit we started with these medial and orbital prefrontal areas and the amygdala and now we're looking at the ventral hippocampus which is a very emotion related area and the fourth region so I like to think of the limbic system in terms of the four pillars of it being medial prefrontal areas the amygdala the ventral hippocampus and the nucleus accumbens which is one of the areas in the basal ganglia which is you know strongly involved in reward based learning and things like that so with these four areas you kind of have a motivation circuit you have emotional salience you have top-down plans and medial prefrontal is complicated but it seems to be involved in things like detecting conflict and error and and sort of notifying the rest of the brain that something's going wrong the hippocampus seems to be involved as everyone knows with memory and also with context and navigation so these areas seem to interact they have strong connections with each other to kind of help an organism decide on a course of action in the case of the hippocampus this seems to involve comparing deciding what context you're in is it a safe context is it a context where it's worthwhile to seek rewards invest time to invest time in something or is it a time to be vigilant and so what we're looking into is the idea that at least some forms of depression involve abnormal context processing so it's the four pillars have an abnormal communication yes and and so the areas that one for can focus on are well any so because these are complicated networks any a disruption in pretty much any node could cause a problem and and outwardly you might not know which node has been disrupted so which is why we we can think about a spectrum of depression like behaviors they may not always arise from the same disruption but so when you're looking at the hippocampus a lot there's a lot of data on the hippocampus being involved in depression but the question is why so so this is what I'm working on currently so you know ask me in a year what if we will what the model is but the shape that I'm trying to take it is that memory of the past and anticipation of the future seem to involve the same region hippocampus and closely related areas and you might ask why so one way you could understand the purpose of memory is to anticipate the future yeah so you know it's great to have memories just to reflect on remember that time but but if we're being good biologists we have to look at everything from the perspective of evolution yes so now what does that mean that means that at least in principle once upon a time traits related to memory had a purpose and so selecting comparing the present to the past in order to choose a course for the future is a way to link all these things together now what does that have to do with depression yeah it seems as though depression is there's many forms of some of them involve demotivation so the it's not so much about sadness all the sadness can be involved as it is an unwillingness to take action so there may be something good just on the horizon but they don't seem to be able to get out of bed to act on that they may be cognitively aware of the opportunity but it doesn't land with the drive to act on it so one possibility that we're thinking about is that and it's not just me a lot of researchers have looked into this but one possibility is that there's an over generalization of a bad context one in which there aren't opportunities like maybe you had a couple of bad days people who are optimistic wake up in the morning and today is not going to be like yesterday even if they've had 10 days in a row of bad days and that's a great form of resilience to say that today will be different from yesterday and this may be disrupted for various reasons it could be prior trauma and one big factor long-term stress unavoidable stress is the worst thing in in terms of human health and the basic physiological needs not being met is one of the worst ways to induce stress yes and you know but you know what's worse social stress how humans treat each other is the number one stress so yeah so be nice to people because because that so Robert Sapolsky talks about he has this awesome data about baboons and how there's a particular tribe in in Kenya that they have so much access to basic physiological needs like food and water that they have plenty of time to be cruel to each other and and and the baboons at the bottom of the pyramid the male baboons have the worst stress related factors and that affects things like propensity to put on weight and all kinds of factors like that whereas the animals near the top are much better in terms of stress so stress can be one way that a system for instance of memory and context becomes biased towards negative context so there needs to be some sort of a of a self-actualization and a self-transcendence some sort of a meaning and purpose and fulfillment past basic physiological needs and usually that means in a social setting that you have these powerful social connections and that you have a goal that you're doing every day and that you're these four pillars are able to talk to each other well about getting you seeing that opportunity and then seizing it and going for yeah so at least some forms of depression involve a difficulty in either recognizing or enabling a good context to drive behavior so it's it's a double tragedy because it's it's not that they don't believe that there are goals out there they might have goals but it's like some sort of access problem like the goal isn't reaching you know is motivating so how to treat it is very complicated and so one of the reasons that we need to develop a model and is to understand all these different antidepressant treatments why do they sometimes work with some groups and not others and why is there still this treatment resistant depression population so once we have a circuit model we can look at different kinds of defects and see well which treatments might compensate for which defects and then we in future we might even be able to target things much better if we can find behavioral clues for well this is depression type A or depression type B that will be ideal I can't promise anything like that anytime soon but that's the goal of theoretical neuroscience fascinating yeah one of the yes yeah totally if you can maybe you know find a power law distribution in depression and identify that a lot of cases follow a type a let's say or yeah yeah yeah and type B is less and it's like it looks like this etc yeah actually what you're teaching about about memory the use of memory is something that we find very fascinating in that you know and the way that you know what the sleep's involvement in that process as well and just if one's ability to look at the stimuli that they've collected throughout their lives as the as their memory and then how a computer carries almost a hundred percent retention of memory whereas we carry whatever way less of a stimuli that we've absorbed right so like a robotic intelligent perception could carry almost infinite if it's being you know stored and and it's so crazy how we then compute the days stimuli with our storage of memory and then figure out what the best future actions are to take that whole that feedback system that we have is so freaking crazy yes crazy is but yeah and you know it's good that you mentioned the computer because it's very popular to understand human memory in terms of computer memory but it's also very important to understand why they are wildly different like for instance see in a hard drive as you pack it with memory eventually it starts to do worse like the more memories it holds but whereas with human memory the more links you have the better your memory is because you can make more connections so associative memory is very different from computer memory associative memory yeah be like if the memory of the computer was learning exactly why and so the the term that's used in computer science and machine learning is content addressable memory we haven't made particularly good content addressable memory and what that so it's what does that mean it's it's that if I talk about say a Beatles song I have all kinds of related information near the address of the Beatles whereas if I have a folder full of mp3s of the Beatles I can't be like well which are the songs that are about such-and-such topic or which which are the ones that are because unless I saved it with with tags or something in the first place that like have you ever tried to search on an unindexed file system it's so bad so you're talking about querying your own memory versus on a computer sending a queries it's much harder but but for the files that you've indexed locally but sending a query to a search engine is different wow because they're doing the connect associating yeah we're doing the associating like we're doing the work yeah yes yes wow wow wow and then it's it's almost as though every single then gmail address that is linked as it's making queries is then creating this massive yeah interplay between the one that's making the queries in the actual database but then also building out the massive model of what that what the queries have been happening and then predicting about what's going to be of interest and all that type of stuff but my favorite version of this is which is very mysterious and I have no idea how to how to model it is let's say you've heard a bunch of music and you have that in your memory somewhere and someone introduces you to a new term let's say a genre name you can retroactively say that oh that song I heard a month ago is in this genre so this new information can can immediately kind of go work back with prior memory so you can take new information and completely recontextualize reorder refile old memories on the fly with no effort yeah oh man yeah that's that's like when you when maybe you experience some profound awareness shift in a scientific field or something and you really go back and you and you realize that the way you are maybe storing the the knowledge before could be updated in such a cool way with the new information that it like makes it more easy to recall the wider of knowledge basis yeah wow that that was such a profoundly fascinating conversation I loved it awesome that was so cool on all things neural circuitry of emotion and also I want to ask you a couple quick questions on the way out of the show like we'd like to ask on every episode two of them okay first question is are we in a simulation I find this the thought somewhat meaningless in the sense that I don't think it matters because we can't step out of it so whatever we see is the real world for all practical purposes so it's fun to speculate about but I don't think it can ever have any consequences for our decisions or even for our experience really it's a bit like you know the new version of are all of us in the dream of Vishnu kind of thing you know like it's an it's just another version of that until we can actually poke with a scientific probe at the simulation and also make our own simulation so we get to you know do both experiments at the same time potentially the up one is a bit harder I don't know whether it's even possible we'll have to figure out nothing potentially it's impossible that's okay and then the last question is what is the most beautiful thing in the world it's a very hard question but I'm going to say the most beautiful thing I don't know whether it's beautiful or not but I love analogies metaphors and analogies and it's almost not just beautiful in itself but it's one of the ways by which we can access many forms of beauty and the more you study it the more beautiful it becomes in itself partly because of the mystery of it like show me the machine learning model that can produce and understand analogies then we'll have artificial intelligence yeah and as a synthesis metaphors and analogies are so crucial and if you can if we can refine something of sheer vast complexity into a very powerful metaphor like right now we've been working so much with this idea of a tree of possibility and the seed so that every mind of a child is a seed born into the world it needs to write nutrients in the root system the basic physiological needs in order for it to flourish and have lots of fruit for the community family and world and then also the tree of possibility that there's so many different roads that the seed can take in its life the jobs it takes the interests that it takes the people at Mary it said that you marry children that you have all this type of way to live so these analogies and metaphors they can drive expanded states of awareness in such a simple form that can get passed on for potentially centuries yeah so yeah I'm really happy that you said that's the first time we've heard that as the most beautiful thing in the world metaphors in the house you said that was the first time we heard that you'll hon thank you so much for coming on to the show thank you Alan this was great it was such a pleasure brother thank you thank you yes and I really hope that everyone that watched that you guys had a great time please let us know your thoughts in the comments below on the episode go and share this type of content with other people get talking to more people about our brain regions how we interact with the world what that means for ourselves and for our friends and communities and our what we what we seek as as our as our ultimate callings in life and also support the people that you believe in Johan's links are below support him follow him also simulations links are below support the artists and entrepreneurs and the organizations around the world that you believe in and go and build the future everyone manifest your dreams into the world thank you so much for tuning in and we will see you soon peace