Loading summary
Transcript1 lines
- [00:00]
A
Foreign. My name is Oliver Golson, and I'm a neuroscientist at the University of Victoria. And in my spare time, I'm that neuroscience guy. Welcome to the podcast. I want to start by saying thank you to all of you that are listening. We host the podcast through a service called Libsyn, and I logged in today. It's been a while, because I know that we've been inconsistent, and I felt bad about it, and. And I saw that we had over a million downloads of the podcast. I remember sitting in my living room, and it was literally when the first COVID lockdown started. And people have been telling me for a while I should do a podcast. You know, I heard it from Rick Mercer from the Mercer Report. I heard it from Bob McDonald from CBC's Quirks and Quarks. And they said, you know, you've got a good voice. You know the science stuff. You're good at explaining it. You should do a podcast. And it took Covid for it to happen because I was stuck in my house. I was bored. So I went online and I bought a Rode Podcaster Pro, and I bought a mic and some other stuff, and I recorded after many, many takes and double takes and retakes, the very first episode of that neuroscience guy. And when I did it and we put it up on Libsyn, I thought, you know, I'm going to listen to it, My mother's going to listen to it, my son and a couple of friends, and that'll be about it. And here we are a couple years later, four, I guess. And we have over a million downloads. And I just want to say thank you so much. It's. When I saw that today, it just reaffirmed that I love doing this and I love telling you about neuroscience, and especially the neuroscience of daily life. I think it's crucial to understand the world around us. So thank you. Now let's get on with it. Okay, so here's where we're at. I'm doing this deep dive into human decision making. We talked about the concept of value and expected value. Then we came up with this simple model of decision making where you always choose the highest value item. And on the last episode, we talked about the Explore exploit dilemma. So basically, this idea that you don't always choose the highest value option. Sometimes you choose a lower value option. And the reason you do that is because you're uncertain of the world or you want to, you know. You know, you want to make sure that you know values correctly because the world might have changed. So I'm just going to review this quickly. And basically, if you think of a flowchart through the brain, sort of what happens first when you encounter a decision point is what we call the representation phase. And what you do in the representation phase is you look at the set of feasible actions. So what choices do I have in this state? All right, and while you're doing that, you have to take into account the internal state of the body and the brain. You know, where am I at? You have to take into account the external state, like, where am I? And then you come up with those values. So that's the representation stage. And in fact, I've misspoken that a little bit. That's just where you figure out what's out there. The next phase is actually valuation, and I sort of always merge the two. But technically valuation comes after this, which is where you assign a value to each action based on the internal and external states. So you, you know, in the, in the representation phase, you're just saying, here are my choices. And most models separate the valuation phase, where you assign value. There's debate there, which is why I sort of said it a bit differently. But let's assume they're separate states. Representation, valuation. Then the next state is action selection. You choose an action based on the valuations or you choose to explore. All right, so there, that's the action selection phase. And then you have outcome evaluation. You basically say, given the action and the outcome, am I happy with what happened? And this is how decision making ties into learning, because from outcome evaluation, you have a learning phase which where you update the representations and the values and action selection process. Now, we're not going to talk about learning that much in this series, but we'll probably do a deep dive into learning in the near future. So that's where we're at right now. Representation, valuation, action selection, and outcome evaluation. Now on the next episode, I'm going to dive into the neuroanatomy of this and I'm going to put places on all of these four, four states and actually the fifth learning phase as well. But before we go into the neuroanatomy, specifically, I want to take one step further, because we know that decision making can't be captured just with this model. You need something a little bit further along. And as it turns out, there are multiple decision systems in the brain. And we've talked about this in the past, but it fits into this series. So I'm, I guess I'm revisiting it. And the most prominent theory with this is the System one, System two theory. System one is your gut hunch system where you make these very sort of rapid gut hunch responses. And system two is a more analytical system where you, you make a more rational thought out choice. If you haven't read it, Daniel Kahneman, he's now passed, but he's a Nobel laureate. He wrote a book called Thinking Fast, Thinking Slow. And it's basically about these two systems. Now, I will point this out, and I mentioned this when I spoke about this many episodes ago. There are two theories about this. One is that it is a dichotomous thing. You have System one and you have System two, and they're two separate things. That is a dual, what's called a dual process model of decision making. I am actually more of a fan of a continuous model of decision making. And I'll walk you through that at the end of this episode. But for now, let's go with the traditional System one versus System two model. So as I said, System one is your fast system. It's these gut hunch decisions. System two is your slow rational system. And in Kahneman's book, he gives these great examples and I'll use them here. You know, what's two plus two? And you go four. And he would say, that's a system one decision. And I'd say complete the phrase bread and, and you would say butter. And that's a system one decision decision. This won't work so well because this is an audio podcast. But when I teach this, I have a picture of two squares, one's big, one small. And I say which square is bigger? And you know instinctively which one is bigger. System one decision. But then if I say something like 13,678 divided by 13, most of us can't rip that off as a System one decision. So we call that System two. You're engaging this slow analytical process. You know, here's another one. The second highest mountain in the world is. Now if I had to said the highest, what's the highest mountain in the world? You would have said Mount Everest, most likely. System one, the second highest mountain in the world. Most people don't know what it is. And to be fair, I'd have to check myself. So I'll throw myself under the bus. I want to say K2, but I don't think that's correct. And there's lots of instances where this is, you know, can be seen. One example I like, and again on an audio podcast, it's hard to, to use it, but chess you know, if you're an a very good chess player, even if you just played a lot of chess, you might be able to make a move very quickly because it's just, boom, this is what you do in this situation. But you might also encounter times where you're forced to really think or deliberate what you want. And that's a System two decision. So it's important to realize that you can have a similar context, and sometimes it's system one and sometimes it's System two. It really, really depends on the context, like I said. So like I said, I'm going to come back to Neuroanatomy in depth on the next episode. But just to give you a precursor, because we've talked about this on previous episodes, typically system one is associated with midbrain structures, all right? And you don't really use the cortex that much, whereas system two is the prefrontal cortex, which we've talked about a lot over the history of this podcast. So if you haven't seen it yet, go to Google images and type in prefrontal cortex. Now, let's just get into the nitty gritty a bit more here. As I said, if you talk about system one and System two, you're sort of buying into what we call dual process theory. I mentioned that. And there's two sets of systems, and these two systems are used to explain reasoning, judgment and decision making. Some of the descriptors that you can think of is that in terms of types of processing, system one is automatic, implicit and associative, whereas system two is controlled, explicit and rule based. All right, makes sense. So system one is this automatic thing that just happens. System two is this more explicit, deliberate thing. If you think of the actual characteristics of the decisions, system 1 decisions are typically thought to be fast, effortful and erroneous. All right? In other words, you can be error prone when you make these rapid gut hunch decisions, whereas system two decisions are spot to be more slow, they're effortful, and as a result they're more accurate. So in terms of brain, another way to think of system one and system two, and this is without going into working into the anatomy. But system one requires working memory. Sorry, system one doesn't require working memory. Working memory, of course, is when you recall things from long term memory. And the reason it doesn't need working memory is because it's an automatic decision, whereas system 2 requires working memory because to make these analytical decisions, you're recalling stuff from that you've learned to help you make the decision. All Right, and basically that makes System one autonomous. Right? It's just works on its own and it gives you an answer. System two is not autonomous. It requires other brain regions and other input. One example that I kind of like, Kahneman uses this, so I'm going to steal it. But system one is basically things that happen to you, and system two is things that you do. So if you make an effortful choice to do something, that's typically System two. Whereas if you're just responding to the world, that's why he says things that happen to you, that's System one. Now I'm going to dive into the literature on this stuff and do some brain regions. But today I just wanted to give you this high level overview of these two systems. All right, so, so let's think about how this fits into what we've talked about before. So you're presented with a situation and you assess the options you have. Then you add value to things and then you select an action. Now if that happens automatically, very quickly, that's System one, all right? And basically in a System one decision, you don't really assess a lot of options. It's more, more like a stimulus response thing. So you, you, you get a certain stimulus, a light goes red. Your response, you put your foot on the brake. All right? System one decision, you're not assessing value. So imagine you're driving your car, you're coming up to a stoplight and all of a sudden the light goes red. You're not sitting there going, well, should I put my foot on the brake? What's the value of that? Or should I, you know, should I just keep going? What's the value of that? No, it's just an response. Okay, you put your foot down. System one decision, however, imagine you're driving your car and all of a sudden you start to slip on ice. Now if you're, if you're very experienced in driving in ice, you might actually have a system one response. But a lot of us would very quickly go, what are my options here? And you would make a system two decision. Now that's pretty rapid, so it might not be the best example, but when you assess values and you assess all, all these choices and you reflect on it, that is kind of a System 2 decision. So our simple decision model of always choose the highest value, all right? If you're just doing that reflexively, you can think of that as system one, all right? You don't even process any other values. Whereas if you go through the Explore Exploit process that System two, all right? Because you're looking at the other options out there. And one way to think of the explore exploit dilemma is when you're exploring, one could argue that you could actually choose the highest value option because you assess it's the best choice. All right, there's a bit of debate there, but you get what I'm. Hopefully you get what I mean. If, if you make a system one decision, there isn't any real assessment of value. You just go with the highest value option. Quick decision. System two, you're more deliberating the choices in front of you and looking at those values or deciding to choose an unknown value. Now, we'll get into this more because I'll do a whole episode on risk and uncertainty. But one way I like to think of system two is that when you're uncertain, imagine that two values are very similar. Well, that's going to be system two, all right? Whereas if one value is clearly out in front of the other values, that System one. Now, the way that Kahneman phrased this is you would either do system one or system two. It's one or the other. That's the dual process model. A lot of people believe in it. It's a thing I believe personally, and I'm not the only one in more of a continuous model. And what I mean by that is your brain is navigating the world. It's faced with decision points, points, and system one is throwing out answers for these decision points. But at the same time, system two, the prefrontal cortex, is sitting there monitoring what's going on. All right? And if there's a reason to engage, uncertainty, risk, any number of other factors, then system two will engage. And it's also how much it needs to engage. Right. System two might just add a little bit of effort to help you make the decision, or it might add a ton of effort. Now, we're going to come back to that concept, but I just wanted to leave you with this idea that system two might not be just. It might not just be this dichotomous thing you're doing. System 1, decisions 1, 1, 1, 1, 2, 1, 1, 1, 1. That may be what's happening, but I think it's far more likely that the prefrontal cortex is always keeping an eye on the situation, kind of like a, you know, like the hound dog who's lying there and opens one eye every once in a while to see what's going on. And if needed, prefrontal cortex brings its weight to bear to help you make the decision. All right, that's an introduction to System One, System two, Hopefully it is clearer than mud. And on the next episode, like I said, I'm going to bring this back to Neuroanatomy. All right, thank you so much for listening. Don't forget the website that neuroscienceguy.com links to Etsy and Patreon, where you can support us by buying merch or donate to the podcast directly. Remember, all of the money goes to graduate students in the Kerg Olson lab. Of course, there's social media at that Neuroscient guy on Instagram, threads and X. Please send us ideas. You know we're not going to just let me ramble on about decision making for the whole season. We do want to know what you want to know about the neuroscience of daily life. And of course you can email us that neuroscience guy gmail dot com. And last but not least, the podcast itself. Like I said at the outset, thank you so much. I cannot believe that we have over a million downloads. Blows my mind. I'm actually almost speechless. It rarely happens. With that note, I'll wrap it up. My name is Olaf Krigolsson and I'm that neuroscience guy. I'll see you soon for another full episode of the podcast.