Max Bennett: belief

Transcript context

…I I think having that second level of of metacognition, it it does a lot of work. Right? And I'm gonna talk a little bit about that now. But 1 of the things is you can infer the the intents and knowledge of others through the same process of doing simulations yourself. So you can kind of imagine, let's say, yourself doing something, but you can kind of swap out the pointer to be someone else and swap out the knowledge to be someone else, and that's incredibly valuable. But the knowledge thing is is really, really, really interesting. So I I asked the question last time, and this is something that I've been quite confused about, and I feel that reading this chapter has actually really cleared up for me, which is about goals. Because when you look 1 level down at the agranular prefrontal cortex, it's modeling intense. And then this granular prefrontal cortex, which is trying to seek explanations about the level below, which is the agranular prefrontal cortex, it it's it's going a level of abstraction up and it's modeling goals, not not intense. But it's actually modeling knowledge. It's what it's doing is it's categorizing. So when you have a simulation of simulate a a simulation of simulations, what it's doing is is it's creating a category. So to the example you just gave before, thirsty becomes a category. So rather than it being a pointillistic intent, you know, it's little bit like saying, can go and have a sandwich or I can go and have McDonald's, or my abstract simulator could kind of draw a boundary around those things, and now I'm getting food. So as well as being able to categorize intents in yourself and other people, you're also categorizing knowledge, and then it can be shared memetically. So it's almost like just going to that second level of metacognition just just gives you so much that you didn't have before. 100%. Yeah. I think the Thinking about the level of granular prefrontal cortex in the new primate regions as enabling something akin to knowledge, I think is a really wonderful way to look at it. 1 from the connectivity analysis, and then 2, just from what we mean by knowledge. So connectivity analysis, if you look at superior temporal sulcus and the temporoparietal junction, these are regions of the posterior cortex that, in simple terms, are at the very top of the hierarchy. I mean, they get multimodal input from all the other regions of sensory cortex. So a very sort of simple, rule of thumb for understanding this is just this models the rest of sensory cortex. I understand the full rendering of the simulation of the external world that is happening, and this is where I build a representation of that. And it is perhaps no coincidence that's also where we see, brain regions light up when you're engaging in things like theory of mind and solving false belief tests. In other words, trying to infer the knowledge of someone else. These same regions light up. And what do we mean by knowledge? I would argue that knowledge can mean a few things. 1 is procedural knowledge, where I just know how to do certain motor behaviors. I don't think that's what we mean. I think we mean more semantic knowledge or episodic knowledge, which would be I know that water is over there. And I know that if I do this behavior, this will be the causal outcome. That type of knowledge, I think, is absolutely rendered in the mental simulation. When I imagine certain things when I imagine the case of lightning hitting the ground, what do I see afterwards? I see fire. And that that that's the source of my knowledge about the causal relationship between these 2 things. So having a layer that models the simulation enables me to reason about my own knowledge and to see what the effect of changing knowledge is on behavior. And this, of course, enables us to flexibly adapt to other people's behavior and predict what they would do under cases of different knowledge and different intents. Yeah. That's fascinating. But you did say that the there was a bit of a riddle about the granular prefrontal cortex because there was 1 study where it could be damaged, and the person would still score really highly in in IQ. But you said it's about being able to project yourself in simulations, you know, this kind of like abstract modeling your own mind. So in this particular case, how could the person still score the same IQ without that part of the brain?…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.