Textbook Group meeting for Parr, Pezzulo, Friston 2022 .

all right hello everyone it is cohort five meeting two and we're in our first discussion of chapter one so to begin with what is anyone's just First Take or on chapter one or where does anybody want to begin and then we'll look at the specific questions but is there any just first overviews or just pieces that people want to share about chapter one or just this first discussion you can just go for it if you want to ask me so my biggest um question is around surprise and what the I guess the parameters and boundaries of that is well sorry I think uh a part of my long time persisting uh the literature is like the use of some terms like these and these might be because I'm also a non-native speaker but like the term of like a normative principle like still not sure what it means or like you know later on they say like epistemic this and stuff and I'm just like very confused by those uh terms uh so maybe clarifying yeah what it means to be a normative principle will be a very useful thing for me okay let's just see if it's been asked overall because as Ollie can probably attest to let's just like at least think of oh does anyone want to add a thought on this like just on uses of terms overall or we can look more specifically at what is meant by normativity like in chapter one foreign was also having some um uncertainty about the use of the term I had pre-released the I had an understanding that normative uh the term described something that uh the way that some something should be as opposed to being descript a descriptive theory for example which is which is simply stating what is observed but I uh in the way that it's introduced here normative is used more I think uh placing an emphasis on the ability to to look at a quantity and measure some behavior in terms of that quantity and um uh emphasizing that's better Behavior or behavior that's more in alignment with with the theory is suggesting will will optimize that quantity uh that's my understanding so far but I'd like to be more clear thank you Ali go for it yeah so normative um I mean the use of the the term normative uh here in active inference literature is closely related to the meaning of this term in uh say philosophical literature and stuff uh so uh basically in uh philosophy or more specifically in ethics when we talk about the normative Behavior or um I don't know a normative action uh it's uh it's interpreted as something ethical uh or moral and the reason we ascribe this kind of meaning to normal to this term normative is because we have a standard of action or behavior against which we can compare our own action and then judge by way of comparing it whether it is ethical or not so by the same token we can talk about normativity here in this context of active inference when we talk about a standard for optimization of our action and perception and namely the free energy against which uh the uh the error or the deviation of the action and perception from that standard can be evaluated and then corrected or accounted for so that's why uh I mean researchers in active inference area use the term normal normative or normativity to refer to this kind of compare comparison between the comparison between the action and a kind of a standard for optimization of that action I would kind of question whether that would be tied to ethics though because Norms don't necessarily align with ethical behaviors and of course ethical is subjective but you know um Norms have been you know for example been leaning more toward coercive and exploitive behaviors which in my in my world is not ethical but uh you know say I don't let I'd like to see the you know those two tied apart especially as far as how people act because ethics are you know are not are not Universal um and coerced coercive behaviors uh tend to you know get more get more benefits in U.S society but maybe not other places yeah I guess it's not like so used to me the connection with that it's you know it comes to these ideas…