Basics of ActInf (Discussion) — Course home

all right July 25th 2023 and we are in the discussion section basics of active inference this is the first discussion section that we've had for this course so thank you all for joining and it's going to be regarding the topics that Ben White introduced in his recent lecture so everyone will be welcome to pop in and introduce themselves and then I know that Ben has some ideas to discuss and many other spontaneous and written questions will come into play so I will uh with that exit for now and pass to perhaps some of our first time live stream guests I guess I've already spoken so uh I may as well continue so I'm Darius um I am Master student I'm just about to finish at UCL and sort of social distributed cognition I work in the social cognition lab looking at salience regulation and attentional mechanisms within social contexts and but all within an active inference framework um within the Bayesian brain hypothesis framework and yeah it's a sort of recent Discovery and Obsession and so I'm sort of going to grips with the both the high road and the low road um and so I've had the opportunity to chat to Mark and some other researchers who have been super informative but always looking to learn more and get my head around even more of the sort of philosophical and mathematical Theory oh sounds good so hey everyone can you hear me yep I'm Francisco balzan I'm a PhD student at the University of Bologna Italy actually really working on the intersection between artificial intelligence and education and I have a background in cognitive anthropology and philosophy of science and I'm at excellent I'm actually in a a few years ago so I stepped out into the rabbit program active inference and currently pretty much interested in uh um multi-scale active inference models are the scientific technicians so pretty interested about the agent level modeling of scientific reasoning the assumption that might emerge from the interaction with the scientific environment so we're referring to Scientific music construction and all this type of top-down environment interactions so super interested to do something from you thank you thank you very much would anybody else like to introduce themselves before we start thank you so you can first okay hi hi I'm Regina Sagi umina I'm from Guatemala and well I'm not from the social sciences I come from biology and Neuroscience and um I'm here because um I'm working on a project as a technical operator in this big project in DX Escape material mines and I do the technical part and experiments but I want to learn the philosophical part that's the background that you know makes The Theory of Everything we're going to you know experiment so I'm here to learn um and also I am like you know when you see the Olympics that you you see all the swimmers and you always say I wish there was a normal guy to see how good these people are so you could compare so that's me today right I I understand half of what's going on but I'm happy to be here so yeah hi oh yeah sorry I'm late um sorry my name's Lee um I'm a PhD student at the uh University of New York um studying systems transformation um and I'm I'm joining because I'm using uh perceptual control theory at the minute to uh to model examples of effective practice um in organizations and and I've read quite a lot of um active inference papers and I understand that there's a lot of resonance uh and overlap between perceptual control theory and um active influence although I understand it's within a predicted framework and also it's uh it's a kind of a level of abstraction higher so you're able to kind of quantify the difference between um sort of the the predictive State or the outcome State and you know in the current state so what I'm really trying to understand is how might that level of abstraction be useful um in what I'm doing actually because I understand it's a you know a much more kind of concurrent Theory and framework than um then perceptual control theory yeah…