Consciousness, Information Geometry, Ethical AI

hello everyone and welcome back to active inference insights I'm your host darus parv Wayne and today I'm delighted to be speaking to Vana Visa van is a postdoctoral fellow at Johannes Gutenberg University mines in the department of Ro University B sorry the internet has lied to me the internet has lied to me I don't like this introduction bit already um but I think this is true nonetheless his research centers on Consciousness medal rep presentation philosophy of cognitive science philosophy of mind and of course Act of inference uh we also both share a love of self- modeling attention autopoetic intentionality and much much more and I'm sure these are just some of the topics that we will discuss today V welcome to the show thank you so much for joining me thank you for having me it's a pleasure this is really exciting uh yeah as I said in the intro we share a lot of interests uh that might appear quite Niche to people but actually you know as far as I can tell pertaining to their lived experience every every moment so it's it's worthwhile dissecting funnily enough the first time I think I'd heard of you was actually from jackob Javi um in the conversation I had with him and he said that you know Thomas metzinger's work perhaps better than Thomas and certainly better than anyone else um so so I I I wonder whether we could start by outlining some of these crucial um Concepts that sometimes seem a bit counterintuitive to people so for example when I say we both share a love of self- modeling or phenomenal self modeling what does that mean what ises this sort of acronym PSM the phenomenal self modeling yeah good so um maybe as a um disclaimer of course um Thomas knows more about this than me he's um uh in invented the term self models and has developed the self model theory of subjectivity over many many years and um actually I I I think I can say that there's probably no living philosopher who has read as much relevant literature about topic as as Thomas because he's devoted his entire career to this and works a lot and of of course I would never um even even if I tried be in a position to know as much about this as as he does and I I think it might be a good idea even to invite him to this podcast he's been invited he has said yes that was a while ago I've tried to push it so Thomas if you're listening which I think he might be you know you never know um please do come on because I do yeah so actually it's quite likely that some of the things that I've be saying might not be 100% true so Thomas if you're listening now you may want to correct some of the misrepresentations of your work um but anyway yeah what what is the self model so for me it's I think easiest to First think about bodily self models so um just imagine some creature or a robot that has a body and has to control the body and as we all know in many cases control can be facilitated ated or improved if you have a model of the thing that you're controlling and if you want to control your body and the interaction with the environment it can be highly useful to have a model of your body and the model of your environment and this is actually then already a form of self model which is part of the world model so the world model would be just be representation of the world or the relevant part of the the world the the parts that are relevant to you and the your environment and your body is part of this world so it's the the self model is part of the world model but of course that's just a form of self model and it's quite primitive probably compared to the self models that we human beings have and one thing that distinguishes our self models is that we don't have only unconscious self self models but actually we have uh we also consciously experience the contents of our self models so this brings us on to to the notion of a phenomenal self model where phenomenal refers to phenomenal experience phenomenal Consciousness so if you have not just a um an unconscious self…