Active Inference: The Free Energy Principle in Mind, Brain, and Behavior

hello and welcome it's February 20th 7th 2023 we're here in active inference bookstream number 2.1 with Thomas Parr as well as Terry and Ali so thank you all for joining this should be a very fun discussion and let's go first by way of introduction and beginning to Thomas welcome and please how did you come to write this textbook in collaboration with pizulo and friston well again thank you very much for having me here um how did we come to write the textbooks good question I think um I think Giovanni was the the person who really had the idea and pushed for this that I think um Carl has already previously been approached by um by the editor who ended up managing application of the book um and when Giovanni had the idea that this would be a really useful contribution um he and I had a number of discussions about it both in in London and in Rome where he's based um and sort of put out a plan and and sort of came up with something that we hoped would be useful for the community I suppose part of the part of the reason why we thought it'd be a useful thing to do is that active inference is a very multi-disciplinary type field that draws a lot from lots of different domains and I think one of the one of the struggles you know we all had when we started out in it and I think certainly what the students and uh people who are interested in the topic say to us is that it's often quite difficult to trying to find all of these different topics trying to find the right background when they don't all come from one field so putting it all together in one book with a consistent notation consistent language and trying to make it complete we hoped would just make the field a little bit more accessible um to others who wanted to get involved how did this dovetail or integrate with your own Education and Research directions like what phases or stages were you moving through well this discussion was happening well I think so when we initially started discussing it it must have been probably the last year of my PhD and so by that stage I'd spent uh three or four sorry two to three years um sort of getting to grips with active inference um having spent you know being relatively recent in terms of learning some of the key material myself I think it was quite a good stage to be at in terms of thinking about what I would have wanted to know a few years previously um the actual writing of the book then carried on over the next year or two um and so by that stage I was I was back in clinical training as well which I think is also a useful perspective in terms of some of the later chapters and thinking about application of active infant site models in the context of clinical conditions and Psychiatry and neurology and those sorts of areas that's actually a really interesting note how did the clinical experiences give a perspective on the book or how did the book or you're learning in these abstract and theoretical domains give a different Insight in the clinical setting it's a good question and I think you're right to highlight the bi-directionality of it that both the useful in in different ways um I think for me the clinical setting is quite useful in terms of thinking you know first of all in interacting the different sort of different sort of um audience so people who may be much more skilled on the clinical side of things maybe add less experience on the mathematical computational side of things but may still find the concepts very useful so from that side of things it's quite a useful exercise in being able to communicate these things um and then there's sort of the interaction on the patient side of things and trying to you know I think it makes a big difference first of all in interacting with patients has an impact on how you might think about the interesting questions and the things you might want to use them to entrance for but also having an understanding of principles behind it helps you I think formulate your questions about how…