What is the Limit of Our Brains in Predicting Future Events?

Hello and welcome everyone. This is Active Inference Guest Stream 88.1 on September 17th, 2024. We're here with Hunar Abdulrahman and we'll be discussing what is the limit of our brains. There's a fun backstory here maybe we'll go into. There will be a presentation followed by a discussion so I'm very much looking forward to it and to reading people's comments and questions too. So Hunar, thank you again for joining and to you for any backstory and introduction and for the presentation. So thanks again. Thank you Daniel for inviting me to this nice stream. Actually it's the first time as I told you before we start to do a live stream using this software so I hope the audience will forgive us if there is any upcoming technical problems. Hopefully there will be not. So thanks a lot for inviting me again. My name is Hunar. I have a PhD in neuroscience. I spent many many years trying to learn how our brain works using functional MRI, EEG. So I did my PhD at from, I get my PhD from the Cambridge University and we will be talking a bit about what I did in my PhD and how I try to link my, the concepts that I learn about brain with the outside world with other concepts like the free energy. So I'm actually a newcomer to the free energy principle. I hope that people with technical background forgive us if I have any technical errors because hopefully maybe I'll try to learn and share my insights about what I have discovered from my searching in the past few years. So is it, can you see the slides if I start? Great. So the topic is actually about what's the limit of our brain and I will try to talk about two, two concepts. I, I'll talk about some few concepts and later on we will link them all together. So first, I'll try to talk about the brain as a predictive machine, usually the predictive coding account, then I will try to ease into the free energy principles. And, and then later on, I will try to talk about what makes prediction difficult. I will talk about extrinsic factors. So even if I have, if we have the best learning machine, the best machine learning algorithm, or even if our brain is perfect in learning, there are still a lot of difficulties, basically because of computational irreducibility principle, which I will be talking more about it in the upcoming slides. So I will try to, I would love to start with repetition suppression is a well known concept in neuroscience. So if, if somebody wants to go into the details, what's repetition suppression, I will recommend to you this article by my PhD supervisor Rick Hansen. But it's a bit difficult to wrap your head around the results. So I will try to make it easy. Basically what repetition suppression means, the activity that we record from our brains, we are using EEG, fMRI, or single cell recordings is the for example, if you see a sequence of pictures of elephants, and then later on, you will see something surprising or novel, like the kangaroo image here, the activity of our brain increases increases. However, when we, when you see repetitions of the same image here, like the kangaroo here, the activity decreases. That's what we call a repetition suppression. If you repeat something, the activity of your brain, your brain will not really be responding in the same way it responded the first time. Then you will see something new like this beer here. Because it's novel, the activity of your brain increases again. So, so this repetition suppression is not only for repetition, it sometimes we call it expectation suppression. So, so even if something is not repeated, but not expected, the activity of your brain increases. But if you expect something, so during, it's usually repetition and expectation are synonymous. If something repeats a lot, it will be expected. If something is novel, it's unexpected. So these results, you will be finding in whether repetition studies or expectation studies. They will tell you, if you see something unexpected or novel, the recordings of your…