1:07:55 KF: But back to the, you know, why not just Bayesian inference, and what’s the difference between minimizing free energy and Bayesian inference? So in a minimal sense, you’ve got a notion of the future through a linear first-order approximation. It seems like the free energy principle and the concept of attractor sets is an extension of the fitness landscape framework which, I hope, will allow to include cultural and personal dimensions into the unifying theory of evolution. But perhaps if you are an internet person, you know him best for his extremely successful blog, Marginal Revolution, that he runs with his colleague at George Mason University, Alex Tabarrok. 0:29:17 KF: The twist, the slightly paradoxical aspect of this is when you move into the future and when you have the beliefs about the consequences of an action, say looking over there or googling a certain entry or going to Wikipedia, before you make that action, you have beliefs about how your free energy is going to change. Other times, we get a little deeper, we kind of get our hands dirty, we get into the weeds, we try to dig into some specific example of something. 0:53:11 SC: Sort of algorithmic compressibility. Well, it basically means I’m trying to find a low energy explanation for my data, whilst at the same time keeping my options open. So that’s the key differences. 0:15:37 SC: That’s asking for trouble, really. Whereas Ariel, you could see that she’s trying to figure something out about what would happen subjunctively if she did something, and her little kitty brain is trying its best. And it’s got its attracting setting, it has its prior belief that the temperature should be like this, and all I need to do is to minimize my prediction error, minimize my free energy. So in some sense are you setting an example, just as much as you’re actually trying to help the world with this gesture? 1:29:34 KF: Thank you. Of course, the roots are much of self-organization to non-equilibrium steady state inherit from the work of people like Ross Ashby who made apparent his ideas through the homeostat. Uncertainty in relation to its model of how we think the world works and in particular how it is situated within that world and sampling from that world. But they live. He supports a family of… His income is really quite low. 12:25 TC: It’s not for me to judge. So there’s a selection process in play, which must in some sense speak to the freewill. 02:34 SC: Now you can argue with this, and of course it’s not quite that naive if you concentrate on short term economic growth and destroy the planet then you haven’t increased anybody’s utility. And you’re saying people alive a hundred generations from now, should count for almost as much. And the solutions, doing the things that you just listed, all which sound great, like you say, “we don’t know the direct path to doing them. 1:25:11 SC: The error bars in some sense. So prediction errors are just the mismatch between what your generative model predicted and what you actually sample, you take the sum of squared prediction errors, you weight them by some precision or inverse variance. 0:56:16 KF: Absolutely, yeah, yeah. So it’s a question of learning how to communicate. Well, you just create a tractable bound. So if you can write down the generative model, you can then write down the differential equations of these, sort of, sensory inactive internal states. Will it work now? So a few of people trading the cards who are not used to doing this, once they own the card, they value it as much more just because it’s theirs, like they identify with the card, it’s mine, but professional traders are in a sense, more ruthless, they’re more willing to just buy and sell at whatever is the best price. So I see traffic is much worse. 1:07:44 SC: Physical measure. 0:31:53 KF: We’re not little marbles or moon rocks. Instead, our system considers things like how recent a review is and if the reviewer bought the item on Amazon. There’s a lot of food for thought. 0:34:24 KF: Yeah, it is. It’s not right before us, but if we invest much more heavily in basic science, we’ll have a chance. 0:16:14 SC: And we would like them to match. One is with the Manhattan Project in World War II, Physicists sort of like myself got a lot more credit than we deserve for making the world a good place, or at least affecting the outcomes that happen in the world. It’s getting people to want to do it. 1:05:17 KF: Yeah, absolutely. Suddenly you become a creature that seeks out sensations, literally sensation-seeking, that resolve uncertainty, you become curious and you go to your discos and you do your bungee jumping, at a certain age. how it performs, what Sean always calls „coarse graining“ (and is the basis for entropy, free energy etc. So, what we are describing are classical reflexes. 0:09:45 KF: The essence of the insight was, well, there’s a much simpler explanation of what’s going on here for this sort of very elemental form of self-organization ensemble dynamics. And this new book Stubborn Attachments, it’s an attempt to outline the philosophical foundations of our economic judgments.