As someone who thinks about the implications of machine learning for our understanding of taste quite a lot, I found this article really interesting. But since I try to integrate embodied/enactive perspectives, I'm not satisfied by such rote applications of the Bayesian brain hypothesis to anything we would normally call "subjective" (like taste.) Kathryn Nave's critique of Karl Friston and FEP is informative: https://direct.mit.edu/books/oa-monograph/5897/A-Drive-to-SurviveThe-Free-Energy-Principle-and
We don’t have to assume neither the free energy principle nor that the brain is doing Bayesian inference. We just need statement that the brain is constantly trying to predict (aspects of) the sensory future in some way or another. Predictive processing in its least restricted form, in other words.
I don’t know what you mean with rote exactly. As far as I can see, for reasons I tried to explain in that post and also in more detail in Beauty as entropic fine-tuning, if you accept predictive processing and its evolutionary explanation as your premise, then a lot of observations about art, taste, and beauty can be explained or partially explained in a unified way from first principles reasoning.
Of course, you can decide you don’t buy predictive processing. I don’t know exactly how compelling lab-based empirical evidence is, but I don’t think we need to rely on it either. Why I believe it is not really any kind of neuroscience work. It’s the combination of two things: first, it describes my subjective experience of beholding and making music and visual arts extremely well and, second, its evolutionarily extremely plausible. Bordering on tautology. Without prediction there is death. In some ways, the fact that PP unifies the explanation of so many aesthetic phenomena, including the mix of subjectivity and objectivity, seems to be some of the most compelling evidence for PP beyond its evolutionary rationale.
As for enactive/embodied perspectives, predictive processing is directly engaged in highly enactive process though, so don’t quite understand where they’d be in conflict. Not claiming that its a passive process. Engagement and play is of utmost importance in the whole story!
I have saved your post for later when I get the time, thanks for sharing.
I think this model explains the beauty of powerful imagery as well. Consider this poem: https://www.poetryfoundation.org/poems/52757/bermudas-56d2317c9465f?mc_cid=201e2ec0e6&mc_eid=5d58076011 Although I’ve never been to the Bermudas, the imagery is so vivid that the place is very literally constructed in my mind, giving a satisfaction at the sense made of the world. Also, such images give expression to the ineffability of subjective experience, enabling richer sense-making as we go throughout life. Enjoying your Substack!
Oh god, I don't know that this was missing from the summary. That was definitely an unintended omission. Thank you for the heads up, I gotta rectify that. Same with the typo!
As someone who thinks about the implications of machine learning for our understanding of taste quite a lot, I found this article really interesting. But since I try to integrate embodied/enactive perspectives, I'm not satisfied by such rote applications of the Bayesian brain hypothesis to anything we would normally call "subjective" (like taste.) Kathryn Nave's critique of Karl Friston and FEP is informative: https://direct.mit.edu/books/oa-monograph/5897/A-Drive-to-SurviveThe-Free-Energy-Principle-and
I come at this from a really different background. I'm not even sure I agree with everything I wrote about this last summer, but for what it's worth: https://elftheory.substack.com/p/taste-as-uncomputable-knowledge
Thanks for your thoughts here.
We don’t have to assume neither the free energy principle nor that the brain is doing Bayesian inference. We just need statement that the brain is constantly trying to predict (aspects of) the sensory future in some way or another. Predictive processing in its least restricted form, in other words.
I don’t know what you mean with rote exactly. As far as I can see, for reasons I tried to explain in that post and also in more detail in Beauty as entropic fine-tuning, if you accept predictive processing and its evolutionary explanation as your premise, then a lot of observations about art, taste, and beauty can be explained or partially explained in a unified way from first principles reasoning.
Of course, you can decide you don’t buy predictive processing. I don’t know exactly how compelling lab-based empirical evidence is, but I don’t think we need to rely on it either. Why I believe it is not really any kind of neuroscience work. It’s the combination of two things: first, it describes my subjective experience of beholding and making music and visual arts extremely well and, second, its evolutionarily extremely plausible. Bordering on tautology. Without prediction there is death. In some ways, the fact that PP unifies the explanation of so many aesthetic phenomena, including the mix of subjectivity and objectivity, seems to be some of the most compelling evidence for PP beyond its evolutionary rationale.
As for enactive/embodied perspectives, predictive processing is directly engaged in highly enactive process though, so don’t quite understand where they’d be in conflict. Not claiming that its a passive process. Engagement and play is of utmost importance in the whole story!
I have saved your post for later when I get the time, thanks for sharing.
I think this model explains the beauty of powerful imagery as well. Consider this poem: https://www.poetryfoundation.org/poems/52757/bermudas-56d2317c9465f?mc_cid=201e2ec0e6&mc_eid=5d58076011 Although I’ve never been to the Bermudas, the imagery is so vivid that the place is very literally constructed in my mind, giving a satisfaction at the sense made of the world. Also, such images give expression to the ineffability of subjective experience, enabling richer sense-making as we go throughout life. Enjoying your Substack!
Hey Patrick. Thanks for sharing this poem, I enjoyed it! Happy you found so value in my writing here.
4. I sense high complexity and surprise tolerance!
Thanks for these remarks dlkf!
Oh god, I don't know that this was missing from the summary. That was definitely an unintended omission. Thank you for the heads up, I gotta rectify that. Same with the typo!