It’s a great relief to know that e.g. Iain McGilchrist isn’t “alone” in this field. Yes, his approach is different, but too many specific criticisms and descriptions are in resonance. The most important to me - as I generally study modernity and its effects on us - is the dialectic of the reductionistic and the holistic. Yes, philosophy and social idealists have indeed brought this forward, especially as modernity took hold some 2-300 years ago. Alas, our idea of education has slowly left general philosophy out, leaving us “disabled” and individualized in a social reality. As in many of us living in tight heaps, societies and a global finite whole. Well, that last bit is on me - I’ve become very biased from my studies and just being… 🔥💛
I don't know why I'm choosing youtube as a means of transferring this idea, but I think 41:33 really shows the limits of control theory on complex systems. As it turns out, no one could have predicted an interaction like this using a systemic lense, because the number of features is practically infinite, and the causal chains are impossible to compute preemptively even worse, politically, the fact that ammonia producing companies use a systemic decision theory could have cost us two consumer goods (and I guess also produce???) because their modus operandi is so rigid. No Agent involved in the system actually wants this catastrophic failure, but it happens because of adherence to profit motive as only means of decision I know for scientists this really doesn't matter much, and low rank interactive causality is for sure more predictive then low rank monodirectional causality like you nicely lay out in your talk But I think in clinical practice, where its not feature outcomes but wellbeing which is in effect the treatment goal, a full rank causal model is a necessity (and I would argue even already in effect as best practice in therapy, which focuses on understanding of patients by the practictioner, a neural description of the entire MH problem with marginal effects included). A systemic lense will always produce the Goodhartian fallacy, because the numbers in your control systems chart don't actually care for the wellbeing of the client. This is a subtle point, but I believe strongly that there must be a a non trivial disconnect between numerical description and actual reality. That no System will ever converge to a Process of sufficient complexity, even when adding details until your computer gives up My personal field of study in philosophy of science aims at replicating full rank causal models in a reproducible way, with explainable AI or Real life analogy, and to be perfectly frank I am just a early 20s non academic nutcase studying this in my room at home, but I thought as this is highly relevant to your talk I'd share this here anyway
A perfect example of why people who categorize should study ontology!! Bravo!
It’s a great relief to know that e.g. Iain McGilchrist isn’t “alone” in this field. Yes, his approach is different, but too many specific criticisms and descriptions are in resonance.
The most important to me - as I generally study modernity and its effects on us - is the dialectic of the reductionistic and the holistic. Yes, philosophy and social idealists have indeed brought this forward, especially as modernity took hold some 2-300 years ago.
Alas, our idea of education has slowly left general philosophy out, leaving us “disabled” and individualized in a social reality. As in many of us living in tight heaps, societies and a global finite whole.
Well, that last bit is on me - I’ve become very biased from my studies and just being… 🔥💛
I don't know why I'm choosing youtube as a means of transferring this idea, but I think 41:33 really shows the limits of control theory on complex systems. As it turns out, no one could have predicted an interaction like this using a systemic lense, because the number of features is practically infinite, and the causal chains are impossible to compute preemptively
even worse, politically, the fact that ammonia producing companies use a systemic decision theory could have cost us two consumer goods (and I guess also produce???) because their modus operandi is so rigid. No Agent involved in the system actually wants this catastrophic failure, but it happens because of adherence to profit motive as only means of decision
I know for scientists this really doesn't matter much, and low rank interactive causality is for sure more predictive then low rank monodirectional causality like you nicely lay out in your talk
But I think in clinical practice, where its not feature outcomes but wellbeing which is in effect the treatment goal, a full rank causal model is a necessity (and I would argue even already in effect as best practice in therapy, which focuses on understanding of patients by the practictioner, a neural description of the entire MH problem with marginal effects included). A systemic lense will always produce the Goodhartian fallacy, because the numbers in your control systems chart don't actually care for the wellbeing of the client.
This is a subtle point, but I believe strongly that there must be a a non trivial disconnect between numerical description and actual reality. That no System will ever converge to a Process of sufficient complexity, even when adding details until your computer gives up
My personal field of study in philosophy of science aims at replicating full rank causal models in a reproducible way, with explainable AI or Real life analogy, and to be perfectly frank I am just a early 20s non academic nutcase studying this in my room at home, but I thought as this is highly relevant to your talk I'd share this here anyway
This is a wonderful presentation. Thank you so much!
Thanks!