Really liked this one. For me it raises a question that we have canvassed before: if some things are unknowable (whether ultimately or immediately), what thinking frame do we bring into making decisions and taking actions. Humility guides us well in determining how confident we can be about the 'truth', but less well in reacting to that truth.
I see a difference between something being (completely) unknowable and something being only partially knowable. Almost every case where we are making decisions, we are dealing with partial knowability. I see an advantage in humility in that it allows us to claim some knowledge, but not base our decisions on things we think we know but are actually wrong. I think it can calibrate our decisions to the evidence, rather than the common practice of building a whole program of work confidently off not very much.
Quantum theory presumes that we know what matter is but it is an hypothesis. That's not a criticism of QFT, just an obvious truth that we forget once the equations get going and the predictions seem accurate. David Bohm was on the right track: What we see (the explicate order) might be a surface-level unfolding of a deeper, hidden structure (the implicate order), where relationships are encoded in a way that isn’t obvious or local. Thank your for your work!
Zoologist JBS Haldane spotted this 'unknowability' presumably from the physics about a 100 years ago. Is it possible however the human mind can 'know', inexplicably but profoundly, from being made of the same stuff? Perhaps much as nature performs roughly the Fibonacci iterations? Or, again, in us partly consciously by what Iain McGilchrist calls attention, a kind of implication from the particulars by analogy? Or yet again to 'know' partly by pattern recognition as you neatly describe for the oddity in model-generated representation of water flow?
I wonder if AI pattern recognition systems trained on real water observation can see the same oddity in the same odd representations.
Thanks. Some ideas I'll keep thinking about. Unknowability is not new, but is fiercely resisted by lots of people. Of course, McGilchrist would argue that is due to the dominance of Left Hemisphere ways of knowing in our culture.
We do have a huge amount of data about weather that we don't have for water flow, especially as the system for water flow is different in every case. Something to keep an eye on.
Thanks. McGilchrist argues from clinical and other evidence for two kinds of attention for selecting and resolving sensory information, these having twin development throughout evolution. I suppose evolution can be seen as engaging aspects (by selection) of the universe available to biomolecules in cooperative relationship becoming living organisms. This to my mind seems to mobilise a 'telos', with an interesting relationship with time, as well as with thermodynamics. (They/we look back as well as forwards?)
McG's work is for me a mountain largely still to climb. I find his position on 'consciousness' difficult (in humans and other species, a lot of energy/work goes on 'unconsciously'), but I guess he has a point when he flags up different modes of 'attention'. The so-called Left Hemisphere (LH) attention might follow 'deterministic', perhaps 'algorithmic' process, perhaps including including pattern recognition. (I presume in organisms, pattern recognition might resolve by comparing by analogy, perhaps in the form of persistently available sensory constellations. I note there was an extraordinary fairly recent revision of estimates, now suggesting vast 'unforgettability', recognition of images in the human memory; a truly extraordinary efficiency. I presume the same applies to other senses. Olfaction is a particular example, even if we are not as acute as dogs.)
Turning to the attention, the emphasis he locates in the RH, I am reminded of Roger Penrose favouring the existence of non-algorithmic procedures available to cognition, which could put this kind of cognition, perhaps conscious recognition itself, in participation with a different aspect of the available universe than algorithmic 'computation', or more broadly 'determinism'. McG seemingly wonders about access to forms of 'knowing', if we are humble enough, that relate intelligibly to understanding where there is plainly mystery.
I wonder what AI can achieve if it does not have a 'RH' connection to 'understanding'?
That kind of spilled onto the page. Forgive my musing! Thanks again.
Really liked this one. For me it raises a question that we have canvassed before: if some things are unknowable (whether ultimately or immediately), what thinking frame do we bring into making decisions and taking actions. Humility guides us well in determining how confident we can be about the 'truth', but less well in reacting to that truth.
I see a difference between something being (completely) unknowable and something being only partially knowable. Almost every case where we are making decisions, we are dealing with partial knowability. I see an advantage in humility in that it allows us to claim some knowledge, but not base our decisions on things we think we know but are actually wrong. I think it can calibrate our decisions to the evidence, rather than the common practice of building a whole program of work confidently off not very much.
Quantum theory presumes that we know what matter is but it is an hypothesis. That's not a criticism of QFT, just an obvious truth that we forget once the equations get going and the predictions seem accurate. David Bohm was on the right track: What we see (the explicate order) might be a surface-level unfolding of a deeper, hidden structure (the implicate order), where relationships are encoded in a way that isn’t obvious or local. Thank your for your work!
Zoologist JBS Haldane spotted this 'unknowability' presumably from the physics about a 100 years ago. Is it possible however the human mind can 'know', inexplicably but profoundly, from being made of the same stuff? Perhaps much as nature performs roughly the Fibonacci iterations? Or, again, in us partly consciously by what Iain McGilchrist calls attention, a kind of implication from the particulars by analogy? Or yet again to 'know' partly by pattern recognition as you neatly describe for the oddity in model-generated representation of water flow?
I wonder if AI pattern recognition systems trained on real water observation can see the same oddity in the same odd representations.
Thanks. Some ideas I'll keep thinking about. Unknowability is not new, but is fiercely resisted by lots of people. Of course, McGilchrist would argue that is due to the dominance of Left Hemisphere ways of knowing in our culture.
And that's an interesting idea about AI pattern recognition and water flow. There are already signs that AI will be better at weather prediction than our existing models as it just looks for patterns rather than tries to calculate the underlying physics. (e.g. https://deepmind.google/discover/blog/graphcast-ai-model-for-faster-and-more-accurate-global-weather-forecasting/)
We do have a huge amount of data about weather that we don't have for water flow, especially as the system for water flow is different in every case. Something to keep an eye on.
Thanks. McGilchrist argues from clinical and other evidence for two kinds of attention for selecting and resolving sensory information, these having twin development throughout evolution. I suppose evolution can be seen as engaging aspects (by selection) of the universe available to biomolecules in cooperative relationship becoming living organisms. This to my mind seems to mobilise a 'telos', with an interesting relationship with time, as well as with thermodynamics. (They/we look back as well as forwards?)
McG's work is for me a mountain largely still to climb. I find his position on 'consciousness' difficult (in humans and other species, a lot of energy/work goes on 'unconsciously'), but I guess he has a point when he flags up different modes of 'attention'. The so-called Left Hemisphere (LH) attention might follow 'deterministic', perhaps 'algorithmic' process, perhaps including including pattern recognition. (I presume in organisms, pattern recognition might resolve by comparing by analogy, perhaps in the form of persistently available sensory constellations. I note there was an extraordinary fairly recent revision of estimates, now suggesting vast 'unforgettability', recognition of images in the human memory; a truly extraordinary efficiency. I presume the same applies to other senses. Olfaction is a particular example, even if we are not as acute as dogs.)
Turning to the attention, the emphasis he locates in the RH, I am reminded of Roger Penrose favouring the existence of non-algorithmic procedures available to cognition, which could put this kind of cognition, perhaps conscious recognition itself, in participation with a different aspect of the available universe than algorithmic 'computation', or more broadly 'determinism'. McG seemingly wonders about access to forms of 'knowing', if we are humble enough, that relate intelligibly to understanding where there is plainly mystery.
I wonder what AI can achieve if it does not have a 'RH' connection to 'understanding'?
That kind of spilled onto the page. Forgive my musing! Thanks again.
Among my favorite quotes:
“The universe is not only queerer than we suppose, but queerer than we can suppose.”
—J.B.S. Haldane