Neat! This part of it helped me get a better model of Eliezer’s model of the lowest levels of subjectively accessible and controllable thinking!
There’s that stuff in the middle of the 30ms to 300,000ms zone where a thought that takes 4 seconds to happen (and which necessarily must have some underlying neurological basis) can sometimes need 4 minutes to explain to a third party… or can’t be transmitted that fast. Or can be explained but not turned into something they could repeat in 4 seconds on their own… or whatever.
Harry’s dark side, as I model it, is not actually supernatural. It is a bunch of stuff that got written into his brain and then erased by childhood amnesia. So he’s got a bunch of habits that chain into each other.
(I parenthetically mention that one of my deflationary hypotheses for why people say they get new thoughts when they’re on drugs, is just that some drugs, like psychedelics, disrupt patterned chains of thought. Normally whenever we think thought X, we then go on to think thoughts Y and Z in a familiar pattern. But taking psychedelics is one way to disrupt those patterns and think new thoughts instead. The deflationary hypothesis is that any kind of mental disruption would do it, that the results are not specific to the drug; you’d need to demonstrate some tighter correlation to get past the deflationary hypothesis for that drug.)
I don’t have a really strong mechanistic idea about habits, but I try to use the word “habit” in a way that is consistent with what I think I know about the basal ganglia… which controls gross motor stuff, emotion, and cognition, and is sufficient to let a de-corticated rabbit stay alive and sort of eat food (but is probably not sufficient to keep a de-corticated human alive, because (insofar as ethical experiments have been possible) it is probably the case that we and chimps and some other higher mammals are way way more “essentially corticated” than the little simple ones).
I had never previously focused on these ideas (habits vs 5-second-level) at the same time, but… it does seem like “sub five second” stuff probably sometimes involves intuitive deployment of valid reasoning leaps (like from mathematics) and this MIGHT actually be simply “based in habit”!?
In “Distinct Contributions of the Cerebellum and Basal Ganglia to Arithmetic Procedures” it looks like maybe reliable iteration (ie “counting”) leans harder on the cerebellum (as if counting was a fine motor skill) and “operational chaining” (like maybe a goto statement in a slightly conscious but still quite low level mental algorithm) leans harder on the basal ganglia, as if long division was the application of a looping habit?
So. Yeah. Maybe “sub five second stuff” just literally IS the application of acquired mental habits that are useful!?
At least this round of falsification pursuit didn’t rule out the hypothesis.
What I was expecting, naively, right after reading Eliezer’s theory, is that sub five second stuff is more “Hebbian, and neurological, in general” than “habitual (and based on the basal ganglia) as such”… which still might be true, but I’m less confident now.
Maybe 3 second rationality skills ARE just “all in fine motor skills and habits, possible directed inwardly, like in kinesthetic imagination”???
Regarding the larger set of data Eliezer talked about, from having tried a hallucinogen once (to gather subjective metaphysical data to see if “reality was really reality” when I was young and foolish)… it very much did NOT seem like the effects were localized to the basal ganglia or the cerebellum.
There was a lot of super low level visual cortex involvement, with things down in the Brodmann Areas 18 and maybe 17 and 19 (and maybe everywhere in the entire cortex?) involved as if “opponent process” processes for things like “motion vs not-motion”, and so on, were falling in and out of calibration.
As subjective data about subjectivity itself (like for trying to figure trying to figure out if Solipsism is true, or whether maybe there is only me plus some Cartesian Demon that is fucking with a hypothetical disembodied mind that is me, with all of external reality as an illusion), hallucinogens did just totally destroy the simple naive hypothesis that “cogito-ergo-sum-style subjective awareness” is independent (not caused by?) what happens in the brain’s firings...
...”the brain” just obviouslydoes cause “the subjective mind”, it turns out...
...unless the Demon’s powers extend to inventing a complex theory of neurology, and taking into account what hallucinogens would hypothetically do to a hypothetically incarnated mind, and then the Demon fed plausible lies along these lines into some hypothetically “metaphysically disembodied awareness” that was me… lol!
((You can’ always generate “an even more paranoid hypothesis”… its just that such hypotheses almost always become negligible under pragmatic anti-paranoid normalization <3))
Anyway. I’m not sure how the idea that “3 second level stuff maybe only happens in the cerebellum and basal ganglia” could be behaviorally applied, to get profits in some way, such as to know the VoI on ruling hypotheses of that class in or out...
But that was a cool part of the interview. Thank you! <3
Neat! This part of it helped me get a better model of Eliezer’s model of the lowest levels of subjectively accessible and controllable thinking!
There’s that stuff in the middle of the 30ms to 300,000ms zone where a thought that takes 4 seconds to happen (and which necessarily must have some underlying neurological basis) can sometimes need 4 minutes to explain to a third party… or can’t be transmitted that fast. Or can be explained but not turned into something they could repeat in 4 seconds on their own… or whatever.
I don’t have a really strong mechanistic idea about habits, but I try to use the word “habit” in a way that is consistent with what I think I know about the basal ganglia… which controls gross motor stuff, emotion, and cognition, and is sufficient to let a de-corticated rabbit stay alive and sort of eat food (but is probably not sufficient to keep a de-corticated human alive, because (insofar as ethical experiments have been possible) it is probably the case that we and chimps and some other higher mammals are way way more “essentially corticated” than the little simple ones).
I had never previously focused on these ideas (habits vs 5-second-level) at the same time, but… it does seem like “sub five second” stuff probably sometimes involves intuitive deployment of valid reasoning leaps (like from mathematics) and this MIGHT actually be simply “based in habit”!?
In “Distinct Contributions of the Cerebellum and Basal Ganglia to Arithmetic Procedures” it looks like maybe reliable iteration (ie “counting”) leans harder on the cerebellum (as if counting was a fine motor skill) and “operational chaining” (like maybe a goto statement in a slightly conscious but still quite low level mental algorithm) leans harder on the basal ganglia, as if long division was the application of a looping habit?
So. Yeah. Maybe “sub five second stuff” just literally IS the application of acquired mental habits that are useful!?
At least this round of falsification pursuit didn’t rule out the hypothesis.
What I was expecting, naively, right after reading Eliezer’s theory, is that sub five second stuff is more “Hebbian, and neurological, in general” than “habitual (and based on the basal ganglia) as such”… which still might be true, but I’m less confident now.
Maybe 3 second rationality skills ARE just “all in fine motor skills and habits, possible directed inwardly, like in kinesthetic imagination”???
Regarding the larger set of data Eliezer talked about, from having tried a hallucinogen once (to gather subjective metaphysical data to see if “reality was really reality” when I was young and foolish)… it very much did NOT seem like the effects were localized to the basal ganglia or the cerebellum.
There was a lot of super low level visual cortex involvement, with things down in the Brodmann Areas 18 and maybe 17 and 19 (and maybe everywhere in the entire cortex?) involved as if “opponent process” processes for things like “motion vs not-motion”, and so on, were falling in and out of calibration.
As subjective data about subjectivity itself (like for trying to figure trying to figure out if Solipsism is true, or whether maybe there is only me plus some Cartesian Demon that is fucking with a hypothetical disembodied mind that is me, with all of external reality as an illusion), hallucinogens did just totally destroy the simple naive hypothesis that “cogito-ergo-sum-style subjective awareness” is independent (not caused by?) what happens in the brain’s firings...
...”the brain” just obviously does cause “the subjective mind”, it turns out...
...unless the Demon’s powers extend to inventing a complex theory of neurology, and taking into account what hallucinogens would hypothetically do to a hypothetically incarnated mind, and then the Demon fed plausible lies along these lines into some hypothetically “metaphysically disembodied awareness” that was me… lol!
((You can’ always generate “an even more paranoid hypothesis”… its just that such hypotheses almost always become negligible under pragmatic anti-paranoid normalization <3))
Anyway. I’m not sure how the idea that “3 second level stuff maybe only happens in the cerebellum and basal ganglia” could be behaviorally applied, to get profits in some way, such as to know the VoI on ruling hypotheses of that class in or out...
But that was a cool part of the interview. Thank you! <3