Notes/ Twitter/TweetThread 1/2
i have never really understood p zombies or qualia tbh trying to imagine a mind where it - senses movement + red light waves in a particular shape in a particular part of its visual field, - does the edge detection and depth perception and pattern matching to a world model to know it is a red cardinal bird 10 ft away in a tree, - simultaneously registers sound waves from approximately that direction which pattern match to "chirping" and are then assigned to the bird as the source of the noise - recalls past experiences with cardinals and, less intensely, other birds and red in other contexts like sweet ripe strawberries and a woman's flushed lips and stop lights and a rose with a thorn, - concludes that the songbird is not a direct physical threat and also updates priors that there's no predator lurking nearby per the bird's sensorium and increases likelihood of nearby ripe berries or seeds - in order to perform the same action as the conscious mind: vaguely smile and otherwise slightly relax muscle tension after initial reflex mild tensing as it first registered the movement, increase likelihood of staying in this location vs moving on, possibly point out the bird to the friend who is hiking next to you - and yet is completely not at all having a subjective experience of redness nor birdness... what IS subjective experience if not all of the sensing, associating, predicting i described which go into producing the same output behavior, as per p-zombie definition?
♥26 likes↻1 reposts↩9 repliescaptured Sep 22, 2026
2/2

one answer i've heard before is to not assume the middle steps are unfolding as i described but that the mind is merely performing a lookup from sensory input to output actions when i imagine the lookup table for the sum of human behavior, the idea of flatly mapping each possible pattern of light + soundwaves + prior context -> set of output actions (+ input to the next lookup) confuses me because of the combinatorial explosion vs the constraint to output the action with identical latency to the conscious mind (the prior sensory info needs to flow either as input to the next lookup or via update to the lookup table to produce the correct identical behavior- the data in a single image != the data in a sequence of images, a human with amnesia != a human with functioning memory who recognizes their friend and remembers the kind words they said yesterday) so then i think, well they must expect their table will be performance optimized for runtime to fit within the physical universe? just a metaphorical table, but underneath it would collapse similar states into wildcards etc? the thing is, as far as i can imagine, to get the scale to fit, you're gonna have to slice your sensorium at the joints- joints like ~birdness and ~redness?