The real answer is that we don't yet know enough about how the brain works to work effectively on this problem. We don't know what questions to ask or how to break down the problem into smaller problems.
We may get there. Read something about how vision works from a century ago, when nobody had a clue. The first real progress came from "What the Frog's Eye Tells the Frog's Brain" (1959).[1] That was the beginning of understanding visual perception, and the very early days of neural network technology. Now we have lots of systems doing visual perception moderately well. There's been real progress.
(I went through Stanford CS at the peak of the 1980s expert system boom. Back then, people there were way too much into asking questions like this. "Does a rock have intentions?" was an exam question. The "AI winter" followed. AI finally got unstuck 20 years later when the machine learning people and their "shut up and calculate" approach started working.)
Personally, I think the brain as a data processing unit is a model that's seriously limiting the way many philosophers and neuroscientists think about this problem. You may be able to correlate an individual's reported experience of periwinkle down to the exact number of potassium ions crossing the cell membrane of every relevant neuron upon there acknowledgement of the color, but you still will know nothing about why it feels like something to see periwinkle.
We may get there. Read something about how vision works from a century ago, when nobody had a clue. The first real progress came from "What the Frog's Eye Tells the Frog's Brain" (1959).[1] That was the beginning of understanding visual perception, and the very early days of neural network technology. Now we have lots of systems doing visual perception moderately well. There's been real progress.
(I went through Stanford CS at the peak of the 1980s expert system boom. Back then, people there were way too much into asking questions like this. "Does a rock have intentions?" was an exam question. The "AI winter" followed. AI finally got unstuck 20 years later when the machine learning people and their "shut up and calculate" approach started working.)
[1] https://hearingbrain.org/docs/letvin_ieee_1959.pdf