Suppose that I were to write a program that decided whether something was X, that program would return YES or NO. That program is what computer scientists call a decision function.
So I ask my friends Bob and Joe to also make programs to identify data
Joe makes Is_Picture_A_Rose(Picture) and Bob makes Is_Picture_A_Dog(Picture)
So we have a “parallel process of creating analysis functions”.
However we can’t just add the results statistical significance, for example, I have 10 pictures, 3 of them are flowers and 2 flowers are roses.
I can’t get impressed with getting a positive for roses(20%) if I already got a positive for flowers(making the chance 66%).
That means we need a “process of determining which tests affect which tests”.
I think schizophrenia is caused by the two processes not working well together, you have too many tests for too little knowledge of their interoperability, hence you get lots of false positives(people look like other people, places like other places).
The medicine probably works by slowing down the first process(creation of analysis tools). You could increase the second processes effectivity by being exposed to:
diverse data(not getting obsessed with a single subject/place)
high quality dreaming(ear plugs, country side, nothing happening in the vicinity of the room, long time of dreaming)
The relation between finding patterns and the immune system could explain why schizophrenics also have false positives(signs of infection) in their blood.
If we could find a drug that increased the dreaming phase of sleeping or somehow helped the second process I wrote about we could get closer to a “cure”.