The characters society is based, on part on neural networks. Neural networks exists in the brain, its how the brain cells connect with each other, how the mechanism of thinking work.
These can of be modeled in computers, not exactly, but close enough to give similar results. When used on a computer neural networks are called activation units.
Traditional computers work serially,no matter how fast they work, the process they compute still happen one after another. For instance the sum (4+5) + (6+8) is worked out as a three step process, first 4+5 are added,then 6+8, and finally both the results are added together.
Neural networks work in parallel, imagine them as loads of independent people working together rather than one person working in steps. In a parallel system 4+5 and 6+8 are worked out at the same time,then the results are added, effectively a two stage process.
If you imagine a computer as a room with a single bloke in it. When information comes in, for instance a picture of a horse, the bloke in the room looks up in a reference book,
step one Animal, vegetable or mineral?
step two Mammal,reptile,insect,bird?
step three how many legs?
step four how big is it.......and so on......
Eventually after all the search criteria has been followed the bloke reduces the information down to being a horse and sends the information out.
Alternatively a neural network could be imagined as a room with loads of blokes in it hanging around. When information comes in, the picture of a horse, something different happens. Each one of the blokes is an expert in a different field. One is really good at identifying colours, one is really into animals, one is really into the number four, one recognises sizes. A bit like when people can identify trains by their sound,or remembers all the different types of car. All at once the fellows recognise different aspects of the horse and come to a consensus weather the legs belong to a horse rather than a chair, as the animal guy knows its a horse, and the colour guy knows its brown, and so on. The knowledge is shared out to different points and processed.
A bit like how knowledge is shared by many different people over the (idealized) internet,no one is an expert but using knowledge as a collective force we can pool our resources and find out more stuff.
I can see parallels here on how we organise our lives, shops sell one type of product, if you need specialist advice to go to a specialist shop, not the supermarket,where it may be available but doesn't have the knowledge base. If specialist shops get to diversified the specific knowledge about one thing gets diluted. Thats why polymaths are not all that common, if people specialise in two subjects they are generally related.
Train spotters, English civil war enthusiast, real ale experts all reflect knowledge nodes, that when seen in isolation as sad enthusiasts but seen together as a solid knowledge database of the world as a whole.
We are a parallel species father than a serial species, we need the farmer, the farmer needs the industrialist.
The characters could be seen as very idiosyncratic neural network nodes. They have a knowledge base but either the knowledge base is wonky (the character that just wants to collect all the objects in the world, without knowing what they are) or what they do with it is wonky (such as the shopper character who looks up products and relates them together by weight rather than type of product). We follow the characters as they negotiate through life through using their knowledge base