Excellent. Thank you, I shall get the book. I presume you are familiar with the work of Pentti Kanerva, also the work of Jeff Hawkins. Put the three together and we have the basis of AGI.
The idea of "change in prominence" is very interesting. When does "prominence" of an attribute change? If a group of biologists decided that whales are mammals now, and the rest of us had to now categorize them as such, would we each find our own unique attributes to create prominence? What is the "shared understanding" of these attributes when it comes to categorization, is the interesting question; and how do we socially affect the categorization of other people here.
I find fascinanting your question. I think the suggested model is very useful, but it says nothing about who does the observing (let's call them the Observers). In semiotics for instance, the Observer has a key role, because is the Observer the one that does the encoding/decoding of signs in a domain (concepts in this video). So in my mind, the modeling of Observers within specific domains need to be taken into account to fully explain the prominence of attributes, because they emerge by the action of Observing the underlaying structure by a specific Observers group.
This is quite cool, and I feel like it needs to figure out how to handle hierarchical structures - for example in recognizing a duck, we probably make use of the fact that it has feathers, but its feathers are correlated with its shape/color, and there are other intermediate features so it's a whole hierarchy of them; also in parsing our raw visual input we first need to figure out which are distinct objects - we can't classify the species of an animal until we know that it's distinct from the carrot it is holding or the hat on its head, but in order to determine that they are distinct, we use some higher-level information like the types of species which we know - if it would clearly fit a duck holding a carrot with a hat on its head then that might be a good fit for what we're seeing, but if it would even more clearly fit an animal holding a carrot and the hat is a part of its head then that doesn't look like any animal, so that hypothesis can't be right - you see how you just need to have some hierarchy?
When a Person gets brain damage in the Dorsolateral prefrontal cortex which controls planning and conceptual thinking that person is the most gullible person. You tell them to do what ever you tell them and they will believe and do what you say. That's because that person lost his or her ability to think and to plan so that person will always look for someone to give them an answer and to plan for them.
