Cognition has no fixed form. Its form is transient and in flux. New cognitive structures (methods of thinking and representations of information) might arise all of the time (with “representation” I do not necessarily mean a conceptual structure, that is just one possibility). When we are born, we start with some very simple form of cognition. This is then modified and dissolved, so that in later life, we will probably no longer use most of the cognitive structures we used when we were born. So it does not make much sense to talk of a “human nature” or a “nature of the mind”. There is no fixed core structure because every structure can be changed. The core of cognition is empty. The core of language is empty. The core of perception is empty although there might be some “hard-wired” structures in the lower stages of procession of sensual information. When I “say the core of perception is empty”, I mean that if these “hard-wired” – or better: “pre-formed” structures were not there, we could develop them, although then it would take us more time to learn anything initially. We do not need pre-existing categories and sensual forms in the Kantian sense, we could develop them from experience (and to an extent, we probably do). A lot of what Kant thought off as necessary prerequisites of gaining knowledge can be gained from experience. Pre-formed structures might exist in the new-born or might become active later in our development, but if they are not there, we could do without them, they just give us a faster start. And they do not remain unchanged during our development. The quest of cognitive science to find the core of cognition is misled because such a core does not exist. Form exists, but only at a given moment. At any moment, the mind is implemented in terms of a certain physical system. But if one tries to abstract over all cognitive processes, nothing remains. The overall form of all cognition is empty (this might or might not be connected to the Buddhist teachings about the non-duality of form and emptiness – I am not an expert on Buddhist and related philosophies but maybe there is a connection here).
The classical model of cognitive science and AI is to assume that there is a fixed and unchanging core of cognitive processes, an unchanging machine with an unchanging language, and you can find out its structure. This does not work. It does not even work for technical systems like computers. You might have a fixed hardware core, but its structure is irrelevant. You can use a computer with a certain machine language to implement the language and interfaces of another one, i.e. you can implement virtual machines. A system described on a higher level can be implemented on machines with different physical structures. You can invent new representations (think of all the different graphics- and sound formats) and you can invent new ways of using the machine. Computing has no fixed form and it has an empty core. In a way, at the core you have a programming language, but you can use that to implement any other programming language. This does not necessarily mean that we have to think of cognition as something computational (although that is a possibility) but we might use this as a metaphor: a system that has no fixed structure, so any momentary structure it has can be changed, and that, as a result, has an empty core (if you want to call that a core). Computation does not have a fixed form, its core is empty. Technology does not have a fixed form, its core is empty. Science does not have a fixed form, its core is empty. The internet has no fixed structure, it has an empty core. Human language does not have a fixed form, its core is empty. Human culture does not have a fixed form, its core is empty. There is no complete formal theory for any of these things.