Yes, as I noted when you first introduced it to us, I think it is a rich idea. And I also agree that it is indeed most likely this neurophysiological process—the dynamic integration (assimilation) of multiple modes of sensory input—from which phenomenal consciousness "emerges." (Incidentally, I would argue this is not a causal relationship in the usual sense.)
Coincidently, the paper
@Constance and I have just read offers a similar explanation of the emergence of phenomenal consciousness:
http://kognitywistyka.umcs.lublin.pl/wp-content/uploads/2014/04/Zlatev2009-CS.pdf
"Still, a basic form of “evaluation” can be performed even without any consciousness (i.e. phenomenal experience), on the basis of a biological value system, as in Damasio’s (2000) somatic marker theory, where “emotions” are defined as bodily states outside of awareness, and not as experiential ones. Some simple forms of learning may be performed on this basis, by connecting “features” of a disjoint Umwelt with such bodily states.
However, “binding” the Umwelt into coherent multimodal wholes – objects, scenes and situations – would be functional for behavioural flexibility, learning, anticipation and “problem-solving”, which are all necessary for navigating in a complex environment (Edelman 1992).
The hypothesis that I am proposing is that consciousness emerged as a biological adaptation in creatures in need of a “common currency” for multimodal perception, action and evaluation, so that attentional resources can be allocated flexibly, and evaluation can be performed efficiently via feelings, e.g. for the purpose of anticipating the results of actions. Thus, (part of) the Umwelt becomes transformed into a Lebenswelt, perceived as separate from the acting and feeling subject. This on its side would lead to the pre-reflective self-consciousness of the minimal self, and to what Husserl called “the correlational structure of intentionality”: intentional objects are perceived as external to the self, but are simultaneously categorized and “felt” on the basis of internal phenomenal value systems. This, I believe, is what is meant by concepts such as “core consciousness” (Damasio 2000) and “primary con- sciousness” (Edelman 1992).
The advantage of such a view is that consciousness is clearly functional (cf. Donald 2001), and ceases to be a mystery for evolutionary theory.
As mentioned in Section 3.2, there are even empirically grounded proposals for the neural bases of this adaptation: the widely distributed and interconnected thalamocortical system present in mammals, but much less developed in e.g. reptiles. As is generally the case in evolution, consciousness even in its “core” and “primary” forms hardly emerged wholesale. But since even a minimal form of what was sketched above would have been adaptive for animals inhabiting a complex Umwelt, it would have been selected for in our distant ancestors, and its evolution into “higher” forms, e.g. with more flexible control of attention, more diverse feelings etc., under way. "
The integration (assimilation) of multiple modes of sensory signals (input) into a unified medium (common currency) is a prime candidate for phenomenal experience.
Pharoah you ask: "How do neural networks assimilate all of these [sensory] inputs?"
As you note, this is a problem neuroscientists are certainly working to answer. But in the meantime, we can ask: If phenomenal consciousness does indeed emerge as an ontologically new phenomenon as a result of the process of sensory integration, how do we characterize this emergence?
(1) The emergence of p-consciousness is ultimately the result of neurological processes, or,
(2) The emergence of p-consciousness is ultimately the result of signal integration processes?