← Back to Program

Qi Zhang

Independent, San Jose, CA, USA
Tuesday, October 13 · Poster Session 1 · Mission Bay Room
Poster
Exploring the Origins of Subjectivity, Awareness, Multiple NCCs, and the No-Report Paradigm Effect with an Artificial Intelligence Architecture

Based on an executable artificial intelligence system, this study presents a system-level cognitive processing model, named the Twin Cognitive Cycle (TCC), together with its computational simulations, to explore the origins of subjectivity and awareness, and how they are connected to the empirical findings of multiple NCCs and the no-report paradigm effect. Subjective experience (qualia), the existence of multiple NCCs, and the no-report paradigm effect can be viewed as different manifestations of the same underlying nature of consciousness and its associated cognitive processes. Therefore, a satisfactory understanding of any one of them is expected to explain them all. The TCC model proposes that the brain's cognitive process is a recurrent loop consisting of two phases: feedforward and feedback (or bottom-up and top-down). The feedforward phase is characterized by a chain of input-specific activations, whereas the feedback phase involves a global activation triggered simultaneously by feedback signals from the central executive. In the TCC model, consciousness is associated with a unique pattern of global activation. The total number of unique patterns can be formulated as the product of two distinct combinatorial systems: a constrained macro-distribution of common features and an unconstrained exponential micro-state space of variables spanning the underlying modality space. This product leads to uncountably infinite unique variations; therefore, every conscious state is inherently unique and entirely subjective. The resulting activation pattern is, literally, the system's echo of the feedback signal, and the portion that can be felt serves the function of evaluating the feedback, which is a necessary condition for reiterating the TCC. Upon receiving an external stimulus, a TCC undergoes up to five waves of activation across four cognitive regions in order to generate a semantic report. These sequential activations align well with ERP waveforms observed in NCC studies and are consistent with the concepts of NCC-prerequisite and NCC-consequence, as well as content-specific NCCs and the full NCC. The executable system can be configured to reproduce activation patterns similar to those recorded by fMRI under both report and no-report conditions by either allowing or blocking information from entering the central executive during the feedforward phase.