Z*-numbers: Augmented Z-numbers for machine-subjectivity representation

Envisaging a futuristic environment of man–machine and machine–machine synergy, this article documents our research on the augmented Z-numbers, the Z∗-numbers, for machineperception encapsulation. The Z∗-numbers have been envisioned as operands of endogenous machine-mind processes underlying bespoke comprehension of the real world. Besides information-certainty, as in a Z-number, a Z∗-number incorporates context, time and affects as essential factors of subjectivity representation. We have proposed: (a) definitions for certainty and affect parameters—arising out of socio-cultural influences on machine-knowledge, (b) a Z∗-number based rudimentary procedure for natural-language comprehension emulation, and (c) primitive perception-operators for ‘machine-mentalese’ simulation using Z∗-number information-equivalents. Our work draws from non-symbolic theories of cognition and ‘mindfulness’, human-mind processes—studied through behavioral experiments, and theories of the ‘self’ and ‘qualia’. The article includes detailed discussions of these experiments and consequent insights, analysis of a theoretical run-through of the defined procedure, and correspondence-studies between the Z∗-number paradigm and philosophies of the self. Our research raises questions on cognitive biases and autogenous mind-processes that highlight crucial practical challenges in the current realization of a synthetic-mind. All ideas herein aim to contribute to studies on the ‘self’ and its machine-embodiment for the synthesis of an empathetic machine-mind. © 2015 Elsevier Inc. All rights reserved.