Towards A Computational Model Of Consciousness: Global Abstraction Workspace
Halim Djerroud, Arab Ali Chérif
Zenodo (CERN European Organization for Nuclear Research) November 1, 2016 preprint DOI: 10.5281/zenodo.1128246 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper |
|---|---|
| Citations | 1 |
| Key points | Proposes that conscious and non-conscious mental functions are produced by biologically adaptive algorithms, and that a computational model called the "Global Abstraction Workspace"—an internal environmental model treated as a multi-agent system—can generate new data, processes, and actions, offering a way to model consciousness. |
Abstract
We assume that conscious functions are implemented automatically. In other words that consciousness as well as the non-consciousness aspect of human thought, planning and perception, are produced by biologically adaptive algorithms. We propose that the mechanisms of consciousness can be produced using similar adaptive algorithms to those executed by the mechanism. In this paper, we present a computational model of consciousness, the ”Global Abstraction Workspace” which is an internal environmental modelling perceived as a multi-agent system. This system is able to evolve and generate new data and processes as well as actions in the environment.