Conscious experiences feel rich and hard to fully describe or recall, a puzzle that partly motivates the explanatory gap—the belief that consciousness cannot be reduced to physical processes. This work offers an information-theoretic dynamical systems framework: richness corresponds to the amount of information in a conscious state, and ineffability to information lost during processing. Attractor dynamics in working memory cause impoverished recollections, language's discrete symbolic nature cannot capture high-dimensional experiential structure, and similar cognitive function between individuals improves communicability. The model advances a physicalist explanation of these puzzling aspects, though it may not settle all questions about the explanatory gap.
A taxonomical framework classifies challenges to the possibility of consciousness in digital AI systems by their level of analysis (corresponding to Marr's levels) and their degree of force: degree 1 challenges computational functionalism without ruling out digital consciousness, degree 2 suggests improbability without impossibility, and degree 3 argues strict impossibility. The framework is applied to 14 prominent examples from the literature. The aim is to disambiguate between challenges to computational functionalism and challenges to digital consciousness, and between different ways of parsing such challenges, without taking a side in the debate.
A method for assessing whether AI systems might be conscious is presented, drawing on existing neuroscientific theories of consciousness. The approach involves deriving indicators from such theories to inform beliefs about AI consciousness. This method can make progress because computational functionalist theories, which are influential, have empirically testable implications for AI. The work does not claim that any current AI is conscious but outlines a rigorous framework for future assessment.
No current AI systems are conscious, but there are no obvious technical barriers to building ones that might be, according to an analysis grounded in neuroscientific theories of consciousness. The report surveys prominent theories—recurrent processing, global workspace, higher-order, predictive processing, and attention schema—and derives computational indicator properties from them. Applying these indicators to recent AI systems yields no evidence of consciousness, but the authors argue that future systems could potentially implement the necessary properties.