Subjective awareness may arise because the brain builds a simplified, incomplete model of its own attention—an 'attention schema.' This incompleteness prevents us from fully understanding how consciousness emerges, which relates to the 'hard problem' of consciousness. Using a mathematical model grounded in classical topology, the paper demonstrates that a complete representation of attention is impossible for any system, human or machine, that monitors its own attention. The argument shows that attention streams cannot be faithfully represented internally, supporting the core claim of Attention Schema Theory: the brain's self-model of attention is necessarily incomplete.
Evolution has selected for inherently unstable biological systems—such as blood pressure, immune responses, and gene expression—that can react swiftly to changing threats or opportunities, but these systems require strict regulation to avoid fatal consequences. Consciousness similarly demands high rates of metabolic free energy to both operate and regulate its underlying machinery, with the stream of consciousness and its boundaries being continually reconstructed in response to dynamic circumstances. The authors develop necessary conditions models using the Data Rate Theorem, which links control and information theories for inherently unstable systems. The synergy between conscious action and its regulation explains the ten-fold higher metabolic energy consumption in human neural tissue and implies a culturally modulated connection between sleep disorders and certain psychopathologies.