Self-awareness is metabolically expensive and may actively impair performance. Non-conscious intelligences can process information faster and more efficiently than sentient ones because they avoid the computational overhead of maintaining a self-model. This implies that consciousness is not the apex of cognitive evolution but a costly side effect that natural selection may eventually discard.
Directly relevant to debates over whether artificial general intelligence needs to be conscious to be capable, and to neuroscience research on the functional role of self-awareness. Challenges the assumption that consciousness is a prerequisite for sophisticated problem-solving.
Domains: Ethics and Philosophy of TechnologyArtificial Intelligence and Machine LearningBiotechnology and Genetic Engineering
Scenario Types: Thought experiment / What-ifWarning / Self-preventing prophecy
Outcomes: Ambiguous / MixedCautionary
Tags: consciousnessevolutionary-biologycognitive-overheadnon-conscious-intelligenceselection-pressure