CONCEPT
Cognitive Bioaccumulation
The process by which sub-threshold cognitive harms from AI tool use compound across exposures, trophic levels, and time—each interaction depositing a thin layer of non-understanding that is undetectable in isolation and lethal in aggregate, exactly as DDT accumulated through food chains.
Rachel Carson’s central technical insight—that a substance safe at any single dose can be devastating in accumulation through a food chain—applies to the cognitive effects of intensive AI use with a structural precision that is not metaphorical but mechanistic. No single AI interaction is harmful. A developer who uses an AI coding assistant to generate a function has saved time and produced working code. A student who uses AI to explain a concept has received a clear explanation. Each interaction, evaluated in isolation, is beneficial or neutral. The label concentration is safe. But the developer does not use the tool once. She uses it all day, every day, for months. Each interaction that eliminates a debugging encounter deposits a thin layer of non-understanding—not ignorance, but the specific absence of the knowledge that would have been built through the
friction of the process. The layers do not wash away. They
accumulate, and after six months