The Informational Argument for Conservation

In the AI Story

The argument proceeds as follows. First: biological diversity is information — specific configurations of genetic sequences, protein structures, metabolic pathways, and behavioral repertoires that encode evolutionary solutions to environmental challenges. Second: this information was produced by a computational process — evolution — of vastly greater scope, duration, and power than any process human beings have constructed. Third: the information cannot be reconstructed. The specific configurations that encode it exist nowhere else and would require the same four billion years of evolutionary computation to reproduce. Fourth: the current rate of extinction is permanently destroying this information faster than human research can extract it.

The AI era intensifies the argument in a way Wilson only partially articulated before his 2021 death. Large language models and other AI systems generate insights by recombining information from the human corpus. This is extraordinarily powerful and fundamentally limited. The human corpus is a tiny, recent, heavily biased sample of the information the biosphere contains. The biosphere's information represents solutions to problems human beings have not yet learned to formulate. As AI tools become more capable of processing information, the marginal value of irreplaceable information — information that cannot be produced by any human or computational process — rises dramatically.

The argument also specifies where conservation investment should concentrate. Habitats containing high biodiversity represent higher information density. Organisms with ancient evolutionary lineages represent longer computational histories. Endemic species — organisms found nowhere else — represent unique information. Generalist species with broad distributions represent redundant information. The argument does not imply that all biodiversity is equally valuable; it specifies the dimensions along which informational value varies and suggests that conservation priorities should be calibrated accordingly.

The strategic implication is that conservation and AI are not competing priorities. They are coupled. A civilization that invests in AI capability while allowing biodiversity to collapse is destroying the substrate from which AI's most valuable inputs will eventually need to come. A civilization that preserves biodiversity while failing to develop the tools to extract and utilize its information leaves the library intact but unread. The consilient strategy invests in both — AI capability to read the library faster, biodiversity preservation to ensure that the library survives to be read.

Origin

The informational argument appears across Wilson's late work, most explicitly in The Future of Life (2002) and Half-Earth (2016). Related arguments have been developed by bioprospecting advocates, by information-theoretic ecologists, and by Wilson's student Bert Hölldobler.

Key Ideas

Information is the frame. Biodiversity is informational infrastructure; conservation is information preservation; extinction is information destruction.

AI intensifies the argument. More powerful tools for processing information make irreplaceable information more valuable, not less.

Priorities follow information density. Ancient lineages, endemic species, and biodiversity hotspots carry more information per unit than recent, cosmopolitan, and species-poor habitats.

Conservation and AI are complementary. The coupled strategy invests in both the capacity to read the library and the preservation of the library itself.