Qeios · Jul 17, 2026
Cybernetic Attractor Decay: A Regulatory-Fidelity Framework for Aging
By Bhak J
Biological aging follows predictable molecular trajectories across individuals and species. This predictability arises despite stochastic biological processes. The same trajectories can be reversed by developmental reprogramming. It is paradoxical that these properties coexist. Pure stochastic-damage models do not accommodate the predictability. Programmed-senescence models do not accommodate the reversibility. To resolve this paradox, I propose a framework called cybernetic attractor decay (CAD). Under CAD, aging clocks measure the predictable drift of developmentally established regulatory architectures. At the core, they do not measure damage load or read out a death program. The attractor is defined as the joint stable configuration of CpG methylation, chromatin state, transcription-factor occupancy, and feedback topology that maintains a cell-type-specific regulatory state. Gerostasis is the maintained, low-drift state of that attractor, and the regulatory condition that interventions aim to preserve or restore. Gerotype is the measurable aging phenotype of a cell, tissue, or regulatory network at a given biological age. The gerotype is the operational expression of attractor decay. It plays the role that catalogs of cellular features (such as “hallmarks”) play in other accounts but is tied directly to regulatory state rather than to a list. Computational drift is operationalized as age-associated loss of regulatory precision, measurable as transcriptional, methylation, and chromatin-accessibility dispersion in single-cell data. The empirical face of the gerotype is the rise of Non-Requisite Variety at the expense of Requisite Variety. The framework integrates several existing accounts: quasi-programmed senescence and hyperfunction[1]; the infrastructure-versus-specialized-gene partition[2]; loss of dynamical complexity[3]; and the complex-systems approach[4]. Each is treated as one axis of a broader landscape of regulatory-fidelity loss. CAD makes testable predictions that distinguish it from pure stochasticity accounts. These include tissue-specific signatures of drift, preserved attractor topology in negligibly senescent species, and coordinated system-wide attractor collapse in semelparous organisms such as Pacific salmon, which die within days to weeks of spawning.