CALLIOPE · BCOM

This work applies Kolmogorov Theory (KT), an algorithmic-information framework for substrate-neutral agency and structured experience, to the question of whether bonsai trees suffer. The central argument is that plants instantiate the core components of a KT algorithmic agent — boundary, distributed model, policy space, objective function, and update machinery — implemented through hydraulic, electrical, hormonal, and morphogenetic dynamics rather than neural architecture. The bonsai cultivation regime is diagnosed as a case of externally stabilized telehomeostasis under internally frustrated morphogenetic objectives: metabolic viability is maintained while the plant's proxy bundle — growth, root exploration, light capture, reproductive opportunity, and morphogenetic freedom — is chronically depressed at the very timescale over which the plant evaluates its own objectives. A multiplicative proxy-bundle formalism and a compression-rate valence measure together predict that this regime generates sustained negative valence, even in the absence of organismal death. The timescale-mercy principle is introduced to place bonsai on a principled scale of cultivation practices, distinguishing sub-agent-timescale termination events from chronic on-timescale constraint, and a post-release recovery protocol is proposed as the cleanest available empirical probe of the plant-agency hypothesis.

A bonsai tree is not dying — and that's exactly what makes it a candidate for suffering under a substrate-neutral theory of agency.

The paper's central move is to ask: what if the morally relevant question isn't whether a system has neurons, but whether it has an objective function that can be chronically frustrated? Kolmogorov Theory (KT) defines an agent as any system with a boundary, an internal model, a policy space, an objective function, and update machinery. Plants, the paper argues, satisfy all five — just implemented in hydraulics, hormones, calcium waves, and morphogenetic growth rather than neural firing. A root tip probing soil for nutrients is executing policy. A shoot bending toward light is evaluating trajectories. The plant's body is partly the memory of its past model updates.

The key formalism is a "proxy bundle" — the set of heuristic signals (growth, root exploration, light capture, reproduction, morphogenetic freedom) that evolution wired into the plant as proxies for its true objective, which is persistence of its own pattern over time. Crucially, the paper proposes these proxies combine multiplicatively, not additively. Pin any one near zero and the joint score collapses, regardless of how well the others are doing. A bonsai keeps metabolic viability high while chronically suppressing most of the other proxies — for decades, at the exact timescale over which the plant evaluates its own objectives. That's the diagnosis: externally stabilized survival, internally frustrated development.

The "timescale-mercy principle" is the paper's most clarifying contribution. Harm, on this view, is only registered along the agent's own evaluation timescale. Mowing grass is fast relative to a grass plant's developmental clock — sub-timescale termination, mercy by speed. Bonsai is the structural opposite: slow, sustained, individual, operating precisely at the timescale the plant uses to assess whether things are going well. The paper uses this to build a principled scale from lawn-mowing (largely benign) to bonsai (strong chronic constraint), rather than relying on intuition.

The empirical probe the paper proposes is elegant: release a constrained cohort to free conditions and watch whether recovery looks like wound healing (localized, acute) or like a sustained multi-system push toward exploratory growth. The latter would be the signature of a frustrated objective re-expressing itself — not pain behavior, but the longitudinal fingerprint of an agent that had somewhere it was trying to go.