From autopoiesis to self-optimization: Toward an enactive model of biological regulation

Abstract

The theory of autopoiesis has been influential in many areas of theoretical biology, especially in the fields of artificial life and origins of life. However, it has not managed to productively connect with mainstream biology, partly for theoretical reasons, but arguably mainly because deriving specific working hypotheses has been challenging. The theory has recently undergone significant conceptual development in the enactive approach to life and mind. Hidden complexity in the original conception of autopoiesis has been explicated in the service of other operationalizable concepts related to self-individuation: precariousness, adaptivity, and agency. Here we advance these developments by highlighting the interplay of these concepts with considerations from thermodynamics: reversibility, irreversibility, and path-dependence. We interpret this interplay in terms of the self-optimization model, and present modeling results that illustrate how these minimal conditions enable a system to re-organize itself such that it tends toward coordinated constraint satisfaction at the system level. Although the model is still very abstract, these results point in a direction where the enactive approach could productively connect with cell biology.

Irruption Theory: A Novel Conceptualization of the Enactive Account of Motivated Activity

“…In addition, the enactive approach has long emphasized that life is a process ended becoming across all scales [12], and this historicity is realized by path-de plasticity of a living system's parameters, variables, and constraints [13]. In order light the role of this plasticity for the enactive conception of life, Froese, Weber, S and Ikegami [11] redrew Figure 2, originally by Di Paolo et al [9], by including p as an explicit aspect of the way in which agency resolves the primordial tension v dination of partial constraint satisfaction over time (Figure 3). Starting from an autopoietic conception of life (lefthand side), the enactive approach highlights the intrinsic precariousness associated with life's thermodynamic embodiment.…”

“…Thus, life requires agency; it must actively coordinate the partial satisfaction of one and the other of its needs over time. Figure taken from Froese and colleagues [11], redrawn from a figure by Di Paolo and colleagues [9].…”

“…On the right, with the addition of unsupervised associative learning, the system begins a process of self-optimization: over time, the network re-organizes its relations such that it generalizes over past equilibria to reach otherwise unvisited, deeper equilibria. Figure copied from Froese, Weber, Shpurov, and Ikegami [11].…”

“…The self-optimization model can also be combined with self-organized criticality such that it has an intrinsic mechanism for the "reset" of neurons to arbitrary states across scales [106]. Given the potential generality of the self-optimization model-essentially, a system that is iterating between equilibrium and far-from-equilibrium dynamics plus some historicity akin to a principle of precedent-Froese and colleagues [11] proposed this to be a fitting model of the notion of adaptivity described by the enactive conception of life. In abstract thermodynamic terms, the living system's need for openness to the environment's energy flows corresponds to a far-from-equilibrium mode of operation (open system), while the system's need for closedness for its structural integrity corresponds to an equilibrium mode (closed system).…”

On motivating irruptions: the need for a multilevel approach at the interface between life and mind

“…As mentioned, the main task of explaining how motivated behavior is non-random even if it is materially underdetermined is placed in the third thesis of attunement . Although Froese (2023) mentions various promising models that could ground and further develop the attunement thesis, such as the embodied approach to habits (Ramírez-Vizcaya & Froese, 2020), implicit body memory (Froese & Izquierdo, 2018), and meta-stable attunement (Bruineberg et al, 2021), he favors the self-optimization model of adaptivity and regulation (Froese et al, 2023). The latter proposes that a system, such as a neural network, can spontaneously optimize its internal organization to better satisfy its constraints over time (i.e., solve a “constraint satisfaction” problem).…”