Abicumaran Uthamacumaran, Felipe S. Abraão, Narsis A. Kiani, Hector Zenil
A recently introduced approach termed “Assembly Theory”, featuring a computable index based on basic principles of statistical compression has been claimed to be a novel and superior approach to classifying and distinguishing living from non-living systems and the complexity of molecular biosignatures. Here, we demonstrate that the assembly pathway method underlying this index is a suboptimal restricted version of Huffman’s encoding (Fano-type), widely adopted in computer science in the 1950s, that is comparable (or inferior) to other popular statistical compression schemes. We show how simple modular instructions can mislead the assembly index, leading to failure to capture subtleties beyond trivial statistical properties that are pervasive in biological systems. We present cases whose low complexities can arbitrarily diverge from the random-like appearance to which the assembly pathway method would assign arbitrarily high statistical significance, and show that it fails in simple cases (synthetic or natural). Our theoretical and empirical results imply that the assembly index, whose computable nature we show is not an advantage, does not offer any substantial advantage over existing concepts and methods. Alternatives are discussed.

Read the full article at: arxiv.org

Michele Braccini, Andrea Roli, Edoardo Barbieri, and Stuart A. Kauffman

Entropy 2022, 24(10), 1368

Systems poised at a dynamical critical regime, between order and disorder, have been shown capable of exhibiting complex dynamics that balance robustness to external perturbations and rich repertoires of responses to inputs. This property has been exploited in artificial network classifiers, and preliminary results have also been attained in the context of robots controlled by Boolean networks. In this work, we investigate the role of dynamical criticality in robots undergoing online adaptation, i.e., robots that adapt some of their internal parameters to improve a performance metric over time during their activity. We study the behavior of robots controlled by random Boolean networks, which are either adapted in their coupling with robot sensors and actuators or in their structure or both. We observe that robots controlled by critical random Boolean networks have higher average and maximum performance than that of robots controlled by ordered and disordered nets. Notably, in general, adaptation by change of couplings produces robots with slightly higher performance than those adapted by changing their structure. Moreover, we observe that when adapted in their structure, ordered networks tend to move to the critical dynamical regime. These results provide further support to the conjecture that critical regimes favor adaptation and indicate the advantage of calibrating robot control systems at dynamical critical states.

Read the full article at: www.mdpi.com