What can AI learn from bionic algorithms?: Comment on "Does being multi-headed make you better at solving problems? A survey of Physarum-based models and computations" by Chao Gao et al.

Physarum polycephalum (literally, multi-headed slime mould) is a multinucleated, unicellular organism that belongs to the protoplast mucus of amoebina. Physarum is increasingly popular in diverse fields including biophysics, evolutionary computation, bioengineering, intelligent algorithms [1–8], because of its striking high-level of biologically intelligent behavior which was first reported in 2010 [9]. For example, inoculated in a maze of corridors on the agar surface with food resources placed at the two terminals of the maze, Physarum polycephalum is able to automatically detect the shortest path along which a protoplasmic pipeline will be formed to connect the food at the terminals [10]. Surprisingly, this is a self-organized process without centralized control. Gao et al. [11] reviewed the latest progress of Physarum-based models and computations. Through a systematic review of publications in the Web of Science, they constructed a network of scientific citations to overview the hot research areas. Their major interests lie in the computational models inspired by the two fundamental features of Physarum’s foraging behavior, i.e., extension and retraction, which are applied as the morphology, taxis and positive feedback dynamics in the top-down and bottom-up modeling techniques. They also surveyed some real-world applications based on the core features of Physarum for solving difficult computational problems. Furthermore, they outlined recent advancements in bionic algorithms that are grounded in the bio-intelligence of Physarum polycephalum.