Академик РИА Pan Wang, Jiafan Zhuang, академик РИА Zhun Fan "Bionic Computing — Shell or Core?"

Академик РИА Pan Wang, Jiafan Zhuang, академик РИА Zhun Fan "Bionic Computing — Shell or Core?" Bionic Computing — Shell or Core? International Joint Research Institute of Wuhan University of Technology - Russian Academy of Engineering, China Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen, Guangdong, China wangpan@whut.edu.cn, 1574586352@qq.com, zhun_fan@126.com This paper thoroughly discusses some aspects of bionic computing, a new branch of bionics. Compared with other branches of bionics, the "bionic" significance of bionic computing is far from sufficient. In short, current bionic algorithms do not live up to their names in the following areas: (1) Environment. It can be said that bionic algorithms differ greatly from the environment in which the simulated substances (organisms) in nature (or the biological world) are located. Bionic algorithms are often a mapping space in the computer environment, and there is almost no comparability with the real natural environment. On the other hand, some other branches of bionics, such as bionic robots, bionic materials, etc., are very close to the real natural environment, even immersing oneself in it (such as the Harrier fighter jet soaring in the blue sky like a bird, bionic underwater robots being underwater like fish; electronic police dogs being placed in the environment of odor detection together with animal dogs, such as operating rooms, warehouses, airports, etc.). Therefore, the rationality and scientific meaning of functional imitation in fundamentally different environments are highly questionable. (2) Goal. The goal of many bionic calculations is optimization, while the primary goal of growth and reproduction in the biological world is adaptation (or survival). Therefore, there are significant differences between bionic computing and biology at the target level. Bionic computing with optimization as its primary and important objective, drawing on or emulating species in the biological world (more precisely, nature) with adaptation as its objective, may result in significant discounts in problem-solving efficiency (as optimization may only be a secondary objective for organisms), and even worse, sometimes it is difficult to achieve significant improvements in problem-solving quality. (3) Methodology. In terms of methodology, positioning bionic computing as mimicking the computational methods of the biological world (nature) is debatable. In fact, bionic computing mostly uses analogies and borrowed methods: analogies to certain processes in the biological world, borrowing biological terms. In essence, there are not many similarities between the two. Viewing the optimization process (or decision-making process) as a biological process such as evolution is subjectively imagined or analogized, and the two are not completely separate in terms of implementation and specific operations; And borrowing biological terms is a subjective imagination replay based on the unrealistic analogy mentioned above. In addition, many bionic calculations are based on outdated or controversial biological foundations, which are relatively superficial. Finally, the authors discussed on how bionic computing can move from empirical science to rigorous science.