The authors argue that evolution has shaped human brains in our ancestral selection environments to prefer modes of organizing ourselves for group activities that are “nearly decentralized”- minimizing differences in the influence each member has on the choice and implementation of designs, yet tolerating some asymmetry to exploit unique strengths and produce convergence in group behavior. They posit that this process of Nearly Decentralized Organizing (NDO) is an attractor in the space of possible design processes, which appeals intuitively even to modern humans; all else being equal, the attractor describes how we would like to organize ourselves. This is an explanation for the disproportionate (relative to economic significance) mindshare occupied by contemporary attempts to organize in a decentralized manner, such as non-hierarchical firms, open-source communities, and decentralized autonomous organizations (DAOs).However, the attractor by itself is by no means a blueprint, as it may be unsuitable to produce designs that meet modern organizational goals. The authors propose that the central challenge for human-centric organizing is to make decentralized organizing effective for large, complex and diverse groups (rather than fix the problems created by centralized organizing). They discuss the role of algorithmic technologies in this endeavor.