Limitations of Information Processing in Top-Down Organizations

Every intelligent agent’s information processing capacity is limited, constrained by the memory, time, and energy required to absorb and process information. Although human intelligence is a product of connectionism, human limitations in memory, time, and energy impose significant constraints on information processing. In organizational management, individuals must assess their own and others’ information processing limitations and devise efficient information processing schemes (making the most efficient and valuable decisions with limited time and energy). Traditional top-down organizations, rooted in mechanism and cybernetics, face limitations as noted in Fundamentals of Machine Intelligence: “Cybernetic systems require standardized judgment and decision-making rules, which are typically ill-equipped to handle complex real-world scenarios with diverse data types and large data volumes (numerous nonlinear features), making them inherently unsuitable for informal, nonlinear contexts.” Moreover, cybernetic systems cannot improve performance by increasing the volume of data.

Traditional organizations (cybernetic systems) often separate decision-making and execution behaviors, with some individuals responsible for applying information to make decisions and others executing the provided paths. This separation creates another issue: decision-making individuals use information, while execution generates it, meaning executors often generate and provide information. However, executors’ decisions are driven by maximizing their own rights transformation efficiency, so the information they provide may be biased. The separation of decision-making and execution concentrates decision-making among a few individuals, increasing their workload as they collect potentially biased information from executors and process all or most of the organization’s information within limited time and energy. Their decisions also aim to maximize personal rights transformation efficiency, leading to issues of biased information sources, decision-making biases, and information processing capacity limits.

Given these factors, cybernetic organizations face significant information processing limitations, with greater decision-making concentration exacerbating risks. However, when information volume is low (small organizations, simple operations, or less information-heavy historical periods), cybernetic organizations can still perform well. As information volume and complexity increase, these organizations may face developmental and survival challenges.