Скачать или смотреть Metagov x Future of Science Seminar - Scaling Digital Support Systems

Abstract: This presentation argues that digitization has been highly disruptive to decision support systems. Using a generalized three-stage architecture, this presentation shows how uneven digital automation across the system's stages creates severe scaling challenges. Early applicability of digital automation to the initial information processing stage induced a bottleneck in the second, largely manual, knowledge processing stage, which organizations have tried to solve by hiring massive teams of analysts. Recent advances in generalized computation, such as Large Language Models (LLMs) and deep convolutional neural networks (DCNNs), offer a path to automating this knowledge work. The presentation concludes by predicting that this digital augmentation of knowledge processing will ease the current bottleneck in partially digitized decision support systems, which has the potential to fundamentally alter the competitive landscape, enabling smaller, more agile organizations to achieve a level of decision-making sophistication that was previously the exclusive domain of tech giants.