6,000 years have led us to a precarious situation. Complicated supply chains have collapsed spectacularly in the past. The modern premise that we can prop up the complication with datacenters running ML/AI is suspect , as the approach requires extra layers of complication.
The deeper and wider we run, the higher the risk. The concentration of knowledge needed to build systems from core requirements with large third party companies, and the corresponding drain on the expertise at many organizations, means that few can adapt quickly enough. This does not mean that organizations shouldn't use third parties to store, process, and visualize information. They should insist on at least a basic overall design of their information flow, and the models and techniques illustrated in this presentation show one way to do this.