| Beschreibung |
In many highly specialized industries, a generational shift is taking place in which expertise is being lost because experienced colleagues are retiring without passing on the knowledge they have acquired over the course of their long careers. Obtaining this knowledge from experienced colleagues is usually a costly, manual process due to the knowledge acquisition bottleneck: experts are busy, and much of what they know is tacit knowledge, so ingrained through experience that even they struggle to put it into words. In this project, we aim to explore approaches that automate knowledge acquisition by involving experts in conversations, creating models of the acquired knowledge, and processing this knowledge in a way that allows others to learn from it. We will use large language models to develop automated interlocutors that are designed to ask the right questions to fill knowledge gaps and to compile appropriate learning materials based on the knowledge acquired. |
