Advances in materials design and discovery have been made in the last few years with machine learning combined with large databases on materials properties. The creation of generative large language models such as ChatGPT is now bound to revolutionize knowledge discovery in materials sciences, which will go well beyond the current achievements in materials discovery. In this lecture, a discussion will be presented of the technologies involving high-throughput experiments and computer simulations exploited in materials discovery, in addition to the proposal of novel approaches to mine scientific literature in materials. The latter approaches encompass natural language processing and network science with which one may obtain the landscape of research on given topics or even scientific journals, and identify materials and processing conditions for targeted applications. For instance, with such methods one may determine the most impactful topics in materials sciences are associated with energy-related materials and organic electronics. Furthermore, tools based on large language models and other artificial intelligence methods permit the development of computer-assisted diagnosis systems for personalized medicine, precise agriculture and different types of automated surveillance. This will be achieved via deep learning to leverage multimodal data from distinct sources, e.g. text, scientific data, images and videos.