It’s fascinating how in a few years, large language models (LLM) went from intriguing new deep learning models (the transformer architecture) to one of the hottest areas of AI research.