If it is news to you, I would like 10 minutes of your time to tell you a bit about #VectorEmbeddings and #VectorDatabases as part of my weekly theme exploring #RetrievalAugmentedGeneration or #RAG (please forgive some past errors where "aided" was my "A" word). The #LargeLanguageModel component will handle the generation, but #LLM is not a great tool for the retrieval. For some problems, vector embeddings are the answer.
I think it is a little dangerous to set out in pursuit of #ArtificialIntelligence. It is much better to have specific goals, and RAG is a technique that fits very well onto some very common megacorporate problems. Specific goals are good because they help spotlight specific tasks that add up to success. Shopping for a vector database is one such task that might turn out be very relevant to you but you will probably not see in the headlines like LLM.