Shah notes that LLMs work well for text and other unstructured data, but also that they’re expensive to operate, requiring vast amounts of storage compared to their graphical model equivalents.