Granted, I feel like munging gigabytes of text data (i.e. G, A, T and Cs) would be something LLMs would be good at