Data Scientists already in the industry more than a couple of years, are you actively learning LLM frameworks? If so, how? If no, why?

It's already established that LLMs are more than a fad and here to stay. Soon there will be very stable APIs for using LLMs like we use sklearn, Tensirflow, Pytorch, Keras etc. We'll know the fundamentals but will rely totally on the already optimized implemented APIs. But, I see a lot of skepticism around LLM specially from DS circles not much from MLE circles though. Any thoughts?

apis data datascience data scientists etc frameworks fundamentals industry keras llm llms pytorch scientists sklearn will