Add to that at one point Claude was the go to models for the very basic use case for LLMs - text generation.
With newer models text generation outputs have gone from probably human readable text to dense philosophical treatise about "load bearing" and incomplete sentences. So much so that now you need skills or another LLM to just parse the output. Simple answers and text generation just doesn't exist.
With newer models text generation outputs have gone from probably human readable text to dense philosophical treatise about "load bearing" and incomplete sentences. So much so that now you need skills or another LLM to just parse the output. Simple answers and text generation just doesn't exist.