#Enshittification – jetzt auch bei AI-Models? > Users report re-reading sentences several times to extract meaning, and some now even routinely pipe Opus output through another model to get a plain summary. https://github.com/anthropics/claude-code/issues/77136 #claude #opus #ai