Unpopular opinion Qwen 3.8 is hard to understand

A user finds Qwen 3.8 27b and Qwen 3.8 Flash Next difficult to understand, citing examples like the use of set intersection symbols and the word "persona" instead of more human-readable explanations. They suggest this density might stem from minimizing "tokens per intelligence" and express concern that Qwen is following Claude 5's path, which was seen as a step backward in human-readability. The user believes Qwen 3.6 was the last easily readable Qwen model.