For example: https://www.v2ex.com/member/openercn

Looking at its historical replies, all of them are:

• single-character verb abuse : "拆" (dismantle), "兜" (pocket/encircle), "接" (connect), "做" (do), "备" (prepare), "跑" (run), "定" (set), "绑" (bind), "落" (fall), "扩" (expand).
• single-character adjective abuse : "漂" (drift/float), "稳" (stable), "变" (change), "行" (okay/run).
• GPT starter phrases : "I will prioritize...", "I will do it according to...", "I will distinguish by...", "It's not... but...", "It is... rather than...", "More like a...", "That's very stable", "The most troublesome thing is", "The real difficulty is", "The key is 'can xxx'", "Know why xxx, why xxx".

Historical posts are all: " Made an X using AI, hope to hear Y's feedback "

It can basically be considered as purely AI-generated content, and it is most likely from the GPT 5.2~5.4 series of models. Shouldn't this be handled according to community rules?

The vast majority of replies are not informative at all, full of repetitive talk, and not helpful to the topic.

@Livid