1
AI and the Fall of Writing
AI生成的文本过于完美，反而暴露了非人类本质，让人反思写作的未来。
We can all spot an AI written text Not by its hallucinations - because it generally isn't any. Not by poor grammar - because it isn't. In fact, that's…
2
MatchLM2Lite: A Scalable MLLM-to-Lite Framework for Reproduced Content Identification
提出可扩展的多模态大模型精简框架，高效识别视频平台重复低质内容，保障内容原创性与用户多样性。
arXiv:2606.14786v1 Announce Type: cross Abstract: Content moderation is critical for online video platforms to ensure content safety, protect creators…
3
Ask HN: What do you do with LLM articles?
当读文时突然发现是AI写的，你会怎么想？HN讨论区用户的真实感受与识别技巧。
I don't want to make a rant about articles written entirely or partially with the aid of LLMs. I do myself sometimes resort to that type of help if I …
4
YouTube to begin automatically labeling AI videos
YouTube即将自动标记AI生成视频，靠C2PA元数据及水印工具识别，但技术细节仍不明朗。
AI videos that are animated, unrealistic, or only have a little AI may still hide their origins, though.
5
YouTube Is Making It Easier to Spot AI-Generated Videos
YouTube推出新AI标签，让识别AI生成视频更简单，并直接点名打击"AI slop"。
AI slop should be easier than ever to identify on YouTube.