There is nothing wrong with AI writing articles. I am using it myself. The problem lies in the "information increment." In the past, we scrolled through Twitter to obtain information. But now, more and more content is produced quickly by AI, with the goal not being the quality of information, but rather the quantity—more posts, more exposure, more rewards. When "writing more" becomes the main KPI, "what was said" becomes unimportant. What we are currently criticizing is actually the proliferation of AI content with no incremental information, not AI itself. My judgment at the beginning of the year was that the channels for obtaining information in this cycle are changing; people are no longer relying on VC research reports or media, but are treating Twitter as their main source of information. But looking back now, there are problems with Twitter as an information source. The posting frequency is high, and the noise is great. When everyone wants to jump on the bandwagon and be the "first to post," our efficiency in obtaining information is actually declining. Perhaps, useful content is being drowned out instead. This has led to several trends: > Creators are being forced to go to other platforms, such as Binance Square, @Sidekick_Labs; > Users are starting to return to private domains, private messages, and community circles. So, when people say that Twitter's traffic has decreased recently, it is not an illusion. It is not that the platform has a bug, but rather that its information density is genuinely declining.