What Users Are Complaining About
The experiences gathered in the posts are much of a muchness: content griping about Meta AI mysteriously throttled, posts teaching people how to turn off AI features flagged as violations, related accounts getting warnings or even bans. Meta's official explanation, as usual, pins it on misjudgment and automated moderation. The problem is that when enough of these "coincidences" pile up, it's hard for users not to wonder—when the platform is itself the vendor of an AI product, is its moderation system still neutral toward content criticizing its own product?
A Structural Awkwardness
The core of this isn't whether any individual case is real, but the structure: Meta is simultaneously the rule-maker and enforcer of its social platform, and a stakeholder in the AI product being discussed—three roles pressing on the same moderation system, and even absent subjective malice, the incentives are crooked. Similar doubts previously landed on Apple's App Store and Google's search results; now it's AI's turn. Betting all your important channels for speech on a single platform has always carried this risk, and this story is just another reminder.
via: Hacker News