After some limited tests, people tend to pick the AI-written post as the real one more than 55% of the time, almost 60%.
So apparently LLMs kinda pass the Turing test when it comes to social media posts and not even general social media posts, but posts of a specific user!?