Using face validity to check AI-generated content might be highly deceptive. I read that one of the ways of identifying them is that they usually have no grammatical or spelling errors. But with the popularity of Grammer checking tools like Grammerly which help writers to produce close-to-perfect writing, this method of identifying these AI writing will not work. As Lucius rightly pointed out, the best option is to check these suspected posts in more than one detecting software.to check its authenticity.
you're right, i may later in the nearest future begin to make use of some of these tools in checking for those plagiarists in disguise for the use of AI in making their post and using the excuse that they only do such in other to make their content carry the weight of thoughts they would have desired to see in them, which all these are flimsy excuses, so as suggested, we can make use of more than one checking tools to confirm the level of how true a post content could be AI generated.