2 ms·
Moderating based on "is it AI" never made a lot of sense, and SE is correct to pull the plug on it. It's a similar problem to the (possibly apocryphal) story of
by lolinder 3y ago
Moderating based on "is it AI" never made a lot of sense, and SE is correct to pull the plug on it. It's a similar problem to the (possibly apocryphal) story of the national parks trying to design bear-proof trash cans—"There is a considerable overlap between the intelligence of the smartest bears and the dumbest tourists." Any anti-AI policy will inevitably catch a bunch of people who just aren't great at English, or are still in high school, or any number of other reasons why their prose might sound stilted. There is no function that can reliably separate the two overlapping probability curves, so trying to do so is pointless.
Moderating based on non-constructive comments makes much more sense—if someone posted 30 answers in one hour and none of them answer the question, moderate that. In practice nothing changes, you're still banning people who are abusing AI, but you're doing so based on a concrete quality metric rather than a flawed algorithm plus gut check.
Obligatory xkcd: https://xkcd.com/810/ https://xkcd.com/810/