There was one paper recently where the AI beat humans at Turing test 2/3rds of the time.

I think it's cause they told it to type like a 13 year old and nobody could imagine AI talking like that.

We don't post-train current frontier models to pass the Turing test, but if we did, it wouldn't be much of a challenge for current models IMHO. It's a dead benchmark. It tests the human machines, not the machines.