"Authenticating that you are talking to a human and not an AI" is just a proxy for something else, most commonly wanting to be able to rate limit something. Because otherwise, why does it matter?
A lot of people are suggesting things like, have the government do it. But that has two big problems. First is privacy. If you have to prove your identity any time you want to do anything, nobody can be anonymous anymore, which is Very Bad.
Second, it assumes the government has some magic incantation that nobody else can use, as if the Post Office knows who you are in a way that your bank doesn't. But they don't. To get a government ID, they just want you to show them some other existing ID. It has no way to bootstrap itself any better than anything else. And some of the IDs they accept are easy to get... without an ID. Because everybody has to start from somewhere. The system has to be set up in a way that it works for people who emigrate from a country with untrustworthy institutions as an adult or if your house burns down and you lose all your documents you can still get new ones. An AI is going to be able to BS its way into a government ID, even assuming criminals wouldn't be able to hack into any state's DMV (as they already have).
It appears that going forward, telling the difference between a human and an AI is going to be hard. Maybe instead of trying to get better at that, we should find a different solution to the underlying problem.
The simple answer is to make account creation cost something. Nothing big, so someone who needs one account isn't paying much, but a spammer who has 1000 accounts get banned every day is out of business. And that's not even hard -- it's finally something cryptocurrency would actually be good for. Because you want a way for people to pay for access to things, while still being anonymous.
The real hard part is, how do you charge for account creation without deterring account creation?