I think these types of tests can only result in showing human failing at them. Because the test is reduced to judging only the single artefact as a thing in itself, no context etc. That's the basic element of all cons: make you focus narrowly on something, where the facade is, and not where you would find out it's fake. Turing isn't about whether something's human, but whether we can be made to believe it is human. And humans can be made to believe a lot. Turing needs to keep you from looking behind the curtain / in the room to make the test work, even in its shape as a thought experiment. The study (judging by the sentences here) is a Turing test in the real world. Why would you not look behind the curtain? This is the equivalent of MIT's tedious trolley problem fixation and calling it ethics of technology, without ever realising that the way out of their false dilemma's is acknowledging nothing is ever a di-lemma but always a multi-lemma, there are always myriad options to go for.

On a web with many generative agents, all actors are going to be assumed models until it is clear they're really human.

Maggie Appleton calls this 'passing the reverse Turing test'. She suggests using different languages than English, insider jargon etc, may delay this effect by a few months at most (and she's right, I've had conversations with LLMs in several languages now, and there's no real difference anymore with English as there was last fall.)

https://web.archive.org/web/20230503150426/https://maggieappleton.com/forest-talk

Maggie Appleton on the impact of generative AI on internet, with a focus on it being a place for humans and human connection. Take out some of the concepts as shorthand, some of the examples mentioned are new to me --> add to lists, sketch out argumentation line and arguments. The talk represents an updated version of earlier essay https://maggieappleton.com/ai-dark-forest which I probably want to go through next for additional details.