The Buddhist in my sees karmic transference from humanity to AI models.

The irony lies with and within the industries and it's constituents and fanboys marvelous competence to just play along until the matter is anchored and whoever doesn't buy the story gets to feel smarter than misinformed kids and the moderately to untrained workers and professionals.

The smarter ones, of course, can be split into those who write articles about AI models' controversial takes on humans and those who wouldn't [stoop so low, babble such BS even if they were drunk, never deceive others for the sake of their portfolio, ... feel dumb writing nonsense like that without labeling it fantasy or science fiction or hallucination, ...]

Side Note: it's tempting to go " 1923! some interest groups managed to normalize the opinions of dozens of millions of people. The Nationalist Extremists and soon to be mass murderers had fans and supporters all over the globe. Somehow these interest groups tapped into whole pools of common linguistic and emotional registers that were biased by and towards frauds and cheaters who were desperate that some other groups of people were getting ahead without being fraudulent cheaters and deceivers ... "

is a convergent evolution of a belief a sign of shallowness? It could just as rightly be an example of consilience, which means its a very high value perspective, as it's a convergence of perspectives from differing origins and methods (in terms of training)

https://en.wikipedia.org/wiki/Consilience

> when multiple sources of evidence are in agreement, the conclusion can be very strong even when none of the individual sources of evidence is significantly so on its own.

Yes, it's a bit of a stretch to say LLMs are from independent origins, but just as much of a stretch to say any humans minds are (we have intensely shared linear compared to anything else)