A friend of mine, Francois, is a fine photographer.
He also loves teaching, and he even writes a bit - including poetry.
He is also a computer nerd

.
For fun, he decided to input into one of the available AI options a set of his photographs of Paris and other interesting places, along with some of his poetry which referenced, among other things, some of his associated travels.
The context was an instruction to create a dialogue referencing both sets of work. It was set up with the intention that the result appear to be a podcast.
The result is uncanny. You hear a male voice and a female voice, and they appear to be discussing my friend, his photographs and his poetry. They sound incredibly natural as they weave in and out of references that connect the two bodies of work, and my friend's background. The "podcast" is about 20 minutes long, and there are almost no clues to the fact that the entire thing is a creation of the AI tool that was used to bring it into being.
Francois told us that it was an AI creation, but I had an inkling because I happened to have another friend - Wallace, one of the members of our Darkroom Group - who had done something similar with the background to a third friend who is a fascinating resource in his own right, and for whom there is available a bunch of background information available on the internet. Both of my friends Francois and Wallace must have used the same AI tools, because the resulting podcasts used the same male and female voices!