For this series of artificial images, I followed a relatively simple yet effective protocol. I've selected some photographs taken in Bali as part of a reportage on the island’s "off-screen" (see here), which I submitted to ChatGPT to generate prompts capable of recreating the images. I then used these prompts along with the original images in MidJourney to produce visuals that display varying degrees of deviation from their references.
The gap introduced through this process ultimately results in a vision very close to the original reportage while seemingly preserving the critical dimension of the initial approach: namely, highlighting the collateral damage of mass tourism. In the end, can this "fake" series also achieve the goal of the original project—to describe a situation in order to critique it?
This question echoes a similar issue I had previously explored when creating a "pictorial print" of these same images through commissioned work by artisan painters (see here). That process introduced a different kind of distance from reality, but isn’t a photograph itself already an interpretation subject to doubt?
Ultimately, what is the reasonable, acceptable, and appropriate distance from reality that aligns with intent? To me, these questions remain open.