The abandonment of AI image generators (for now)
In an update to my previous post, I have abandoned AI image generators for the time being. The randomness factor is just too high to be able to realistically grant any authorship to the prompter. It isn’t recreating the image you have in your head. In fact, it was unable to manage this in any of the experiments I made. At best, it was able to give me a couple of ideas for future images. But even this I don’t like. Will an AI image now be the source for much of the art that we come into contact with? Pandora’s box has truly been opened. It’s not that artists have never borrowed, but at least they were borrowing off fellow humans, meaning the final image in question was distinctly human. AI is trained on humans, but its final output is random, if random can truly exist in a computer. To be completely honest, there was an initial period where I waxed lyrical about AI image generators capabilities, but now this has been replaced by a sickly feeling. I’m not sure if this is partly due to the massive backlash from actual artists using a physical medium, or whether it is just from using it more extensively, and thus seeing the artifacts/errors that paint the image as distinctly AI. Either way, I have seen enough to know that I will be abandoning it for the foreseeable future, and will instead look to continue with photography (which, although similarly lambasted during its infancy, is still capable of representing a more accurate artist’s vision from mind to medium). It will be slow, but I hope to make more pro-life images in the future, within the photography medium, and with other human artists as my inspiration.