I think you are mistaken on this point ; it does not sample images, it does not make a kind of puzzle based on the images. It learns from all of them, with millions parameters, transformed into concepts and ideas stored. But it does not actually keep any of its images in its database.
And it is for these reasons that I think this AI is REALLY similar to the brain. Because apart of our consciousness, the brain works like a machine, where our unconscious records billions of data and stores them.
So, for example, the idea of a "bike" has this or that image in our head: it can be bigger or smaller, of this or that color, it can be found in this place, it can be associated with this or that thing,...
This is basically how IA works, for then creating new images, original contents. Because it is an image that does not exist. Humans do not have the exclusive right to the word "original" I expect 😅
So... no, I don't give up, the functioning of this AI is very similar to our functioning as artists and beings with brains.
But, platforms like ArtStation makes it possible to disable allowing AIs to train with their images, if users so wish. I think it's a good compromise to lower tensions.
You are viewing a single comment's thread from:
I hear what you're saying and I understand it doesn't literally collage images together. I've tested it A LOT, I see what it does. I honestly think the technical details are less important than the fact that it can't do anything without that data. Less data, and it would be worse. More data and it will be better. While no image that it makes is exactly like an image from the training data, that does not equal original imagery.
Whether it has been trained to steal artists work, or it directly collages from billions of images, everything it creates is based on the work and labor of other people. Actual artists, even if they're referencing other stuff, THEY still have to create it. Their mark comes from them. So yea, I respectfully disagree.