Number one: It ain’t AI. It doesn’t have a will of its own.
Number two: Can’t add or create any materials or textures on its own. If no one adds the text, the image, or the texture, like a color or skeleton, it doesn’t exist in its library.
Number three: It can only follow the rules or perimeters created and set by a human, and can’t make or set its own in anyway.
Number Four: Can’t do anything without it being told to. If you don’t tell it to, it can’t do it.
Number five: If it’s outside of its rules and perimeters or it doesn’t exist in its library, it can’t do it.
Basically, how it works is that it doesn’t generate, as people believe, but Constructs by using the skeleton of one thing in its library and then following its rules and perimeters to put textures, materials, and text in its library on the skeleton.
Example: If I want to create an image of a pink rose, I need to add the image of a rose, aka the skeleton, then add the color pink, set the perimeters for where it to cut and change the color, and then tell it to do it.
And I got an image of a pink rose. Now, if I want to change it to blue, I have to add the color blue and tell it to do the same thing with the pink.
And that will be always in its library for future use and improvement.
So, if something CONSTRUCTED by Machine Learning is bad, it’s not its fault but the person who added to its library and set the rules and perimeters for it.