It's helpful because they know more about what they're looking at, I guess? I'm a bit confused by the question - why wouldn't consumers want to know if a photo they're looking at had a face-slimming filter applied?
You're not thinking like a compliance bureaucrat. If you get in trouble for not labeling something as AI-generated then the simplest implementation is to label everything as AI-generated. And if that isn't allowed then you run every image through an automated process that makes the smallest possible modification in order to formally cause it to be AI-generated so you can get back to the liability-reducing behavior of labeling everything uniformly.
It may not be relevant. What if I want ro pyt up a stock photo with a blog post. What benefit does knowing whether it was generated by multiplying matrices have to my audience? All I see it doing is increasing my costs.
The benefit is that your audience knows whether it's a real picture of a thing that exists in the world. I wouldn't argue that's a particularly large benefit - but I don't see why labeling generated images would be a particularly large cost either.
It's helpful because they know more about what they're looking at, I guess? I'm a bit confused by the question - why wouldn't consumers want to know if a photo they're looking at had a face-slimming filter applied?
You're not thinking like a compliance bureaucrat. If you get in trouble for not labeling something as AI-generated then the simplest implementation is to label everything as AI-generated. And if that isn't allowed then you run every image through an automated process that makes the smallest possible modification in order to formally cause it to be AI-generated so you can get back to the liability-reducing behavior of labeling everything uniformly.
In fact this is exactly what happened recently with sesame labeling requirements: https://apnews.com/article/sesame-allergies-label-b28f8eb3dc...
It may not be relevant. What if I want ro pyt up a stock photo with a blog post. What benefit does knowing whether it was generated by multiplying matrices have to my audience? All I see it doing is increasing my costs.
The benefit is that your audience knows whether it's a real picture of a thing that exists in the world. I wouldn't argue that's a particularly large benefit - but I don't see why labeling generated images would be a particularly large cost either.
4 replies →