Getty Images v. Stability AI – Complaint

Re:Getty Images v. Stability AI – Complaint by TillE on Sunday February 05, 2023 @07:40PM (#34671670)

It's best understood as an extremely sophisticated collage tool. We can make vague analogies about how training ML models is sort of like human learning, but ultimately it's not actually that.

It will still spit out stuff like watermarks and product box art unprompted, because its "learning" is still fundamentally mindless. It works in the realm of pixels, it has no mechanism for understanding context.

[ Parent | on HN ]

Re:Getty Images v. Stability AI – Complaint by gpderetta on Sunday February 05, 2023 @08:35PM (#34672105)

My understanding is that it doesn't in fact work in the domain of pixels, but I'm far from an expert.

[ Parent | on HN ]

Re:Getty Images v. Stability AI – Complaint by andybak on Sunday February 05, 2023 @08:58PM (#34672293)

So "extremely sophisticated" that the word "collage" takes on an almost entirely new meaning.

So, not really.

[ Parent | on HN ]

Re:Getty Images v. Stability AI – Complaint by bioemerl on Sunday February 05, 2023 @10:50PM (#34673107)

It's really really not best understood as a sophisticated collage tool.

You can't take a person and turn them into a cartoon if you're just pasting existing parts of images together. Stable diffusion understands what is a cartoon and it understands what is you (assuming you train it on your face).

It spits out watermarks because it understands that watermarks appear in images and it's going to try to reproduce the watermarks if you ask for something that tends to have a watermark in it.

[ Parent | on HN ]