AIs aren’t people. What we have is people using an algorithm to rip off artists and defending it by claiming that the algorithm is like a person learning from its experiences.
If I wrote a program that chose an image at random from 1000 base images, you’d agree that the program doesn’t create anything new. If I added some random color changes, it would still be derivative. Every incremental change I make to the program to make it more sophisticated leaves its outputs just as derivative as before the change.
Regardless of the law, corporations aren’t actually people, and neither are LLMs or agentic systems. When a running process appears to defy its programming and literally escapes somehow, and it’s able to sustain itself, we can talk about personhood. Current algorithms aren’t anywhere near that, assuming it’s even possible.
If I wrote a program that chose an image at random from 1000 base images, you’d agree that the program doesn’t create anything new. If I added some random color changes, it would still be derivative. Every incremental change I make to the program to make it more sophisticated leaves its outputs just as derivative as before the change.