[thewebguyd]

The USCO's decision hinges on whether or not a human has predictive, mechanical control over the final output. The ruling applies to Generative AI, the USCO made a separate distinction for assistive AI, which image classifiers would fall under.

> “Authors have long used such tools to create their works or to recast, transform, or adapt their expressive authorship. … what matters is the extent to which the human had creative control over the work's expression and actually formed the traditional elements of authorship.”

The USCO doesn't care what type of algorithm is used, it cares who determined the traditional elements of authorship. If a human dictates the expression, and then uses a computer to clean, translate, or refine it, it is copyrightable. If a human just provides an idea and a generative algorithm creates the specific expression, the output is public domain. One is using spellcheck, the other is telling the computer "Write me a novel" and letting the computer generate it.

[make3]

AI cost companies billions of dollars to make the time of hundreds of expert researchers. You're saying that my tool became too good so I can't make money from it anymore?

What about protein language models that create new protein drugs, can pharmaceutical companies not protect them too? That's straight up the same time of model as LLMs and image models.

[thewebguyd]

"I'm" not saying anything, I didn't make the ruling, the US Copyright office did.

Copyright is a human concept, to protect the exclusive licensing of a human creation. We don't grant copyright to machine output.

Your later example would be more than likely protected via patent, not copyright.