←back to thread

258 points signa11 | 1 comments | | HN request time: 0.239s | source
Show context
aithrowawaycomm ◴[] No.42733199[source]
> Meanwhile, we will see more focused efforts to create truly free generative AI systems, perhaps including the creation of one or more foundations to support the creation of the models

I understand this will be free-as-in-beer and free-as-in-freedom... but if it's also free-as-in-"we downloaded a bunch of copyrighted material without paying for it" then I have no interest in using it myself. I am not sure there even is enough free-as-in-ethical stuff to build a useful LLM. (I am aware people are trying, maybe they've had success and I missed it.)

replies(2): >>42734226 #>>42738687 #
1. ASalazarMX ◴[] No.42738687[source]
I don't think blindly abiding to copyright is the higher moral instance here, even if it's the law. Knowledge wants to be free, and the way AIs need to be trained now is a sign that copyright laws have become unreasonably restrictive and commercialized.

Not only AIs should be allowed to train on pirated content, humans should too. Copyright laws need to be scaled back so that creators are protected for a reasonable period, but humanity is not gated out of its culture for decades. The cheaper culture distribution has become, the harsher copyright laws have evolved.