from Hacker News

Generative AI Will Be a Boon for Public Domain Unless Copyright Blocks It

by SeenNotHeard on 12/29/23, 7:31 PM with 5 comments

• by NemoNobody on 12/30/23, 4:01 PM

  Is anyone asking actual creators what they think of AI?

  Like, you wrote that huge thing about how it's replacing a line of work but that isn't the case. I'm not very good at editing images and tho this has changed my world I the sense of what I can initially accomplish - I still need a graphic designer.

  I like to occasionally make websites that I never really do anything with, it's been years since I've made one and I'll just say with coding this changes everything for people that actually know what they are doing - it saves so much time.

  Our time is the most valuable thing we have. AI can help creators accomplish things in a fraction of the time it would take without it.

  Copyright law isn't for creators it's for corporations - don't pretend otherwise

• by Tanoc on 12/30/23, 5:11 AM

  I seriously question how they believe this, because it doesn't help bring things into public domain when all the training data itself is copyrighted and possibly pulled from trademarked brands. Which is a problem public facing offerings from OpenAI, Google, and Facebook refuse to solve both due to legacy baggage, complexity, and the potential lost profits. Unless you can absolutely 100% prove that all training data was captured and input by yourself at every stage courts will not entertain the idea. If anything this hurts smaller artists or groups while helping larger corporations and organizations because only the large corporations will have the resources to gather that unique wholly-owned training data. The article and comments section also don't seem to understand that corporations will and do use such tools to find the very legal edge in order to consolidate power even more than they already have.

  There's the first part of that consolidation that if a tool can easily reproduce signature elements it makes it harder to defend against copyright and trademark infringement in court because the copyright and trademark can be diluted. That is a double-edged sword, but the sword is swinging in your direction and not the corporation's. The public will work faster than the court ever can to dilute that copyright or trademark as shown with things like memes using copyrighted images, meaning by the time the case is actually heard it might already be lost. A corporation will always have the resources to disseminate things better than an individual can, allowing them to dilute the copyright or trademark quickly so they can profit off of their own widely distributed derivative version that they can copyright and trademark. An infamous example is the 1960s Loufrani smiley face, where the public diluted the copyright and trademark enough that Wal-Mart attempted to associate it with their brand and trademark it. Wal-Mart only barely lost that court case to Franklin Loufrani.

  The second part of that consolidation is that it's enabling the eradication of an entire line of work and removing even more financial power from individuals. By using AI that's abusing copyright to fight corporations abusing copyright you're removing an entire group of thumbnail, insert, layout, and title artists that are already being hurt by corporations and are already extremely versed in getting around the broken copyright laws. Those who use AI keep said artists from being involved and paid, making it harder to influence industry standards and for everyone else driving up the cost of getting the work done while reducing the income of the people hired to do the work. You cannot compete with free and lightning fast.