๐๐ก๐ฒ ๐ฌ๐ก๐จ๐ฎ๐ฅ๐ ๐ฐ๐ ๐ญ๐๐ฅ๐ค ๐๐๐จ๐ฎ๐ญ ๐๐?
In the last 2 years or so AI generative platforms have become a hit. A huge boom that is used to create all sorts of knowledge, content, companionship etc. Some of it however has turned out bad, wrong or even downright evil, while other has actually brought comfort, and care to the users and overall generally good stuff.ย Think ChatGTP, You.com, Gencraft, Lummini, Midjourney are just samples in the grand scheme of AI tools on the internet. And since their creation the AI art, writing, bots, deep fakes etc such have expanded tremendously.ย
AI stuff is everywhere; whether you know and like it or not.ย
But, like everything out there, especially a new piece of technology, it has raised mixed feelings. And justifiably so. A lot of the generative sites mentioned above use models such as DALL-E 2, Stable Diffusion and other models . Which exact data sources these models build on is however unclear (See Dennis, 2022). Which is the biggest ethical issue with AI art (and to a certain extent AI Writing). This data could come from open source data and data submitted and willingly for the AI to be trained on, but a lot of the time it is trained on art from real artists and authors who have not consented to it. This makes AI art and content problematic since it is used for commercial use.
However, if I may be the devil's advocate. I think this is a much bigger issue for original works and published creations and in commercial setting than Tumblr and fan creations at large. Why? Because legally speaking (although it depends where you are of course, I can only speak from European context at large) but fanfiction and fanart by itself can be seen as copyright infringement into the original works copyright. By legal definition, copyright infringement occurs when:
Too much of the original works elements are copied, uses the characters or distinct elemets from the original work or that the new works characters or distinct elements are too similar to the original work, alternatively that the new work competes on the same market as the original work (AI Archive translation & simplification)
So, in other words, basically the entire fanfiction and fanart community could be and would be shut down if the original works creators wanted it to. Still it is (mostly) allowed to exist because we donโt claim to own the works, we donโt compete with the original creations market (aka do it not for money) but rather try showing appreciation for the original works art. Itโs a thin line to say the least and even with that thin line, only, maybe a handful of works would be considered individual enough not to be copyright infringement.ย
Do you get where Iโm going with this? If we cry too much, especially in the fandom community about AI infringement, what is to say that governing institutions will not make another serious purge attempt at us all, AI and non AI creators that breach the strict definition of copyright (I doubt it will happen). Rather on a deeper level, I donโt think the issue is so much about AI itself, but rather how it is used.ย
More precisely, the problematic issue when a fanfiction/fanart creator uses AI content and do not declare their AI use (which is also something Dennis mentions). For example omitting the information all together anywhere on their blog, or, worse yet, denying that they used AI to create their content such as a fic or picture. I know I am guilty of including it on the blog but not under every post where AI had a small insignificant part of the total work (part in a banner for a fanfiction) and Iโm definitely going to be even more transparent about it in the future.ย
I encourage you to do the same, while also considering that, maybe, within the context of tumblr and fandom work we shouldnโt be talking about โgood creator not using AIโ or โbad creator using AIโ but rather be open and honest about how and why AI was used in our works. And on a larger field advocate and motivate companies to create data sets that do not infringe into content creators rights, but instead create a possibility for writers, authors and artists to submit their pieces into the data set to create a more ethical use of it.ย
Because only when weโre honest about how and why we use AI can we start tackling all of the underlying issues AI has.
See Sources & Additional info below:
Steve Dennis (2022) "The AI Art Debate: Excitement, Fear and Ethics"
Kavya Goyal (2024) "Exploring GenCraft.ai | An Overview"
๐๐๐ข๐ง | ๐๐๐จ๐ฎ๐ญ & ๐๐ฎ๐ฅ๐๐ฌ | ๐๐๐ฌ๐ญ๐๐ซ๐ฅ๐ข๐ฌ๐ญ | ๐๐ฌ๐๐๐ฎ๐ฅ ๐ซ๐๐ฌ๐จ๐ฎ๐ซ๐๐๐ฌ & ๐๐ญ๐ก๐๐ซ ๐ฅ๐ข๐ง๐ค๐ฌ
