Some org (who I will name after this poll ends) published this poll on Twitter (🙄). They used the results to try to validate their POV on AI theft.
Though it won’t provide meaningful research data, I am curious to see how Mastodon responds.
(Please boost so we can get good numbers. 🙌🏻)
—
Question: should openly licensed content (images, music, research, etc.) be used to train AI systems? (Reply with reasoning if you feel called.)
#AI #ArtificialIntelligence #Art #Music #Copyright
> openly licensed
That's not specific enough. If it's "public domain", "do whatever you want with this and no attribution required", then sure, use it for whatever.
But if it's a license that requires attribution, like the CC licenses, then it's wrong to "train" so-called "AI" systems on that, because what they will do is to throw away the info about who the author was, and then generate "new" things as if they were "authored" by the "AI" system.
