Towards a definition of “Open Artificial Intelligence”: First meeting recap https://blog.opensource.org/towards-a-definition-of-open-artificial-intelligence-first-meeting-recap/ #ai #DeepDiveAi
@osi Looks like #Open #Source #AI is going ethically in the wrong direction.
Just because #attribution is hard, it doesn’t equal ”useless”.
”One group member pointed out that “attribution” for a dataset might result in a 300-million page PDF. “Completely useless. It would compress well, because most of it would be redundant.””
Yeah. Most, but not all, and you can’t know which because you can’t figure our whose #work is going to get #stolen in advance.
Bullish nerds being bull.
@osi Ps. If anybody is listening, please read my blog about ”a middle way”. That we must do not because it’s easy but because it’s hard. Otherwise intellectual property as an incentive for encouraging innovation is done for.
https://gimulnaut.wordpress.com/2023/01/13/copyright-wars-pt-2-ai-vs-the-public/
@osi PPS. If all you #ai #artists think I’m sidelining you, don’t jerk that knee yet. I see you, and the value your work carries.
But our economic and political systems don’t see it in a fair and equal way. We need a new class of art to finally be coded into law: The #transform.
#ai #legislation #art #aesthetics #theory
https://gimulnaut.wordpress.com/2023/04/20/ai-art-is-a-remix-the-djs-of-pictures/
@gimulnautti the comment you quoted doesn't say it's **hard**. It says it's **may** not be useful. The discussion is not over, that was only the first meeting.
@osi Thank you for reading my comment. I realise now I did use harsh language and I apologise.
But to get to the point, I tried to offer a statement not a quote. Making reliable attribution is much more difficult than making models. This was stated from a technology perspective.
If you read my articles in the other thread, I offer some reasons there to keep attribution around, even though right now it might seem like just another piece of redundant data.
Hope to talk to u again