I appreciate everyone who raised concerns about the thumbnail artwork used in the video below. It was the first time I used AI art in a thumbnail, though I have used it before within videos from time to time. I understand that some channel viewers are against the use of AI-generated art in principle and feel very strongly about it. Over the past couple of weeks I have been thinking about it, as I am always concerned about doing the right thing.
Some people don't like "AI slop" simply because it is crappy. This post is not for such ones. If you think it's poor art, that's understandable. I have made many crappy thumbnails in my time, and many of them are still up, but I never got much hate for any of those.
First, as I said in some of my replies to the comments under the video, I have great respect for human artists and love their craft. My use of AI-generated imagery isn't a dismissal of human creativity but simply an accessible alternative given the financial constraints of my budget.
Several people have accused me (either directly or indirectly) of stealing from human artists. But no copyrighted imagery was used in the thumb. Stealing would imply using a specific, identifiable piece of artwork created by someone else without permission or compensation. This AI-generated image is a completely new illustration that does not replicate any particular existing piece of art or any character created by a human artist.
It doesn't even copy any particular artist's style, or art studio's style. It draws broadly upon general cartoon illustration conventions, blending several influences to create an original look. Specifically, it integrates elements commonly found in editorial cartoons and caricatures, retro boxing posters and vintage advertisements, and a dash of influence from Ligne Claire comics. It's an original synthesis, and that is exactly what human artists also do. Human artists continually build upon the works of those who came before them. AI-generated art is trained similarly: it analyzes patterns, styles, and techniques from a vast pool of publicly accessible data to produce original compositions.
Now, someone can argue about whether DALL-E (by Open AI), the app that I used, violates copyright by drawing from other artist's material, and I know that there are legal cases pending, but the way I see it, the AI is only doing what people themselves do. It looks at other people's art and is influenced by it. The only legal cases I think have legs are those in which the app outputs actual images that are copyrighted or characters that are copyrighted.
Thank you all again for voicing your concerns. I realize that I cannot give this complex ethical topic its due justice in a relative brief post like this, but constructive dialogue helps us all better understand this issue, so feel free to chime in. I hope this clarifies my position and demonstrates that my intention was never to disrespect or undermine human artists.
World of Antiquity
I appreciate everyone who raised concerns about the thumbnail artwork used in the video below. It was the first time I used AI art in a thumbnail, though I have used it before within videos from time to time. I understand that some channel viewers are against the use of AI-generated art in principle and feel very strongly about it. Over the past couple of weeks I have been thinking about it, as I am always concerned about doing the right thing.
Some people don't like "AI slop" simply because it is crappy. This post is not for such ones. If you think it's poor art, that's understandable. I have made many crappy thumbnails in my time, and many of them are still up, but I never got much hate for any of those.
First, as I said in some of my replies to the comments under the video, I have great respect for human artists and love their craft. My use of AI-generated imagery isn't a dismissal of human creativity but simply an accessible alternative given the financial constraints of my budget.
Several people have accused me (either directly or indirectly) of stealing from human artists. But no copyrighted imagery was used in the thumb. Stealing would imply using a specific, identifiable piece of artwork created by someone else without permission or compensation. This AI-generated image is a completely new illustration that does not replicate any particular existing piece of art or any character created by a human artist.
It doesn't even copy any particular artist's style, or art studio's style. It draws broadly upon general cartoon illustration conventions, blending several influences to create an original look. Specifically, it integrates elements commonly found in editorial cartoons and caricatures, retro boxing posters and vintage advertisements, and a dash of influence from Ligne Claire comics. It's an original synthesis, and that is exactly what human artists also do. Human artists continually build upon the works of those who came before them. AI-generated art is trained similarly: it analyzes patterns, styles, and techniques from a vast pool of publicly accessible data to produce original compositions.
Now, someone can argue about whether DALL-E (by Open AI), the app that I used, violates copyright by drawing from other artist's material, and I know that there are legal cases pending, but the way I see it, the AI is only doing what people themselves do. It looks at other people's art and is influenced by it. The only legal cases I think have legs are those in which the app outputs actual images that are copyrighted or characters that are copyrighted.
Thank you all again for voicing your concerns. I realize that I cannot give this complex ethical topic its due justice in a relative brief post like this, but constructive dialogue helps us all better understand this issue, so feel free to chime in. I hope this clarifies my position and demonstrates that my intention was never to disrespect or undermine human artists.
5 months ago | [YT] | 75