Not sure this is where to post this question, if not I apologize. There are so many discussions on Society6 regarding artists using AI instead of their own creations. Is what I create using Repper, which I absolutely love, considered an AI creation?
Georgiana, while we wait for repper’s reply on this, let’s think about it for a moment.
The new kind of AI involved in the current computing ‘revolution’ is called ‘generative AI’. It produces or generates for us by trying understand a query or command from us, a ‘prompt’ from which the AI generates a piece of writing or a photo. This is not how Repper works at all. Repper is definitely clever computation and I don’t know what it’s called but I would bet the farm that it’s not AI.
But for those who worry about AI stealing ideas and methods from artists, we humans have done this for ever. Every time you look at an image, listen to a piece of music, you are training your own brain like AI. We all create from our memories or self-training. It’s affectionately called ‘influenced by’ but no one can create from nothing. All artists, writers, actors take as their first steps a role model, an idea and we try to be or create something as good as that. Eventually we find our own styles. So, these people moaning about AI are actually moaning about themselves.
As @Bernardus correctly says, Repper isn’t AI-based. We use geometry and visual effects to turn part of an image into a pattern. Only the source image is used as input, everything else is your designs decision in the software and the calculations it powers.
The great thing about the way Repper works, is that if you know you have the right to use the source image as input, everything that comes out of it is yours to use as well.
To be clear about right to use source images, because this is a complex subject:
You need the right to use the source image for remixing purposes (sometimes also called derivative work)
If you don’t want to credit the original artist, you need the right to give/share/sell the pattern without attribution.
If you want to use the resulting pattern commercially, you need to right to use the source image commercially. Commercial rights can get quite complicated, like on which medium, to what quantity, or in which geographic areas. In case of doubt, but cautious and double-check.
Because of these complexities, the easiest thing to do either to create your own source material. This is easier than you think, because Repper doesn’t need much to create amazing patterns! It also gives you a lot of control over the resulting patterns in terms of colours, style, etc. Our content creator @Michel_Repper made a useful blogpost about this a while ago.
Alternatively, find good sources for images that have permissive licenses like Unsplash (which is built into Repper, see this help article) or the public domain library from Rawpixel.
(Note: this post is not legal advice. If you use source images you did not make yourself, always make sure you have a right to use them in the way you plan to)
@Bernardus, while I agree that art is always based on prior experience, there is of course a large grey area between truly novel work and 1-to-1 copying. Exactly where AI text-to-image generation sits, I find hard to say at this point.
The creative possibilities of generative AI are amazing, both for trained artists and those new to the field. It can help spark new ideas, improve on existing work, and help people learn and grow. On the other hand, it’s been shown that what comes out of these generative tools can sometimes be very close to the original works of art the models where trained on, and so using the output becomes little different from copying someone’s work directly and selling it without their permission.
I’m still on the fence about the legal and ethical aspects of AI generation, but it’s an interesting and important discussion. In the meantime, use generative AI tools with care
You know, there is an easy way to get past this bump in the road. Obviously copying is not on, even the copycats aren’t getting true satisfaction from staright-forward copying, 99% are just playing anyway let alone the idea that someone is profiting from another’s work. Rights and copyright about this is an absolute jungle of morals, ethics, philosophies and any other mind-twisting thing we can think of. You know, when is an idea original or not original. I prefer the journalistic approach to stories. Media is my field, radio actiually. The journos chase stories and who the heck knows where they originate from? An outlet has to actually state ‘exclusive’ to protect their story and even then it is often argued that once a story become the subject of national gossip it’s not exclusive any more and other laws come in to play. I came across a site recently that made me laugh out loud. It was a famous quotes site, and they had included a number of notices to the effect that their (famous) quotes were exclusive to them and that if you wanted to use a quote you had to get permission in writing. Well, what a load of nonsense, right?
Back to the idea that came to me. If there was a rule that stated when using all generative art, that your AI work, if you were to use styles of famous writers, artist etc had to be a mix of at least 2 and preferably more different artists/writers. There it is. You would have to combine the styles of 2 artists, 2 photographers, 2 anything. Then at that point we could all agree that the new AI work has to be original. Yes, it would still be ‘derived’ but a combination of two artists would be like making a baby
Hi Wouter,
Thank you so much for your explanation and thoughts on this matter. I can now confidently move forward with Repper, which I absolutely love!
Thank you all for this thoughtful discussion. I’m happy that we can have these kinds of exchanges on this forum, it gives me confidence that we’re doing the right thing as a community!