Exploring solutions and challenges for the controversial debate about AI art
It’s been lower than a yr since DALLE•2 was released and blew our minds away with the variability of art it could generate from a straightforward yet descriptive prompt. Other tools like Midjourney and Stable Diffusion quickly followed.
The initial stages of pleasure were quickly replaced by concerns in regards to the of AI-generated art. Were these AI systems creating original art? Or were they copying other artists?
Artists’ concerns over the past yr materialized right into a class motion lawsuit filed against Midjourney, Stability AI, and DeviantArt. The suit alleges copyright infringement, unfair competition, breach of terms of service, and more.
I’m not a legal expert, so I can’t weigh in on each of those claims.
Technically, these AI tools generate latest images, pixel by pixel, each time someone uses them. But the issue is that these models were trained on datasets containing images and art found online artists’ consent.
For instance, Midjourney and Stable Diffusion were trained on a dataset with 5 billion images containing work from artists without their permission.
Should artists and designers be compensated?
I actually have some ideas on how it could occur. Perfect solutions don’t exist, but I hope this creates a discussion regarding the longer term of AI-generated art. When you’re done reading, share your thoughts within the comments.
To present my ideas, I’ll describe them using a AI image generator called .
Training
: training the AI model behind ArtFair.
Unlike other AI art tools, ArtFair pays artists a for every image they use to coach the model. This might be an expensive upfront cost for ArtFair, nevertheless it’ll create an incentive for all artists to sell their art to them. And it advantages each parties:
- can earn a living from the portfolio of art they’ve built over their careers, which not generates income for them.
- could have the legal right to make use of the pictures without worrying about lawsuits that may lead to hefty legal fees, settlements, and/or fines.
Revenue Sharing
ArtFair will generate revenue by selling credits to users, similar to current tools. To maintain things easy, let’s assume:
- 1 credit = 1 image
- 10 credits = $1
So, if I wanted 10 AI-generated images, I’d pay $1.
Revenue from credits might be split 50/50 between the platform and its artists. The 50% for artists might be distributed based on how much art they contributed to training ArtFair’s AI.
For instance, if Samantha contributed 30 images to coach the model and George only contributed 5, she would get the next share of the revenue from credits sold on ArtFair. This incentivizes artists to sell more art to the platform since they might be rewarded accordingly.
This creates one other positive growth loop that advantages each ArtFair and artists:
- Artists receives a commission a one-time fee for images they contribute to ArtFair’s training dataset and receive additional royalties based on how much art they contributed.
- ArtFair uses those images to coach and improve its AI image generator, attracting more users who will turn out to be loyal to the platform.
- Attracting more users results in more credits being sold, giving ArtFair artists more cash.
Yet one more thing…
I’ve proposed a option to align the incentives between ArtFair and artists, but .
AI-generated art doesn’t all the time look great, but it could be a for getting ideas on the art we want. Some people will want higher quality art or a more consistent look across multiple images they generate.
Here’s a portrait I generated using AI. I really like every thing about it, just like the colors and art style, but the additional shoulder ruins it:
Imagine you desired to create an illustrated children’s book. You could possibly use ArtFair to generate your initial concepts, ideas, and storyboards for the book you ought to create.
ArtFair could then provide recommendations of artists that match your required art style. This is feasible since each image that ArtFair is trained on is related to a particular artist, who could have a profile on the platform.
You could possibly contact the artist and share the (imperfect) AI-generated art you created to your book. The artist would use your rough art as a guide to creating the art you would like with all the main points you would like: characters, scenes, colors, and magnificence.
So, ArtFair can turn out to be a source of passive income for artists (through the 50/50 revenue share from credits) a lead-generation tool to search out clients willing to pay for higher-quality, skilled art.
All the pieces I’ve said to date painted an version of how an AI tool like ArtFair could pay artists for his or her work.
But, it’s necessary to focus on the and of such an approach:
Midjourney and Stable Diffusion were trained on 5 billion images.
Even when ArtFair only needed to pay for a of those images, the bill could be . I don’t doubt that big tech firms and VC-backed startups can find the cash to achieve this.
AI art tools are here to remain, they usually’re already becoming a staple of our online toolkit. This significant upfront investment might be definitely worth the payoff because the industry consolidates in upcoming years.
The opposite side of the coin is that artists may have to sell themselves short to contribute to ArtFair’s training dataset.
I discussed that ArtFair would distribute 50% of the revenue it generated from credits. Each artist would get a sum based on what number of images they contributed to training ArtFair’s AI.
But, the flaw on this approach is that images with different art styles have different demand. For instance, 3D art is perhaps a more popular category than oil paintings.
Using the Samantha and George example again, assume:
- Samantha contributed 30 images of her oil paintings
- George contributed 5 images of his 3D art
- The 3D art style is significantly more in demand on ArtFair than oil paintings, which generates more revenue from credits
So, shouldn’t George get the next share of returns because he contributed to a higher-demand category? If that’s the case, what’s the precise option to compensate everyone fairly?
I admit that the thought behind ArtFair is optimistic and idealistic. Whether a platform like it should ever exist will depend upon the consequence of the lawsuits we’re seeing today. If AI tools avoid penalties or repercussions, they’ll keep doing what they’re doing.
But, in the event that they lose the case, a business model like ArtFair’s won’t be out of the query.
Or… these firms will hire “art farms” in accordance with someone within the industry who commented on considered one of my recent videos:
Oh well… 🙃