From digital art to NFTs, and now AI generated art, technology has slowly been claiming its place in the art field. So where do we draw the line? With AI self-portraits recently taking the social media world by storm, more conversations are taking place regarding the ethics of artificial intelligence.
What is AI art and how does it work?
AI Art is generated using artificial intelligence that depends highly on machine learning. Many AI art generators use the Stable Diffusion Model. This is a model that can create and modify images based on text prompts. Essentially, the algorithm is fed AI images with associated captions collected from around the web. With enough data collected, the AI is able to create content based on prompts, out of information it has already gathered.
Why are artists unhappy?
Apps like Lensa sample artwork from artists that haven’t given consent for their work to be used. Many have even found pieces of their artworks within these AI generated images. Unfortunately, since the art produced is not a replication of one person’s work, but rather a collage of multiple pieces, companies can exploit a legal grey area to their advantage. Lensa has made $16.2 million so far, and no artists have been compensated.
For instance, Sarah Andersen has added her voice to the conversation to raise her concern. In a recent article, she explains how her webcomic was modified and posted to an alt-right forum, and then stolen by an AI art app.
Greg Rutkowski is another example of the damage that AI apps can do to artists. His style is one of the most used prompts in one of the AI art generators. This article by Technology Review shows the striking resemblance between his digital work and AI generated prompts.
A large component of art is the originality and creativity that goes into its creation. If we remove these human elements, is it still art?
What are some other criticisms?
As human cognitive biases are lost with artificial intelligence, many other ethical issues arise. Many users of Lensa have reported the app changes their facial or body features. A concern that has come up is oversexualization of female users. The generator often produces images of nude women or women in suggestive poses, even when unprompted.
Another concern is the app anglicizing features and having difficulty creating accurate self-portraits of people of colour, often making them whiter. It can also create unrealistic body standards, adding to larger debates around filters and social media use.
Is there hope for AI art?
Although artificial intelligence as-is raises a lot of ethical concerns, it doesn’t mean it can’t improve and have a more positive impact on the art world. It could, for instance, be used as a tool for inspiration. Artists could brainstorm new works or explore different genres or styles.
There is hope for a future where artificial intelligence and artists can collaborate. For this to happen, artists need to be included in discussions and be compensated for any of their work being used by AI art apps or programs. Tag us @palbricart and share your thoughts with us, so we can continue to centre artists in this conversation!
Written by Jillian Simpson and Nina Surugue