Jason Allen’s A.I.-generated work, “Théâtre D’opéra Spatial,” took first place in the digital category at the Colorado State Fair(The New York Times, 2022).
Artificial Intelligence and Painting
In recent years, as artificial intelligence (AI) technology has continued to advance, it has gradually replaced many laborious, mechanical, and low-skilled jobs. The art of painting was once considered a field that was not easily encroached upon by AI and was relatively safe due to its unique professionalism and creativity. However, the AI painting “Edmond de Belamy”, which was auctioned for $432,000 in New York in 2018, and “Theatre D’ opera Spatial”, which came out on top at the Colorado State Art Fair ……, are announcing to people that AI painting as an emerging form of artistic expression, is constantly impacting the traditional art field.
Crawford (2021) argues that artificial intelligence is neither artificial nor intelligent; it consists of natural resources, fuel, manpower, infrastructure, logistics, history, and classification, and is the intersection of the natural and social sciences.
The development of artificial intelligence (AI) painting can be traced back to the 1960s when researchers began to explore how computers could be used to generate artwork. The early stages of computer painting focused on generating lines and basic geometric shapes, such as the work of computer artists like Vera Molnar and Michael Noll. As computer technology and algorithms continued to evolve, the ability and accuracy of AI painting continued to improve. Today, AI painting technology is very mature and can generate extremely realistic and detailed images and videos. AI painting has gained wide application not only in the art field, but also in film production, game development, virtual reality, and others. At the same time, some AI art trading platforms based on blockchain technology have emerged, such as Nifty Gateway and SuperRare.
figure 1. Vera Molnar, (De´s)ordres, 1973
figure 2. SuperRare, https://superrare.com/
AI painting has undoubtedly lowered the threshold of art creation and increased productivity and inclusiveness. AI painting software only requires users to enter some keywords to instantly produce a painting that would take the artist a day or more of labor to complete. That makes the artwork more accessible. Although AI technology can not yet perfectly reproduce the descriptive text typed by the user, people still revel in the fun of participating in generating art. Such a trend has also led many art creators and scholars to worry that people will be domesticated by AI in the process of using it in reverse, thus degrading their creative abilities. Moreover, for those companies whose purpose is to make a profit, AI painting can bring them lower economic costs and faster output. It means that the emergence of AI painting may make it difficult for many artists to make a living.
“Reproduction technology frees what is reproduced from the traditional realm, and because it makes so many copies, it replaces the unique presence with a multitude of copies. ”
– Benjamin Walter
Users boycott AI paintings by swiping
Artstation is a comprehensive visual arts site that allows professional artists in games, film, media, and entertainment to display and share their art. In December 2022, the site’s home page was flooded with the same image with a“No AI Art” logo.
figure 3. Alexander Nanitchkov, https://www.artstation.com/artofinca
The design of this picture is very simple, but it is also a work of art designed by a live artist. Alexander Nanitchkov, a Bulgarian illustrator, produced the image. Each of his illustrations on ArtStation is watermarked with this loge, in case his work is automatically referenced by the AI. On December 5th Alexander tweeted: Current AI ‘art’ is created on the backs of hundreds of thousands of artists and photographers who made billions of images and spend time, love, and dedication to have their work soullessly stolen and used by selfish people for profit without the slightest concept of ethics(Edwards, 2022).
figure 4. Nanichkov’s ArtStation page, https://www.artstation.com/artofinca
The image later sparked a backlash against AI painting on ArtStation. Many of the artists who took part in the protests were hostile to Ai painting and did not want to see an art website filled with AI-generated“Non-art” content, more do not want their work by AI capture plagiarism. Others are less hostile, complaining that ArtStation lacks a systematic policy for AI-generated content.
In the face of user protests, ArtStation officials first removed the initial large batch of swiped images but failed to stop more users from participating in the swipe. In a statement about the removal posted to Twitter, ArtStation said, “For the sake of site availability, we are reviewing posts that violate our terms of service. We understand the concerns about AI and its impact on the industry. We will be sharing more information about improvements that will give users better control over what they see and how they use ArtStation in the near future.” (Weatherbed, 2022)
Figure 5. ArtStation, https://www.artstation.com/?sort_by=community&dimension=all
Similarly, on March 6, 2023, Lofter, an original community platform, launched an “avatar generator” feature that allows users to automatically generate drawing images by entering keywords. The launch of this feature aroused strong opposition from creators in Lofter. Although Lofter officials repeatedly stressed that the original purpose of the feature was to facilitate users without drawing the ability to create their favorite avatars, there was no profit purpose. And the training data set used is all from open source data, not infringing on the rights of original artists. However, a large number of content creators and artists on the platform still started to follow the practice of artists on ArtStation last year and said: “no to AI painting”. Some creators even stated to leave the platform and cancel their accounts to show their attitude. By observing the users who objected, they considered that the main reason for their protest was that using the “avatar generator” to generate images was against Lofter’s positioning as a platform for creators, and the algorithm might also be suspected of plagiarizing original works on the platform. As content creators, they expressed their reluctance to use their work to feed the AI, and they wanted to oppose any possible infringement.
Figure 6. Lofter, https://www.lofter.com/trend?act=qbview_20130930_01
Ethical Issues in Artificial Intelligence Painting
AI painting is essentially machine learning and requires training by borrowing image data sets from existing artworks. This data is available via the Internet without the permission of the original creator. While AI painting is not a direct copy of an existing work, it does draw inspiration from the artwork that trained it. The original creators can neither take back their work nor prevent the AI from seizing more.
AI paints might be used to forge pictures, such as documents and photographs. Furthermore, AI-generated pictures could be utilized to make highly fabricated movies that are able to convey disinformation or propaganda (University Wire, 2023). Such activities might have significant ramifications, such as fraud and deception. Last year, for example, a fake video of Ukrainian President Zelenskyy went viral on social media, urging Ukrainians to surrender.
- Data Privacy
AI paintings often require the use of large amounts of data resources to train models, and these resources may need to resort to third-party cloud computing platforms or storage services. If these service providers do not have sufficient data privacy protection measures, it may lead to the leakage of users’ privacy. Therefore, AI paintings not only draw from existing paintings, but also may contain users’ personal information, sensitive information, and other copyright information, etc.
- Copyright Attribution
Traditional paintings are created by human artists. Their copyright and intellectual property rights are attributed. In contrast, the ownership of AI paintings is difficult to prove because AI technology models, programmers, artists, and end users may all have an impact on the creation process of the work.
- Awareness issues
AI paintings are created by machine algorithms, which do not have human perception and consciousness and may lack uniqueness, creativity, and expressiveness compared to real works of art. As a result, many artists and scholars believe that AI paintings may not have real artistic value and may lead to the downgrading and devaluation of art and creativity.
- Cultural and Ethical Issues
The cultural and ethical issues raised by AI paints also need to be taken seriously. With the prevalence of AI painting, the threshold of drawing has been lowered because people can “create” their paintings as they wish. For example, after Lofter launched its “avatar generator”, it was found that someone used the Ai painting function to generate images with sensitive or inappropriate content, such as violence, pornography, or racism. In addition, when AI drawing first became popular, the portraits it generated were either abstract or anime and did not look like real people at all. But just a few months later, AI painting technology can now grab jobs with real cosplayers. That is a huge advancement in AI technology. However, AI painting out characters too much like real people may also cause problems such as infringement of portrait rights, disinformation, dissemination of pornographic and violent material, and so on. Therefore, there is a need for appropriate laws and policies to regulate such AI-generated image behavior.
figure 7. Kanyon Industries, AI coser, https://twitter.com/auditore_k/status/1625034250423787521
What can we do?
- Government should establish and improve laws and regulations to guide and regulate the application of artificial intelligence. As the pace of technological development is increasing, the government should also keep updating more appropriate policies. More and more detailed policies are needed to regulate the development and application of AI technologies to prevent the impact on public interests and humanistic values due to the overuse of technology. For example, (1) integrating ethics into the entire life cycle of AI and providing ethical guidelines for its development in various fields; (2) establishing laws and regulations related to copyright protection and privacy protection to protect the reasonable rights and interests of people under the impact of AI technology; and (3) paying attention to the cultural and ethical issues caused by AI, particularly the need to regulate the use of minors in this reg, etc.
- The platform should implement its responsibility as the main body. As the most direct benefit recipient in AI painting technology, the platform is also the most direct subject of responsibility. Therefore, the platform must consciously follow the dual constraints of laws and regulations, and ethics, and improve the regulations related to the use of AI technology in the platform. At the same time, the platform should also pay attention to users’ demands and balance technical development and humanistic values to promote the development of AI technology for good.
- Artists can explore new paths of art creation in conjunction with technological developments. Technologies such as artificial intelligence seem to be only getting more powerful at the moment. Therefore, it is important for artists not to focus all their attention on opposing the use of AI technology. As they argue, the paintings produced by AI are based on the learning and imitation of human paintings, while human imagination is infinite. Therefore, artists still have an irreplaceable position, and even the thoughtful and emotional works of artists are made more precious under the backdrop of AI creations. At the same time, in the face of the threat of AI technology, human creators need to continue to learn and maintain a love of life to create better works. In addition, artists may also explore a new path of human-machine collaborative creation, which is not a great attempt.
- For us ordinary users, understanding more laws and regulations, not abusing AI paintings, always keeping a clear head, and paying attention to the ethical issues in the use of AI technology are helping the development of AI and society.
For art creators, I think there is no need to feel overly anxious because of the high ease of use and low cost of AI. The artworks created by humans reflect specific historical backgrounds, cultural contexts and are rich in emotions, aesthetic values, and profound meanings, which are difficult to be reached by AI paintings at present. Moreover, even if AI can do part of the basic painting work well, human participation is still needed to make it an excellent work. Using AI to assist and establishing a healthy partnership with it can free people from repetitive instrumental labor, which is no less an advancement in the art. In any case, AI is changing our future in spectacular ways, and all the challenges we currently encounter are because we haven’t known this tool long enough to have a complete methodology for using them. Once we regulate the use of AI and improve the regulations, AI may be just like any other tool our ancestors have used, just for us humans.
Artstation – Alexander Nanitchkov. (n.d.). Retrieved April 14, 2023, from https://www.artstation.com/artofinca
Benjamin, W. (1982). The Work of Art in the Age of Mechanical Reproduction. In Modern Art and Modernism: A Critical Anthology (1st ed., pp. 217–220). Routledge. https://doi.org/10.4324/9780429498909-39
Benjamin Williams. (2021). Painting by numbers: Copyright protection and AI-generated art. European Intellectual Property Review, 43(12), 786–792.
Collect digital art. SuperRare. (n.d.). Retrieved April 12, 2023, from https://superrare.com/
Crawford. (2021). The Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press.
Edwards , B. (2022, December 16). Artists stage mass protest against ai-generated artwork on artstation. Ars Technica. Retrieved April 15, 2023, from https://arstechnica.com/information-technology/2022/12/artstation-artists-stage-mass-protest-against-ai-generated-artwork/
Is AI painting a bright or bleak future? (2023, Jan 24). University Wire http://ezproxy.library.usyd.edu.au/login?url=https://www.proquest.com/wire-feeds/is-ai-painting-bright-bleak-future/docview/2768792276/se-2
Kanyon Industries. (2023, February 13). It’s a pity that Twitter can only post four pictures at a time #ai #stablediffusion #arknights #Texas #アークナイツフェス #明日方舟. Twitter. Retrieved April 16, 2023, from https://twitter.com/auditore_k/status/1625034250423787521
LOFTER. (n.d.). Trend | LOFTER. Retrieved April 16, 2023, from https://www.lofter.com/trend?act=qbview_20130930_01
Molnar, V. (1976). (Des)Ordres [Dessin ordinateur au plotter sur papier Benson]. VERA MOLNAR. Retrieved April 11, 2023, from https://www.oniris.art/en/exhibitions/16-vera-molnar-alchimie-des-lignes/
Roose, K. (2022, September 2). An a.i.-generated picture won an art prize. artists aren’t happy. The New York Times. Retrieved April 12, 2023, from https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html
Translated by Content Engine LLC. (2022). Thousands of artists protest at ArtStation against AI-generated images (English ed.). ContentEngine LLC, a Florida limited liability company.
Weatherbed, J. (2022, December 24). Artstation is hiding images protesting AI art on the platform. The Verge. Retrieved April 16, 2023, from https://www.theverge.com/2022/12/23/23523864/artstation-removing-anti-ai-protest-artwork-censorship