FEATURE
Art Under Fire: How AI Challenges Artistic Authenticity and Integrity
by Kelly LeBlanc
Using generative AI to produce “art” is impressive from a technical standpoint, amusing as a novelty, and maybe even useful for some organizational goals, but at what cost? We must consider its impact on human artists and contemporary art, as well as AI’s role in creating and perpetuating bias.
USING AI TO GENERATE ‘ART’
AI is a useful tool for yielding insights, making predictions, and ensuring 24/7 availability of services such as chatbots. However, its use in the creative domain has a profound effect on artists, their works, and the art field itself. AI has developed new possibilities of creative outputs based on machine learning methods, but it also raises concerns regarding its integration alongside human artists, the differentiation between real and generated artistic outputs, and the potential for bias. The visual arts field, referring to graphic art, studio art across all mediums, and art history, is significantly affected by AI.
HOW DOES AN ART GENERATOR WORK?
Art generators take a text-based prompt and use it to create an image. The prompt can be anything (even an abstract idea), as long as the application can interpret what you want. The AI algorithms are based on image-text pairs. They can recognize, or learn to recognize, the subtle differences between the cubist styles of Pablo Picasso and George Braque, for example. The details of how the AI understands depend on the specific art generator, how it was trained, and what data it used. Hidden parameters affect the output.
IMPACTS ON ARTISTS AND THE ART FIELD
AI tools that can learn from networks, datasets, and other sources blur the lines among art, artist, and machine. This raises questions not only about creative rights, but also about how we define art and creativity. Should machine-generated content be considered art and included in galleries and juried shows? While one could argue that the prompt that is input into these AI systems is a form of creative expression by humans, the output—such as a painting or graphic rendering—does not require any level of skill or training. This use of AI takes away opportunities from professional artists who have spent their life honing their skill set and who rely on commissions and contracts for their livelihood.
Companies may consider the use of AI-generated art to be a cost-saving measure, which could potentially lead to a decrease in demand for skilled graphic designers or other artists. However, this decision could have larger implications, as it devalues the skill, knowledge, and education of human artists in favor of AI. It’s important to consider the broader impact of these decisions on the art community and the value we place on human creativity and expression.
AI-GENERATED BIAS
AI-generated content depends on the quality of the models, datasets, and content that it learns from. This means that AI tools might have biases that are either deliberate or unintentional. The user’s text prompts will influence the output, but machine learning also has a role. AI cannot utilize a knowledgebase that it does not have or hasn’t been trained with because it’s not a sentient entity. In this way, bias can come from training data or from a user’s text prompt. The understanding and mitigation of bias are essential to avoid automating and scaling discrimination in AI-generated content.
Artists who work with organizations may face bias due to instructions, company culture, or design parameters, for example. Thus, organizations must be informed about potential AI-derived bias. Human biases may be easier to notice and can be resolved through communication and understanding, while AI biases can be harder to spot, especially since the public is not privy to the AI training data. This leads to questions about how to find and reduce bias in AI, especially when its origin may not be clear.
ARE THE VISUAL ARTS PROGRESSIVELY LESS VALUED BECAUSE OF AI ART GENERATORS?
AI tools have a negative effect on professional opportunities for contemporary artists, craftspeople, and creators. This negative effect does not imply that AI stops one from improving or developing one’s artistic ability, but rather, it implies that it reduces work and contracts, lowers the monetary value of art in our modern world, and even undermines the creativity and skills required to be an artist. This doesn’t even touch on the impact AI has on the meaning of art, which is a separate topic entirely.
If anyone can type a text prompt, get an image, and use that image however they want, where is the creative merit? Are AI users now artists in a general sense? Is everyone an artist? I’ve been illustrating for decades; I consider it a preferred medium that exemplifies my artistic work. But if I’m feeling silly one day and ask an art generator for an illustrated etching of Bastet standing against the Ishtar Gate in Rococo style, is that art? No. Did it require me to have any artistic skill? No (but having some knowledge of art history helped for the prompt). If I submit my AI-generated sample to a judged show, is that fair? No. (These “no” answers are my opinion and come from my perspective and experiences.)
AI-GENERATED ART FEEDS THE ILLUSION OF REALITY
AI-generated art can risk deception. In certain circumstances, it’s similar to propaganda in that images (such as paintings or photographs) created by AI can mislead the public because of their realistic appearance and often convincing narratives. Social media platforms are a prime example of how this form of generated imagery can be used for misinformation. Instances of this deception may include political messages, pleas for charitable donations, or seemingly harmless content that is used as clickbait. All of this can border on propaganda, if not meet its definition outright.
When the public cannot determine what is real, and there is no attribution to something that was AI-generated, there is a danger that fake becomes fact. It’s not solely the responsibility of AI technology to address this deception, but accountability must be assigned somewhere.
HOW ARTISTS CAN BENEFIT FROM AI
A fair way to look at AI tools is that they can be helpful if they support artists creatively or simplify more mundane tasks. For example, tools such as Copilot, Claude, or ChatGPT could be useful in writing biographical statements or social media posts for advertising an upcoming exhibition. Also, artists might use text prompts for inspiration, in which the text itself becomes a creative component. When artists use AI to improve their work, persona, and professional presence, it creates a harmonious relationship between technology and humans. These partnerships between humans and AI are positive and mutually beneficial, but it’s important to remember that AI tools are not necessarily the ones that generate “art.”
FINAL THOUGHTS
In my opinion, AI in the art world presents more challenges than opportunities. While AI art generators can produce impressive results, they raise questions about the value and definition of art, the role of human creativity, and the potential for bias. The livelihoods of human artists and the whole art industry can be adversely affected by the use of AI-generated art. It is important to consider its ethical implications and to ensure that it is used in a responsible and transparent manner. At the same time, AI (generally, text-based AI) can be a valuable tool for artists, assisting them in their creative process and helping to streamline tasks. Ultimately, the key is to strike a balance between the benefits and risks of AI in the art world and to ensure that human creativity and expression are valued and respected.
|
