Artist or Algorithm? On AI and Its Role in Art

The earliest definition of the word artist dates from the 13th century, due in part to the Old French word art - described simply as "skill as a result of learning or practice." The word also carries a suffixed form of the root "ar-", meaning "to fit together" (Etymonline).

As simple as these roots and meanings may seem, in our digital age defining art is a paradoxical and complicated endeavor. The proliferation of technology for editing and making creative works have pushed the boundaries of artists' ideation and creation processes. With tools like Photoshop and Illustrator, for example, artists gained unprecedented freedom to experiment with colors, shapes, and textures, blurring the lines between traditional and digital art forms. The accessible nature of technology, moreover, has somewhat democratized the creative process, allowing anyone with access to a computer to dabble in artistry.

Generative AI in particular has made its name as a groundbreaking frontier of such technology forms, amazing users and arousing fervent debate within the arts industries. With artificial intelligence in the picture, how does technology fit into the original definition of art, and what exactly constitutes an artist in our modern society? 

How it works: Generative AI and Image Generation

Watercolor painting of a dark and stormy night in NYC in the year 2004

Generative AI refers to algorithms that are trained to generate new content, such as images, music, or text, based on patterns learned from existing data. Rather than relying on explicit instructions, generative AI operates autonomously, producing original outputs that mimic the style and characteristics of inputs they are given.

Most generative AI platforms designed for art begin by prompting the user to write a specific request - for example, "watercolor painting of a dark and stormy night in NYC in the year 2004.” They then render new images from those requests, usually in a short timespan and at a low or free cost to the user. The user is also able to request images to be generated in certain styles, such as "impressionist" or "modern". As such, the images are largely dependent upon the wording and detail that the user provides in their request. 


Images outputted by Generative AI are sourced from vast datasets, including the internet, digital archives, and proprietary databases. Although some of these sources are published for the purpose of public use, their owners do not often explicitly define them as usable by AI models. They contain a diverse range of visual content, from photographs and illustrations to paintings and sketches. By analyzing and synthesizing this data, generative AI algorithms learn to generate new images that resemble those in the training data.

Who Owns AI Images?

Ever since the introduction of these models, artists have persistently contested the intellectual property rights and ethics of using copyrighted material to create AI-generated art. As these AI models are trained on large amounts of data, often sourced from the internet, the images themselves may include copyrighted or borrowed material without proper credit. This raises many questions about who owns these images, as well as why artists are not being compensated for the use of their work in the AI image generation process. 


As for assessing the creative strength of AI-generated art, it is reasonable to state that AI algorithms lack empathy, emotions, and subjective experiences that human artists bring to their work. As artists are revered for their abilities to evoke emotions and change perceptions, the use of AI images potentially diminishes their individual identities and threatens the potential for art to reflect and express the complexities of humankind. 

It is also important to consider whether AI images reflect the biases of existing creators and perpetuate existing inequalities in art, given that they are generated from various man-made sources. This is an extremely difficult question to answer–as different image generators have been produced by different creators and companies; what a user is and is not allowed to create depends largely upon the generator they use. This article, for example, demonstrates how OpenAI's image generation model Dall-E 3 blatantly refuses to generate any images of politicians, public figures, or symbols that carry reputations for being widely misused or misunderstood. With that said, all art arguably holds the potential to be misused for malicious purposes, and the extent to which it may perpetuate inequalities present in art that was manually created would be nearly impossible to quantify.

Pride or Practicality?

One of the most common arguments which AI art supporters bring is the fact that Generative AI models offer time and cost efficient solutions for many business needs. If you could generate many free images in seconds that are tailored to your specific needs and stylistic preferences, why would you spend the time and money to commission an artist to manually create works which you may not find satisfactory?


On the other hand, even if art produced by AI and by humans require the exact same resources and return equally satisfactory outputs, would it do the definition of art as "skill as a result of learning or practice" more justice if we were to choose the human-produced work? Regardless of whether AI could outpace human artists in terms of speed and cost, would these factors overwrite all the hours of practice which human artists pour into honing their crafts and developing their distinct styles?


How you answer these questions may reveal how you define who an artist is. Although an artist is a creator, they can also be considered a curator, marketer, and storyteller. In a world driven by social media and digital content, artists are expected to showcase not only their finished works but also their creative process, their studio spaces, and even glimpses of their personal lives. The ability to market and promote one's art effectively has become as essential as the art itself, shaping the identity and success of modern artists in unprecedented ways.


While technology continues to shape the artistic landscape, it is essential to uphold the principles of creativity, ethics, and human expression, ensuring that art remains a reflection of our shared humanity. While AI may offer efficiency and convenience, it is our collective values and principles that constitute the heart of art. Whether we choose human-produced works or AI-generated marvels, let us ensure that the freedom of creativity and human expression remain at the forefront of our endeavors.

Krystal Wu

Krystal is a junior majoring in Data Science and minoring in Creative Writing. She enjoys alt rock and J-pop music, melodramatic and experimental poetry, nice cafes, fashion, and thrifting. Her hometowns are Ellicott City, MD and Taipei, Taiwan.

Previous
Previous

World Design in Video Games and Why It Works, from Pokémon to Hitman

Next
Next

My Cousin, The Director