0
0
In the face of image generating AI models, artists and designers around the world are confronted with life changing questions as society recontextualises its relationship with art. In this article, our author takes us between the thin but meaningful boundaries between human and artificial, the individual and capitalism, history and a possible future for the industry.
Lingering for centuries, the definition of art has been debated. The rise of AI-generated art reignited the conversation. While some are skeptical of whether AI creations can be considered true art, human-made art will retain immense value in the future, just as it has throughout history.
Art has long been considered the pinnacle of human expression. It transcends aesthetics, holding profound cultural and historical significance. Traditionally, art has been a reflection of society and its values.
The emergence of generative adversarial networks (GANs) marked a significant turning point for digital art. GANs leverage deep learning to discern aesthetics from extensive image datasets for increased efficiency. This enables the system to autonomously generate new images that adhere to learned aesthetic principles from countless datasets for images and art. However, the emergence of the art generation software blurs the line, making the artistic process feel somewhat redundant.
Art in a historical context
However, creativity isn’t just about producing an image; it’s about the freedom of expression and the artist’s unique perspective. Some say art lies in the creation process itself, while others emphasize the viewer’s interpretation. Human creativity thrives on context. Even the most outlandish prompt fed to DALL-E would be based on human experiences and carry specific meanings. AI might require an unfathomable amount of data encompassing the entire spectrum of human experience to truly understand and fulfill such prompts.
All art builds upon what came before; every artist is a student of art history and its movements. This constant dialogue fuels creativity and contextualizes new works. The biggest challenge for AI isn’t replicating existing styles; it’s about pioneering entirely new art forms and movements that redefine the very concept of art. Consider Cubism, initially ridiculed for its radical departure from traditional forms. Only after our aesthetic sensibilities adjusted, thanks to movements like Impressionism, did we appreciate Cubism’s groundbreaking nature. Innovation starts disruptive and unexpected, eventually becoming a familiar part of our artistic landscape.
This artistic evolution is deeply intertwined with social, political, and cultural contexts. It’s a complex process that can’t be captured by software simply analyzing the brushstrokes of Van Gogh. The act of creation extends beyond the canvas. Van Gogh’s paintings were the concrete expressions of a singular individual pushing artistic boundaries.
This line between technological feats and genuine creativity is crucial.
Humanity sits outside AI’s canvas
“The Next Rembrandt”, a computer-generated painting developed by several computer scientists and developers at Microsoft, TU Delft, and some other firms, was an attempt to “understand the relationship between art and algorithms”. Rembrandt van Rijn is regarded as one of the best European painters, defining the Golden Age of Dutch painting. If “The Next Rembrandt” is viewed based on mechanical features, such as colors and brushstrokes, the AI image looks indistinguishable from the very idea and imagination of Rembrandt’s work. While it is impressive that a computer program can replicate Rembrandt’s artistic vision and physical attributes, it reinforces our existing idea – AI can replicate, duplicate, or imitate. German Philosopher Dorothea Winter writes, “We need to understand how creativity manifests in human life before evaluating if AI can replicate it. DALL-E may generate impressive, seemingly original works based on keywords. However, from a philosophical standpoint, it doesn’t create autonomously. DALL-E essentially remixes existing elements, no matter how sophisticated or visually stunning the results may be.” This freedom, this autonomy, is essential for artistic expression. In other words, AI can not imagine or innovate, for therein lies an absence of freedom and hence, an absence of creativity.
Furthermore, art, at its core, has something metaphysical, a uniquely human quality. It’s about expressing emotions, experiences, and the artist’s worldview. Consider the works of Saadat Hasan Manto, his searing satire reflecting the brutality of India-Pakistan’s partition, or Picasso’s Guernica, a powerful response to the Spanish Civil War. Such art evokes a deep human connection, forcing us to confront the artist’s context and beliefs, forcing us to interpret the artistic expression. On the other hand, AI art risks stripping away these emotional nuances and historical significance – pretty sure that the neural networks we train cannot feel.
AI may be able to mimic artistic styles, but it lacks the human experience and the capacity for true creative expression that will forever elevate human-made art. AI knows all the words but not the song.
This is important because all technological innovations, including art generation by AI, are positive as standalone.
AI in a capitalist context
Ideally, this technology should lighten artists’ workload, most of whom work wage jobs for companies requiring them to produce great amounts of graphics, animations, videos, text, etc., but we are collectively let down by the shackles of capitalism. Technology should be used to help the worker. The problem is that technology is used against the worker: Instead of reducing the workload, it’s used to reduce the workforce and wages. According to the Arts Workers Japan survey, 58% of Japan’s illustrators and writers are fearful of losing their jobs to AI. Similar surveys show a huge number of creative jobs being cut in the upcoming years.
No matter how you look at it, the advent of AI-generated art will disrupt the bottom line for a lot of artists, especially in the digital domain. Independent artists, who already struggle to protect their style from being poached on by human competitors, will now be exposed to models scraping the internet. Artists employed by studios would be at serious risk of being laid off and replaced by models that do the grunt work for longer hours and at a lesser cost. Investment would naturally flow towards purchasing the computing resources and developing the most reliable models, and therefore away from the hands of original artists. What is bland and what caters to the lowest common denominator tends to dominate the popular market in a winner-takes-all economic system, and AI is far better equipped to find an optimal solution for a particular artistic ‘product’ than humans. Designers who are trained in the sector would still be needed to provide some creative directions to the models working on a project, but the development of the next generation of maestros will be stunted if the jobs that act today as steppingstones are swept away. In conclusion, many of the values that we attach to art today: originality, authenticity of emotion, and sociopolitical relevance, are human values and face serious threats from the development of AI-generated art.
