Art In The Wake Of AI
From Brushstrokes to Bytes
Technology and Humanity
Since the dawn of civilization, we humans have harnessed the power of technology to understand and shape the world to our benefit. It allows us to record, express, and communicate our thoughts and experiences. Each technological advancement, from the invention of the printing press to the dawn of the digital era, has altered how we interact with reality to express ourselves.
Today, we stand on the precipice of a new era marked by the relentless progress of Artificial Intelligence (AI), blurring the lines between creator and tool and challenging our perceptions of artistry and creation.
Each human invention comes with its dilemmas. The camera revolutionized our perception of artistic authenticity, and initially, conservatives didn’t think of photography as art. Similarly, we find ourselves in the wake of generative AI, compelled to ask: should we consider AI’s creations as art?
Artificial Intelligence
AI has recently flourished through machine learning and neural network advancements, emulating human intelligence using silicon and statistical processes. Technically, AI is a catch-all name for computer systems that perform tasks that would normally require human intelligence, such as visual perception, speech recognition, reasoning, and text and image generation. It is a technology with learning and problem-solving capabilities that allow it to make impactful decisions and generate creative output autonomously.
We know AI systems are not perfect yet. However, Generative AI products like ChatGPT and MidJourney are poised to disrupt the domains of artistic and intellectual production, kickstarting copyright nightmares and fueling existential crises from white-collar workers who thought their creative work was safe from automation. One example is the Writers’ Guild strike, having writers demand guardrails around AI’s encroachment on their work. Another example is Getty’s lawsuit against Stability AI for using their images for training their AI model, Stable Diffusion.
Whether you approve of the strike or the lawsuit, it is evident that AI represents a pinnacle in our ongoing pursuit of creating autonomous decision-making systems. It is a potential paradigm change where humans become spectators rather than the driving force of creation and discovery, forcing us to re-evaluate the future of creativity and art.
Technology and Art
Technology is to art what instruments are to music. Every technological revolution has created new modes of expression and has fueled hot debates between modernists and conservatives.
For example, in the dialogue Phaedrus, Plato thought writing “will create forgetfulness in the learners’ souls, because they will not use their memories; they will trust to the external written characters and not remember of themselves.”
I wonder if Plato is turning in his grave today as students use ChatGPT to write their class assignments.
Another example is how the Industrial Revolution gave Warhol and Duchamp the liberty to explore art concepts that enraged some conservatives. For instance, Marcel Duchamp’s “Fountain” (1917) is a ready-made urinal he signed with a pseudonym. This piece challenged traditional notions of art by presenting a mass-produced object as a work of art. Similarly, Andy Warhol’s iconic “Campbell’s Soup Cans” series (1962) leveraged the imagery of mass-produced consumer goods to create art, reflecting and critiquing the consumerist culture that the Industrial Revolution engendered. Both Warhol and Duchamp used the technology, products, and ethos of their times to expand the boundaries of what could be considered art, much like what AI technologies are poised to do today.
So, if the mass production of goods and services shook the foundation of art, think of what AI-generated text, images, video, and music will do to it.
What is Art?
The purpose and definition of art have been debated for centuries among philosophers, artists, and scholars.
Are cavemen paintings art?
Is Da Vinci’s Mona Lisa art?
Is Monet’s impressionist sunrise art?
What about Duchamp’s Fountain? Is this art?
Is Warhol’s Campbell Soups art?
And finally, what about this — an AI-generated image of The Mona Lisa holding a Campbell soup reimagined by Monet?
According to the Oxford Dictionary, art is:
The expression or application of human creative skill and imagination, typically in a visual form such as painting or sculpture, producing works to be appreciated primarily for their beauty or emotional power.
So, AI is not human; therefore, it’s not art. As the definition above says, art is the application of human creative skill!
Really though? Let’s put AI aside and test this concept.
We can find many moving creations and performances in nature. There is a male puffer fish in Japan that creates fascinating mandala-like shapes on the bottom of the ocean to seduce the female puffer fish. The red-capped manakin bird performs an impressive acrobatic dance as part of their mating ritual. The peacock evolved mesmerizing feathers, becoming the ultimate example of pure form and no function — as beautiful as those feathers are, the peacock can’t fly. Some elephants have been trained to create paintings.
Who would dare say something like the puffer fish’s creation or any other is not art just because it was not made by a human? In view of such natural creativity, traditional definitions of art crumble, and I’m reminded of Ambrose Bierce’s provocative stance in The Devil’s Dictionary:
Art (noun). This word has no definition.
As an object, art is an elusive concept, and if you think otherwise, you probably haven’t been to the MoMa. Once, a post went viral with an impromptu social experiment where someone intentionally dropped a glove on the MoMa’s floor, and visitors carefully walked around it, unsure if it was part of the exposition.
But what if we define art as a verb? In this way, art is anything that moves you or anything you pursue with enough passion and dedication to move others. It’s a process with two sides: creation and consumption.
Creating Art — AI & Humans.
Unlike human creativity, it’s relatively easy to explain how Generative AI creates art. It uses neural networks to analyze data, find patterns, and create new content. Neural networks are inspired by the human brain structure, consisting of layers of interconnected nodes (neurons) that process data collaboratively. In layman’s terms, the AI model ingests data — images, sound, and text — to produce its output. It allows these models to turn text inputs into images, music, narratives, or even video and images into text.
But how do humans create art? As with the attempt to define art, you can find many definitions and approaches to answer this question. They range from the scientific, describing how the brain operates, to the metaphysical, like artists shedding light on their creative process. Let’s take a few examples of creative giants.
In her book ‘The Art of Fiction,’ Ayn Rand provides her insights on the role of the subconscious in the process of writing. She believes the subconscious mind stores all the knowledge and experiences a person has absorbed throughout their lives. Authors call upon this accumulated knowledge to compose their prose, sifting through every piece of information before using the conscious mind to polish the work.
Salvador Dali wrote ten guidelines in his collection of essays, ‘Fifty Secrets of Magic Craftsmanship,’ for anyone who wants to become a painter. One of them states that an aspiring painter must ‘first of all, learn to paint like the old masters. Then you’ll be able to paint the way you want (like yourself) and everyone will respect you.’ In other words, you must learn from artists who came before you so you can one day develop your own creative output.
In an interview, Steve Jobs said, ‘Picasso had a saying…good artists copy, great artists steal, and we [Apple]…have always been shameless about stealing great ideas.’ Steve Jobs ‘stole’ the concept of the Macintosh computer after seeing a similar computer at Xerox’s Palo Alto Research Center. Ironically, he would later sue Microsoft and Android for ‘stealing’ Apple’s ideas. Also, it’s unclear whether Picasso actually said that quote or if it derived from T.S. Eliot’s book The Sacred Wood: Essays on Poetry and Criticism written in 1920: ‘Immature poets imitate; mature poets steal.’
In conclusion, for any human creative endeavor, it’s impossible not to be influenced by the world around you and the information you consume, and so it goes for AI.
Humans and machines alike are a mashup of what they choose to let into their lives (or hard drive). The quality of any output is derived from the data the intelligent agent ingests.
Consuming Art — AI & Humans
How does AI consume art? For AI models, art is not an experience but an object — data. Through an emotionless process, AI models analyze bits to identify patterns and generate outputs. Their consumption of art is devoid of any cultural, historical, or emotional context.
Companies crawl the web to scrape data and train their AI models with it. However, this might not be as feasible in the near future. Training models with the indiscriminate use of data might deteriorate these tools as more AI-generated content dominates the web. The authors of the paper, ‘The Curse of Recursion,’ explore this scenario. In their words — What will happen to GPT-{n} once LLMs contribute much of the language online? This can lead to model collapse, a degenerative learning process where models start forgetting improbable events over time as the models become poisoned with their own projection of reality.
Imagine a pianist capable of playing a wide range of music genres. If, over time, the pianist only gets feedback for one genre, like pop music, he or she will only play pop music, forgetting other genres like blues, jazz, or classical music. Similarly, model collapse represents AI’s constriction as it starts producing a monotonous echo of its training data, potentially eroding the rich tapestry of artistic expression. Models will become biased, and their output will conform to the mean.
For humans, art is a deeply emotional process. We consume art because it moves us and makes us feel alive through our emotions. It becomes a union among humans. That is the beauty of art. You discover that your longings are universal longings, that you’re not lonely and isolated from anyone — you belong. This profound connection reaffirms our collective consciousness and shared human experience, a nuance that AI cannot replicate.
In computer science, there is a concept called GIGO — garbage in, garbage out. It reminds us that in any system, the quality of output is determined by the quality of the input. This not only applies to computers but also to humans. You become the information you consume. So, just as programmers need to be aware of what data they use to train their AI models, humans should scrutinize what art they decide to consume.
Art in the Wake of AI
In a world saturated with AI content, human-generated art will not be just a preference but a necessity. It will be an essential data point to represent our cultural richness and diversity, which will make AI a better tool for future generations.
AI is here to stay, and it has the potential to be a revolutionary technology that can improve the world in unimaginable ways. Its role in art generation will bring innovation and disruption as other technologies have done in the past. So, how should we react to it?
Already, tools are being developed to fight AI models, protecting the content-creator’s data before being uploaded to the web. One example is Nightshade, an open-source tool that alters pixels in a way that is invisible to the human eye but detrimental to the model, forcing developers to retrain their models from scratch. In my opinion, this won’t stop companies from developing AI and will only result in more expensive services and will be costly for the environment.
Rather than fighting AI, we should embrace it as a tool that can foster human creativity. As part of a capitalistic society, I believe in the power of the consumer. Therefore, it’s important that we, consumers of art, embrace and demand a future with data dignity, a system where individuals are compensated for the use of their data, especially when it’s used to train AI. This, in turn, will create a world where human creativity is the most valuable and priced asset.
To be human means that you have an innate desire to create. Creativity — in whatever form it takes — is a fundamental aspect of being human. Just because AI can create images, text, video, and music doesn’t mean that humans will stop creating art, just as humans didn’t stop playing chess when computers defeated grandmasters.
So, what is art in the wake of AI? Art will remain the same — an elusive concept, an emotional and human process.
One thing is certain. For there to be art, there needs to be a human in the loop, whether it is in the creation or consumption part of it. Without a viewer’s eyes, a painting becomes blotches of color scattered on a canvas. Without a beating heart, a story becomes hieroglyphic characters on a page, words reduced to statistical probabilities. Without a listener’s ears, music is just vibrations in the air.
Art is the experience of becoming.
