Table of Contents
In seconds, with the right prompt, AI can do almost anything. It can paint portraits, write poems, and generate an entire page of visual concepts. With it's improvement in technology and rapid generation, it's no oner that one AI-generated artwork has already won an art competition. Ai-Da, a humanoid robot artist, has exhibited work internationally.
So, asking whether AI can create art isn't what we should be questioning. Instead, we should be looking at what happens to human creativity when machines can do so much of the creating for us.
What counts as art has been a contentious topic for decades, long before Midjourney entered the picture. In 1917, Marcel Duchamp submitted a porcelain urinal titled, Fountain, for exhibition and challenged assumptions that art required beauty or technical skill. More than a century later, generative AI is creating a similar headache.
Philosopher, Alice Helliwell, made an obvious but uncomfortable point. If a urinal can be art, why should something produced by an algorithm automatically be excluded? The distinction of what can be considered art or creativity becomes even harder when we look at how humans create in the first place.
Creativity researcher, Margaret Boden, defines creativity as "producing ideas that are new, valuable, and surprising." Believe it or not, AI can arguably tick all three of these boxes. And as mathematician, Marcus du Sautoy, points out, humans don't create in a vacuum either. Artists absorb other artists, movements, experiences, and cultural references before producing their own. Similar to the process of generative AI.
But there is one thing that AI doesn't have. Life experience. For years, neuroscientist, Adam Green, has been researching how creativity works inside the human brain. The ideas that appear in our minds emerge through networks shaped by our individual experiences. Someone might associate a watermelon with a soccer ball because of something that happened to them years ago. But another person would never consider that connection. Multiply those strange, personal connections across billions of people and you get an enormous diversity of ideas.
AI, on the other hand, draws from a shared system instead. Green's recent research examined more than 370,000 college admission essays written before and after ChatGPT's release. The newer essays contained greater variety of words, yet often contained fewer original ideas. However, despite this finding, creativity experts still rated some of these pieces as more creative anyway leading to the point that while AI may not think differently, it can still sound somewhat creative.
This distinction matters in the classroom. Now, don't get me wrong, the question has come up of "so why not respond to this by banning AI altogether?" But we have to be realistic. Students already have these tools in there pockets, and are using these systems everyday. Therefore, the better question we should be asking is how they should be using them.
Adobe executive, Scott Belsky, describes AI users on a spectrum between "outcome-oriented" and "process-oriented". One person may ask an image generator for a cowboy in space, take the first decent result and move on. Another might generate concept after concept, pull apart compositions, change elements, and use them as a starting point for something distinctly their own.
This way of using the generative AI enables the process of using the tool to be identical, but there is a change in the creative involvement. Plenty of artists are already working this way, whether it be through training algorithms exclusively on their own artwork or using AI to rapidly test concepts before developing them further. When used in this way, the tool acts less of an artist replacement and more of an additional creative medium. But the fear of it being taken too far still lingers
History has seen this panic before. When photography arrived in the 1800s, some feared it would make painters redundant. Truth be told, it didn't. Instead, it helped push painting away from realism and towards a new form of abstraction, and AI could produce its own version of this shift.
However, there is always a catch. Green has argues that engagement in the creative process has value in itself. If we continually hand a bank page and prompt to AI, as artists, we may lose some of the messy thinking that comes with producing different ideas. As he puts it, AI asks us to trade process for product.
For educators, this distinction is worth protecting. Ultimately, as AI continues to evolve, there is no point pretending it doesn't exist. Nor does every use of AI deserve to be treated as a creative achievement. Instead, as educators, we will need to pay closer attention to everything that happens before the finished product. Consider the decisions, experiments, mistakes, references, revisions, and intent behind the piece.
Yes, AI can give us an image in seconds. But teaching students why they made it that way may become far more important to the final product.
