Lead image: Creativity in the 17th century as interpreted by an AI art bot. Creativity has been a fundamental aspect of human society since the dawn of civilization. It has been and continues to be the cornerstone of our ability to innovate, communicate, and express ourselves through art. Although we are often united in reaping its practical or aesthetic rewards, there is a never-ending struggle over who or what (in the case of AI art) can lay claim to the title “creative.” Is this title reserved for only humans? Is it tied to originality? And if so, are any creative works truly original and devoid of existing influences? None of these questions has easy answers. Reinventing The Familiar: Creativity is About Finding New Perspectives “Creativity is a remix”. The words of Kirby Ferguson in his decade-old TED talk. A presentation that audaciously challenged the foundations of our understanding of creativity and originality in art (and all aspects of human endeavor). https://youtu.be/zd-dqUuvLk4?embedable=true His argument was as follows: The act of creation, be it art, music, literature, or anything else, is fundamentally built on the principles of copying, transforming, and combining. Ferguson demonstrated how the music legend Bob Dylan built his early work by referencing the melodies and structures of traditional folk songs. The result was a sound that was new and fresh but made up of pieces of old songs. Away from Ferguson’s example, there is no shortage of contemporary artists in the world of paints and canvases who attribute their unique styles to the works of artists from older generations. An excellent example is Kehinde Wiley, a wildly successful and insanely skillful African-American painter who is known for creating pieces of black figures by referencing the Old Master style. Does it matter that he references old work? Is his creative proficiency in question as a result? Not at all! Everything about Wiley’s work exudes peak creativity. The elegance of his subjects, the masterful brushwork, and mesmerizing use of colors. Everything! Looking at these examples, one can argue that the essence of creativity is taking existing elements and combining them to create something new — new perspectives. What About AI-generated Art? The technology behind AI-generated art is built on the same principles of copying, transforming, and combining. Machine learning algorithms are trained on vast datasets of images, including those of existing artworks, and then they use that knowledge to generate new art. The AI-generated piece is not simply a copy of existing content. It is a new creation, built from the building blocks of existing content. What makes these works of art different is how they combine these elements, the context in which they are presented, and the emotional impact they can have on the viewer. They reflect the biases and preferences of the algorithms or training data and the approval of the human who curates and presents the final product. But much like the case with the invention of photography, the emergence of AI art has spurred a frenzy of backlash in the lair of artists and art critics. They argue that AI is a string of emotionless code, and its lack of human touch signals an impending death sentence for true creativity and art, by extension. If that is true, then AI is surely the enemy. Death to the machines! End of debate! Well, not so fast. Beyond The Machine — AI Art is More than Just Code The argument that AI-generated art lacks human input and therefore offers nothing in emotional resonance doesn’t present the complete picture. Typically, creating AI art is a collaborative effort between the artist and the machine. And the latter serves as a tool that facilitates the creative process, not one that replaces the artist altogether. Midjourney, one of the most popular AI art generation programs today, creates images based on textual prompts provided by users (humans). Basically, the user imagines a prompt, enters it, and the program generates a set of four images. From this point, it is up to the user to select the most compelling or accurately rendered output(s) based on the provided prompt. Look at that! “Lacks human input” is not so accurate, after all. If this were a game of call-a-spade-a-spade, “involves human input” would be the only correct answer. So, what is so bad about AI-generated art? Looking Deeper Into The AI Art Resistance Although AI art scores some points for sharing underlying similarities with its human counterpart, it is still far from gaining favor among a growing jury of artists, art critics, and enthusiasts. Frankly, this fact comes as no surprise, as the technology continues to weave a sticky web of hard-to-navigate issues. Let’s look into some of them. The Authorship Conundrum The issue of authorship in AI art is complex and challenging, as there is no standing protocol for attributing creative ownership to a particular artist or individual. Furthermore, as some AI art programs learn and evolve — further pushing the limits of machine sophistication, deciding the extent of human involvement becomes even more challenging. What this means for all is a brightly burning flame of unending ethical and legal debates. In the interim and maybe long run, a potential solution is to consider viewing AI art for what it really is — a collaboration between humans and machines. The individuals who create the original algorithm or provide prompts receive the creative credits, while the AI programs go into the tool/collaborator category in the creative process. Sadly, suggestions are not enough. We are yet to reach any consensus on acceptable ways of defining ownership in the context of machine-generated art. Hence, the resistance remains. Intellectual Theft Yet another complex and multifaceted challenge powering the resistance to AI art. As AI programs can generate artwork strikingly similar to existing works of art, concerns about plagiarism have become increasingly pertinent. The issue of mimicry is particularly pressing, as AI-generated pieces that closely resemble the works of known artists could be construed as intellectual theft if they are marketed and sold without appropriate attribution or permission. Also, the fact that AI art programs run on data sets that include existing copyrighted artworks raises further concerns. Without ethical guidelines and best practices within the field, including standards of transparency and consent, attribution, and protocols for utilizing existing and copyrighted material, AI art stands a long way from winning over the resistance. Good Old Evolution As creatures keen on survival, humans are biologically hardwired to resist change — the unfamiliar. On the one hand, this is understandable. On the other, it is maybe responsible for our resistance to the new AI art technology. Think of it; until recently, the art domain remained exclusive to humans. So much so that when thinking of art, the mind immediately conjures images of trained hands and gifted minds as they transform time, patience, mistakes, and emotions into masterpieces. Pure sorcery! We even go as far as spending years and thousands of dollars in tuition fees to learn the inner workings of this sorcery. What would it mean to make room for a system that has reduced it into machines and algorithms? What happens to the art degrees and years of devotion? How about the artist’s feelings and career security? Without “good” answers, questions like these aren’t just questions. They feel like threats. Whether or not AI art programs pose any threats seems to be secondary to the feeling of being exposed to such threats. And since evolution has designed our survival instincts to kick in the wake of threats, making efforts to understand AI art and identify its true risk status might be vital in easing the resistance. Parting Thoughts The point of this post isn’t to argue that AI art programs are just as creative as humans. That would be a futile mission for many good reasons. One of which is the fact that humans are capable of being dynamic. We are also capable of experiencing powerful emotions that influence our creativity. The same isn’t true for even the most advanced AI models. However, considering the shared similarities with the human creative process, there is a case for rethinking our understanding of creativity and AI-generated art. If art serves to create beauty, evoke emotions or drive narratives, should the matter of how it comes to exist take from or add to its creative essence? Should it matter that it is human-made or ai-generated? Should we focus on what makes AI art imperfect to justify the resistance? Or do we make necessary improvements and explore a new path to limitless creative possibilities? What say ye, brethren? Let me know in the comments. Also published here. Lead image: Creativity in the 17th century as interpreted by an AI art bot. Lead image: Creativity in the 17th century as interpreted by an AI art bot. Lead image: Creativity in the 17th century as interpreted by an AI art bot. Creativity has been a fundamental aspect of human society since the dawn of civilization. It has been and continues to be the cornerstone of our ability to innovate, communicate, and express ourselves through art. Although we are often united in reaping its practical or aesthetic rewards, there is a never-ending struggle over who or what (in the case of AI art) can lay claim to the title “creative.” Is this title reserved for only humans? Is it tied to originality? And if so, are any creative works truly original and devoid of existing influences? None of these questions has easy answers. Reinventing The Familiar: Creativity is About Finding New Perspectives “Creativity is a remix”. The words of Kirby Ferguson in his decade-old TED talk. A presentation that audaciously challenged the foundations of our understanding of creativity and originality in art (and all aspects of human endeavor). https://youtu.be/zd-dqUuvLk4?embedable=true https://youtu.be/zd-dqUuvLk4?embedable=true His argument was as follows: His argument was as follows: The act of creation, be it art, music, literature, or anything else, is fundamentally built on the principles of copying, transforming, and combining. The act of creation, be it art, music, literature, or anything else, is fundamentally built on the principles of copying, transforming, and combining. Ferguson demonstrated how the music legend Bob Dylan built his early work by referencing the melodies and structures of traditional folk songs. The result was a sound that was new and fresh but made up of pieces of old songs. Away from Ferguson’s example, there is no shortage of contemporary artists in the world of paints and canvases who attribute their unique styles to the works of artists from older generations. An excellent example is Kehinde Wiley , a wildly successful and insanely skillful African-American painter who is known for creating pieces of black figures by referencing the Old Master style. Does it matter that he references old work? Is his creative proficiency in question as a result? Kehinde Wiley Old Master Not at all! Not at all! Everything about Wiley’s work exudes peak creativity. The elegance of his subjects, the masterful brushwork, and mesmerizing use of colors. Everything! Looking at these examples, one can argue that the essence of creativity is taking existing elements and combining them to create something new — new perspectives. What About AI-generated Art? What About AI-generated Art? The technology behind AI-generated art is built on the same principles of copying, transforming, and combining. Machine learning algorithms are trained on vast datasets of images, including those of existing artworks, and then they use that knowledge to generate new art. The AI-generated piece is not simply a copy of existing content. It is a new creation, built from the building blocks of existing content. What makes these works of art different is how they combine these elements, the context in which they are presented, and the emotional impact they can have on the viewer. They reflect the biases and preferences of the algorithms or training data and the approval of the human who curates and presents the final product. But much like the case with the invention of photography, the emergence of AI art has spurred a frenzy of backlash in the lair of artists and art critics. They argue that AI is a string of emotionless code, and its lack of human touch signals an impending death sentence for true creativity and art, by extension. If that is true, then AI is surely the enemy. Death to the machines! End of debate! Well, not so fast. Well, not so fast. Beyond The Machine — AI Art is More than Just Code Beyond The Machine — AI Art is More than Just Code The argument that AI-generated art lacks human input and therefore offers nothing in emotional resonance doesn’t present the complete picture. Typically, creating AI art is a collaborative effort between the artist and the machine. And the latter serves as a tool that facilitates the creative process, not one that replaces the artist altogether. Midjourney , one of the most popular AI art generation programs today, creates images based on textual prompts provided by users (humans). Midjourney Basically, the user imagines a prompt, enters it, and the program generates a set of four images. From this point, it is up to the user to select the most compelling or accurately rendered output(s) based on the provided prompt. Look at that! “Lacks human input” is not so accurate, after all. If this were a game of call-a-spade-a-spade, “involves human input” would be the only correct answer. So, what is so bad about AI-generated art? So, what is so bad about AI-generated art? Looking Deeper Into The AI Art Resistance Looking Deeper Into The AI Art Resistance Although AI art scores some points for sharing underlying similarities with its human counterpart, it is still far from gaining favor among a growing jury of artists, art critics, and enthusiasts. Frankly, this fact comes as no surprise, as the technology continues to weave a sticky web of hard-to-navigate issues. Let’s look into some of them. The Authorship Conundrum The Authorship Conundrum The issue of authorship in AI art is complex and challenging, as there is no standing protocol for attributing creative ownership to a particular artist or individual. Furthermore, as some AI art programs learn and evolve — further pushing the limits of machine sophistication, deciding the extent of human involvement becomes even more challenging. What this means for all is a brightly burning flame of unending ethical and legal debates. In the interim and maybe long run, a potential solution is to consider viewing AI art for what it really is — a collaboration between humans and machines. The individuals who create the original algorithm or provide prompts receive the creative credits, while the AI programs go into the tool/collaborator category in the creative process. Sadly, suggestions are not enough. We are yet to reach any consensus on acceptable ways of defining ownership in the context of machine-generated art. Hence, the resistance remains. Intellectual Theft Intellectual Theft Yet another complex and multifaceted challenge powering the resistance to AI art. As AI programs can generate artwork strikingly similar to existing works of art, concerns about plagiarism have become increasingly pertinent. plagiarism The issue of mimicry is particularly pressing, as AI-generated pieces that closely resemble the works of known artists could be construed as intellectual theft if they are marketed and sold without appropriate attribution or permission. Also, the fact that AI art programs run on data sets that include existing copyrighted artworks raises further concerns. Without ethical guidelines and best practices within the field, including standards of transparency and consent, attribution, and protocols for utilizing existing and copyrighted material, AI art stands a long way from winning over the resistance. Good Old Evolution Good Old Evolution As creatures keen on survival, humans are biologically hardwired to resist change — the unfamiliar. On the one hand, this is understandable. On the other, it is maybe responsible for our resistance to the new AI art technology. Think of it; until recently, the art domain remained exclusive to humans. So much so that when thinking of art, the mind immediately conjures images of trained hands and gifted minds as they transform time, patience, mistakes, and emotions into masterpieces. Pure sorcery! We even go as far as spending years and thousands of dollars in tuition fees to learn the inner workings of this sorcery. What would it mean to make room for a system that has reduced it into machines and algorithms? What happens to the art degrees and years of devotion? How about the artist’s feelings and career security? Without “good” answers, questions like these aren’t just questions. They feel like threats. They feel like threats. Whether or not AI art programs pose any threats seems to be secondary to the feeling of being exposed to such threats. And since evolution has designed our survival instincts to kick in the wake of threats, making efforts to understand AI art and identify its true risk status might be vital in easing the resistance. Parting Thoughts Parting Thoughts The point of this post isn’t to argue that AI art programs are just as creative as humans. That would be a futile mission for many good reasons. One of which is the fact that humans are capable of being dynamic. We are also capable of experiencing powerful emotions that influence our creativity. The same isn’t true for even the most advanced AI models. However, considering the shared similarities with the human creative process, there is a case for rethinking our understanding of creativity and AI-generated art. If art serves to create beauty, evoke emotions or drive narratives, should the matter of how it comes to exist take from or add to its creative essence? Should it matter that it is human-made or ai-generated? Should we focus on what makes AI art imperfect to justify the resistance? Or do we make necessary improvements and explore a new path to limitless creative possibilities? What say ye, brethren? Let me know in the comments. Also published here. here.
