Last semester, for a different final project, I used Discord and embarked on a journey into the unknown, starting my experiences with AI image generation with Midjourney for my DTC 475 class. It was both mesmerizing and unsettling—a cool yet creepy experience that felt like learning an entirely new language. Learning the basics of this powerful tool came quickly, but achieving even a pinch of fluency required a sometimes frustrating cycle of trial and error. It was in this landscape of digital alchemy that I asked and contemplated my creative question: Are AI image-generation tools like Dall-E and Midjourney truly all-powerful and a threat to the livelihoods of trained artists?
This essay and accompanying project come from a deeply personal place, as both a digital and traditional artist navigating the relatively new world of AI art. With a plan to generate my own images, I aim to explore the nuanced differences between AI art generation tools like Dall-E and Midjourney. More than just a technical comparison, the purpose of this process is to understand how the quality of prompts affects the final image, revealing the necessary skills required to properly utilize certain AI tools.
At the heart of my journey is a drive to explore the ethics of AI in art—a topic that resonates with me because of how much I value humanity and its contributions to society. This technology not only reshapes the landscape in which we create but also holds a mirror to the values, fears, and aspirations of those who use it. As someone involved with the artistic community, I wish to uncover how these digital tools affect us—not just as creators but as individuals navigating a world where art and artificial intelligence come together.
Artificial intelligence (AI) is making it easier for people who aren't traditionally trained in art to create artistic works. This technology provides access to those who might lack resources like art supplies or formal education. AI's role in the arts began with tools like Photoshop, which introduced some AI features for photo editing. Over time, AI has evolved into specialized programs such as MidJourney and Dall-E for image generation, ElevenLabs for voice synthesis, Suno AI for music, and Runway for video production.
Public opinions on AI vary. Some people are indifferent, possibly because they aren't aware of it or it doesn't affect them directly. Others see AI as a powerful tool that can open up art to a wider audience and have positive impacts across many areas. However, some view AI as a threat, worrying that it could take over jobs from skilled professionals or even pose a risk to humanity itself. This spectrum of opinions is the basis for the debate itself and leaves individuals with the feeling that they must “pick” a side to stand on, even though AI is ever-changing and evolving minute by minute.
Originally, I was skeptical about AI, especially in the arts. As an artist myself with many talented and hard-working artistic peers, I worried that it could undermine what makes art special: the human experience and emotions behind it. But my views are starting to change. While I still recognize the potential dangers of AI, I'm more open to considering its benefits and challenges. This project will explore these aspects, examining AI's role in art and whether it threatens or enhances artistic expression and the livelihoods of classically or traditionally trained artists.
Ethical considerations in AI-generated art are becoming more and more important as the technology becomes more intertwined with the creative industry. One of the primary concerns is that of originality. Critics of AI art argue that while traditional and digital artworks by humans might also be derivative, they still offer a personal touch, innovating upon existing ideas. In contrast, AI systems, trained on large quantities of human-created artworks, essentially recycle these inputs to produce variations on existing themes. This brings up concerns about whether AI can truly create "new" art or simply remix old pieces without original thought or emotional engagement.
Another major ethical issue is copyright. AI models use artworks created by individuals who have dedicated significant time and effort, often years or lifetimes, to mastering their crafts. When these works are used without permission to train AI, it leads to questions about the legality and morality of repurposing these artworks without consent or compensation. This situation complicates the traditional understanding of intellectual property—can AI-generated art be subject to copyright infringement? If so, who is responsible? Is it the developers who created the AI, the users who input prompts, or the AI itself?
Addressing these ethical challenges requires a nuanced approach, balancing the innovation AI brings against the rights and recognitions owed to human artists. The goal should be to find a middle ground that allows creators to explore AI technologies while ensuring that artists keep control over their intellectual property. This balance is not easily achieved, but will be worth the effort. It involves creating policies that do not suppress the creative possibilities of AI or ignore the rights of artists.
Implementing such policies comes with its own set of challenges, including enforcement and ensuring that regulations do not unjustly get in the way of technological progress. The complexity of these issues explains why a solution has not yet been found. However, the ongoing debate itself is a step towards understanding and potentially resolving these conflicts. As we continue to explore the capabilities and impacts of AI in art, it is highly important to keep these ethical considerations as a major focus to develop a framework that respects both human creativity and technological advancement. In this attempt to find a solution, it is unlikely that a perfect one will be found to satisfy all parties. A compromise that respects the contributions and rights of artists while being open to new technologies is necessary and possible. I believe that with time and continued conversation, a reasonable and fair resolution can be reached that keeps in mind the interests of all parties.
Navigating AI art generation tools like Chat-GPT, Dall-E, Midjourney, Stable Diffusion, Suno, and Runway ML has been a complex journey for me. These tools, while innovative, require a blend of technical know-how and artistic vision to produce quality outcomes. My experience has primarily been with Dall-E 3 within the Chat-GPT system, where I've seen firsthand how the quality of input significantly influences the output or result. This principle holds across various AI platforms, reinforcing the adage, "you get what you pay for."
Learning to create effective prompts is crucial to mastering AI art generation. A badly made prompt, such as "give me an image of a black and white dog," is often too vague, leading the AI to make assumptions based on previous inputs from the user or popular trends, which rarely matches a specific artistic vision. A well-crafted prompt includes detailed descriptions not just of the subject, but also of the composition, color palette, and even the mood of the image. For example, a more effective prompt might specify the dog's breed, characteristics, and setting, like "An image of a black Border Collie with white paws and tail tip, sitting on a tan pillow in a mid-century modern living room during early morning, with a natural color palette and a 5:9 aspect ratio." Being specific in this way and using the proper language guides the AI, resulting in artwork that much closer resembles the intended vision.
The learning curve for these tools can be steep. Generating art with AI is not as simple as giving basic commands; it requires understanding the unique 'language' each program uses. Midjourney and Dall-E 3, for example, each have their own nuances in how they interpret prompts, creating a period of learning and adaptation to effectively use each tool. Additionally, mimicking specific artistic styles using AI requires more than just casual use; it demands an in-depth understanding of both the technology and artistic composition.
My experience in this class has significantly re-shaped my views on AI in the arts. Initially, I found the reliance on technical knowledge over traditional artistic skill highly concerning. However, as I've grown to learn more about these tools, I've come to appreciate the blend of art and technology they represent. Effective use of AI art tools is not just about manipulating software but about translating one's artistic vision into a language that the machine understands. This does not replace traditional artistry but rather introduces a new medium that requires a different type of skill and creativity.
Reflecting on AI's changing role in the art world has shifted my perspective significantly, especially because of my hands-on engagement with the tools that I once viewed as potential threats to my career as a traditionally trained artist. This course has not only enhanced my understanding of AI art generation tools but also deepened my insight into my own artistic journey and future aspirations within this new technological landscape.
Initially, I was apprehensive about AI art generators, fearing they might undermine the skills and talents that artists spend years mastering. However, my direct experiences with these tools have revealed that they are not as straightforward to use as one might assume. They require a unique set of skills, blending technical knowledge with a creative vision that cannot be easily replicated by those unfamiliar with the medium. This realization has somewhat alleviated my fear that AI could render traditional artistic skills obsolete (at least, for now).
In addressing whether AI art generators are a threat to artists like myself, my stance has evolved. These tools do not simply complete tasks or create art autonomously; instead, they require directive input and a creative partnership with a human to produce meaningful results. This highlights a different but equally valuable skill set that artists must develop to coexist with and harness these technologies effectively, something that I see myself diving further into in the future.
Despite recognizing and understanding the utility of AI in art, I am still concerned about the ethical aspects, specifically regarding copyright and the originality of AI-generated works. These issues are complex, with no clear solutions on the horizon. As AI technology continues to advance—evidenced by the recent emergence of sophisticated video generation tools—it is crucial to stay informed and engaged with the ongoing conversations around these developments.
Looking ahead, I believe it is essential for artists and the broader community to keep up to date with technological advancements and the ethical debates that come along with them. Listening to the concerns of those who feel marginalized or exploited by AI is essential for fostering a balanced and informed approach to integrating these tools into our artistic world and society. AI is unlikely to disappear; it will continue to evolve, mirroring trends seen in computer technology and the internet over the past decades.
My hope is that through continued dialogue, experimentation, and advocacy, we can develop ethical frameworks that address the concerns of all stakeholders involved. This includes ensuring that AI complements and enhances human creativity rather than replaces it, essentially enriching the art world while respecting the rights and contributions of traditional artists. By preparing to adapt and engage with AI, I aim to not only enhance my own capabilities as an artist but also contribute positively to shaping how these technologies are integrated into the creative professions.