AI & Artistic Identity Research Presentation

On April 28th, 2023, I had the privilege of presenting my research on AI, artistic identity, and copyright law at Arizona State University's Lincoln Center of Applied Ethics Undergraduate Research Symposium. The presentation, titled "A.I. and Artistic Identity," delved into the complex legal and ethical implications of AI-generated artwork and the need for adapting copyright laws to emerging technologies.

My research journey led me to explore theories of consciousness, creation, freedom, expression, aesthetics, ethics, and law. The goal was to challenge existing mindsets and shift perspectives on how we view AI, art, and copyright in our rapidly evolving digital landscape.

Key points from my presentation included:

1. The similarities between human and machine learning processes

2. The paradoxical nature of creativity and originality

3. Challenges in defining and testing AI sentience

4. Controversies surrounding AI art generators and copyright infringement

5. Rethinking the concept of authorship in the age of AI

   
   

I proposed a shift in how we approach copyright, suggesting that artists' creative processes, rather than the art itself, should be the focus of protection. This perspective allows for a more open and collaborative approach to art creation and dissemination, benefiting both human artists and AI advancement.

The presentation concluded with the idea that in the debate between art and artist, art itself is winning. While there's immense value in human creativity, the products of that creativity should be free to inspire and influence the development of aesthetics, culture, and technology.

I'm grateful for the opportunity to share my research and contribute to this important dialogue on the future of art and technology. As we continue to navigate this new frontier, it's crucial that we remain open to evolving our understanding of creativity, ownership, and the role of AI even beyond just artistic expression, accounting for our expanding connectivity as humans and respect for others on this planet.

Transcript

Good afternoon, everyone. I'm Samantha Robyn, and today we're exploring the fascinating intersection of A.I., artistic identity, and copyright law in our increasingly interconnected world.

I began with this question: "What are the legal and ethical implications of using A.I. to create artwork, and how should copyright laws adapt to emerging technologies?"

This inquiry led me down a rabbit hole, exploring consciousness, creation, freedom, expression, aesthetics, ethics, and law. What I discovered wasn't a destination, but a paradigm shift - a challenge to our existing mindsets about creativity, ownership, and the very nature of art itself.

The goal of my presentation is to guide you through this journey, revealing the complexities of this question and hopefully aligning you with a new way of looking at A.I., art, and copyright that promotes accessibility and innovation while still valuing human creativity.

—

Let's start with a thought experiment. Recall your first childhood drawing of a tree. Imagine that rudimentary sketch, perhaps resembling more of a shadow than a tree, caught between your perception and your ability to capture an accurate image.

Now, think about a fully formed tree. Envision the leaves, the way the bark climbs up the trunk and through the branches. If you were to draw that tree again today, how different would it look?

—

You'd likely create something far more accurate and true to life. The difference lies in your accumulated knowledge and experiences - the sensory data you've collected about trees over the years. This process of learning and creating reflects the interconnected nature of consciousness and art, as philosopher Alan Watts might say, "You are an aperture through which the universe is looking at and exploring itself."

—

Machines learn similarly. The more data they're provided, the more they learn, and the more accurate their output becomes. This parallel between human and machine learning raises profound questions about the nature of creativity and consciousness.

AI art and media generators compile vast amounts of data from pre-existing images, videos, text, and audio, composting it together to form something seemingly new and inventive.

—

They render unique content by compositing other content, essentially imitating, copying, and reusing it. This process mirrors how we form our understanding of the world from information gathered through our senses and experiences over time.

In this way, AI is remarkably similar to us. After all, the processes used to code AI are modeled on human brain processes. But if something acts like us, and even outperforms us, does it deserve to be treated as one of us? This question touches on the heart of existential philosophy, challenging our notions of human uniqueness and purpose.

This is one of the key questions I explored in my research. We've never before seen anything outside our human race create multimedia content so compelling, beautiful, and ingenious.

—

Traditionally, we assign copyrights to such creative outputs. These give artists and authors the ability to own, reproduce, and distribute their original work when fixed in a tangible medium of expression.

—

However, copyright laws only protect humans. This limitation raises profound questions about the nature of creativity and ownership in an age of artificial intelligence.

—

In my research, I explored post-humanistic views that suggest AI could deserve similar rights to "their" outputs. Posthumanism challenges the idea that humans alone hold special value, agency, and moral supremacy. It entertains the notion that AI could be just as sentient and creative, potentially granting AI the status of creator, artist, or author.

—

But what does it mean to have agency and be sentient? To perceive and feel, to act on one's own accord, to have consciousness? These questions have puzzled philosophers for centuries, and now they're at the forefront of AI ethics.

—

The Lovelace test, named after Ada Lovelace, the pioneer of modern algorithms, attempts to test AI sentience by challenging it to output something indistinguishable from what its designer programmed. The idea is that if the output reveals something new and surprising, the AI must be conscious.

—

However, this model has limitations. It's challenging to create something entirely unique within the confines of one's knowledge. Designers might always trace how their AI arrived at its outputs.

—

For this reason, Mark Riedl, a professor at the Georgia Institute of Technology, believes it's impossible for AI to pass the Lovelace test. He proposed the Lovelace 2.0 test, which places constraints on the AI through set expectations. If the AI surprises the judging agent, it's a measure of its intelligence, allowing us to quantify and compare the creativity of different artificial agents.

—

This leads us to another fundamental challenge: understanding creativity and originality. There's no universally accepted definition for creativity. The processes involved in human creativity closely resemble how AI produces creative outputs. Both require imitation, synthesis, and reinterpretation of existing ideas.

As the artist Pablo Picasso famously stated, "Good artists copy, great artists steal." This paradox is at the heart of the AI and copyright debate. It highlights the tension between individual expression and collective inspiration, a concept central to many philosophical discussions on art and creativity.

—

Since AI has become mainstream, controversy has surrounded AI art and media generators, particularly regarding their use of human works to create their own outputs. Some artists and authors feel threatened, as if their work has been stolen and used without consent or credit. Several active cases involve copyright holders suing AI companies on this basis.

—

However, AI art and media generators don't necessarily steal. The references they're fed are used for learning, not replication. Any appearance of copyright infringement comes down to adjusting an AI's algorithm, as all AI outputs require inputs. Questions of copyright infringement in AI processes result from human intervention, specifically in what information the AI is fed to learn from and produce with.

This dilemma brings us back to questions of authorship and ownership. If someone's work is used to create something else, is part of that output theirs? Who owns what? The software company that developed the algorithm? The user feeding instructions to the AI? The artists whose work contributed to the output? The copyright owner? Or the AI itself?

—

At this point in my research, I began to question why authorship is so important to us. What makes authors and artists necessary for art to exist? Are we over-emphasizing and over-empathizing with the identity of the "artist"? What is art separate from these ideas, and could we allow it to exist in a free and uninhibited way?

These questions align with the philosophy of art as a "strange tool," as proposed by Alva Noë. He argues that art helps us understand our perceptual engagement with the world. In this light, AI art challenges us to expand this engagement even further, pushing the boundaries of what we consider art and creativity.

—

I surveyed 70 people both inside and outside my personal community to gauge public opinion on this matter. 80% of respondents identified as artists, with about half agreeing to using AI generators frequently. While nearly half saw AI as a tool, they also perceived it as a threat to artists' work and validity. I suspect this is due to a lack of recognition around the creative process and the nature of inspiration itself.

—

Current copyright laws in the US align closely with utilitarian ideology, prioritizing societal interests over individual ones. The welfare theory, stated in our Constitution in Article 1, Section 8, aims for works to "promote the Progress of Science and useful Arts," benefiting the greatest number of people. Copyrighting work is in our best interest because of the direct, positive impact observed when works are commercialized, strengthening our economy.

However, this utilitarian approach may need to evolve in the age of AI. As philosopher Peter Singer might ask: If AI can create art that moves and inspires us, does it deserve moral consideration? How do we balance the rights of human artists with the potential of AI to democratize art creation?

—

My survey revealed that when considering artists' work as physical property, as current copyright laws do, over 80% of respondents believed the works belong to artists. However, when considering art as both physical and metaphysical property, over 60% thought that art in its subjective and conceptual form should belong to either no one or everyone.

—

Our current copyright laws don't acknowledge art separate from its physical manifestations. They recognize canvas, paint, film, paper, and data, but fail to measure artistic value as a subjective conceptualization.

For instance, the Mona Lisa is a painting by Da Vinci, but who owns the concept of "a brown-haired woman sitting in front of a scenic view with a subtle smirk"? Is it still Da Vinci's?

—

What our copyright laws protect are authors and artists, upholding societal standards. But if we understand art in this abstraction, it can exist free and uninhibited from individual creators and law, potentially benefiting the advancement of art, technology, and much more.

—

This perspective aligns with the concept of "commons-based peer production" proposed by Yochai Benkler, where information and creative works are treated as a shared resource, fostering innovation and cultural growth.

Studying AI art and media generators against copyright concerns revealed that this debate need not center around authorship, agency, or creativity. Instead, it could unveil an often-overlooked ideology promoting a dualistic sense of art, demonstrating how we should address property in our current copyright laws.

Overall, survey respondents placed more importance on identifying artists by their physical and mental processes and valued art based on those processes. This suggests that art, standing alone, cannot express value without its creator. And perhaps this is the case.

—

Maybe artists' and authors' processes are their property, which should be protected by law, while the art and media itself exists separately, belonging to no one and everyone simultaneously. It could lay outside the bounds of creators and copyright, objectified in its subjectivity for the purposes of learning and evolving without permission or compensation for any originator.

—

The benefits of adopting this mindset are clear. As artists and authors already borrow and draw inspiration from one another's work, AI should be able to access and learn from as many data points as possible to further technological advancements. The objective of using someone else's work should be to learn, evolve, and progress, not just to profit.

This approach aligns with the philosophy of Hannah Arendt, who emphasized the importance of 'natality' - our capacity to begin anew and bring novel ideas into the world. By allowing AI art to be free and accessible, we create more opportunities for this 'natality' to occur.

—

What my research with AI has taught me is that in the argument of art versus artist, art is winning. Yet, there's a great deal of power artists and authors can maintain through this. No one else can replicate their efforts and creative processes - not even AI. There's immense beauty and value in our human experience and expression of creativity.

But the product of that creativity is its own entity and should be subject to both the public and AI algorithms for inspiration and influencing the development of aesthetics, culture, and technology. By embracing this perspective, we open ourselves to a world where creativity knows no bounds, where inspiration flows freely, and where the advancement of art and technology can flourish unimpeded.

Thank you for joining me on this journey of exploration and reimagination. As we move forward into this brave new world of AI and art, let's carry with us the understanding that creativity, in its purest form, belongs to us all.