Thought this week​

The media’s hype surrounding AI may be slowing, but as AI increasingly influences creative processes, industries such as Fashion must grapple with critical questions around the evolving collaboration between humans and AI. The nexus between technology and creativity presents opportunities for innovation and artistic expression, while also raising important discussions around the legal and ethical implications of AI's growing role in new concepts, products and art.

AI's capacity to instantaneously create art based on an artist's style has raised novel intellectual property challenges in the past year and while AI is a formidable tool, attributing credit and ensuring authenticity must be addressed alongside preserving human creative value. Alarmingly, RetailWire recently reported OpenAI and Anthropic have been disregarding web scraping rules for bots. Artists whose work is extensively indexed or falls within a prevalent genre find their styles captured by generative AI models, something we’ve heavily seen with, for example, Wes Anderson’s signature style used by many in AI generated content.

Artists and designers alike must respond to the capture of the style they developed over a lifetime of work. A philosophical divide is emerging among creators with some prioritizing protecting their intellectual property, while others, like Grimes, embrace AI as a collaborative partner. By sharing earnings from AI-synthesized songs using her voice, Grimes exemplifies the potential for AI to augment production without replacing human creativity. Despite ongoing debates within creative communities, both perspectives demonstrate a willingness to adapt and harness AI's potential.

The possibility of generative AI systems inventing products also raises questions. The case of Dabus, an AI system granted a patent in South Africa, highlights need for legal frameworks to accommodate such advancements. As AI increasingly contributes to or even fully generates inventions, artists and creative workers worry about their skills and job prospects with a recent IMF report claims that 60% of workers will be affected by AI-driven job automation. Recent labour rights disputes, such as the 2023 Writers' Guild strike in Hollywood underscore these concerns.

Implementing regulations will be necessary to establish clear guidelines that balance the interests of inventors, creatives, industries, and society as a whole. Silicon Valley's "build first, ask permission later" approach is diminishing as regulatory oversight increases due to mounting concerns with recent efforts such a US presidential order on AI and the EU's AI Act.

However, many artistic professions exist in a regulatory grey area, posing challenges in extending such protections to creative workers. Ensuring that regulatory efforts effectively address the unique nature of artistic labour is crucial to safeguarding livelihoods. But as Herndon Dryhurst Studio points out “All media is now training data.”

New strategies are emerging to address the challenges of this "data as shadow labour" paradigm. Researchers from Google, DeepMind, ETH Zurich, Princeton, and UC Berkeley have shown that AI models like Stable Diffusion can memorize and reproduce images, even those with trademarked logos. To address this, digital watermarking techniques are being developed, such as Google DeepMind's SynthID, which embeds invisible marks in images for AI-generated content identification. Similarly, Tracklib, co-founded by Pär Almqvist, simplifies music sampling while ensuring copyright ownership transparency.

For individuals engaged in creative fields, AI serves as an invaluable tool to bolster innovation and bridge the gap between creative vision and technical execution. While AI cannot replace the human touch, it can augment and inspire new insights when used consciously. Recent research indicates that AI excels in creativity, with a University of Montana study, ChatGPT outperformed nearly all college students on the Torrance Tests of Creative Thinking, which measure human creativity skills. These findings suggest that AI can be an exceptionally creative collaborative partner. Notable examples include Tyler Hobbs, a visual artist whose artwork focuses on computational aesthetics “how they are shaped by the biases of modern computer hardware and software, and how they relate to and interact with the natural world around us.”

The question remains: what makes a great work of art? Acknowledging an artist's context, story, and interpretation gives art its meaning. Our ability to make meaning through unique personal experiences is difficult to replicate artificially. As Keith Y. Patarroyo points out in SUM Journal Issue 18 “Can we quantify the unquantifiable, compute the incomputable, or capture the infinite with the finite? This seems to be the problem one faces when trying to develop a theory of creation. The creativity of the human mind or the creativity of nature seems to be far beyond any type of understanding, moreover many elements seem to hide even more when one tries to explore them. In a way, it is just like when you are not looking for love, but it suddenly appears in front of you, elements of nature’s creativity seem to marvellously manifest when you are not looking for them.”

Adam Moss explores this intangible magic in his book The Work of Art, attempting to break down “the work—the tortuous paths and artistic decisions—that led to great art”. Embracing the messiness of creation allows us to reach unforeseen possibilities. We can do this with a variety of techniques including AI. But what seems clear and impossible to replicate, dictate or automate is human’s propensity for exploring serendipity, editing and iterating which is the essence of all great creative endeavours.

By Geraldine Wharry