Creativity in
the age of AI

In 2018, Christie’s became the first auction house to sell an artwork created by Artificial Intelligence (Portrait of Edmon Belamy sold for $432,500). Four years later, the boom of text-to-image AI generators like Dalle, Midjourney, Stable Diffusion, OpenAI’s DALL·E 2 and Google’s Imagen and Parti, which, like LaMDA and GPT-3, also learn to perform transformations from vast data sets, continue to raise complex questions about art, design, and the creative process. Designer Stefan Sagmeister described DALL-E as “absolutely fascinating and will touch likely everything we do. As these programs improve, enormous questions about authorship will arise. I am kind of speechless.” Can we move beyond this speechlessness? As improvements in AI image generation move beyond the uncanny valley and become indistinguishable from human creations, what does this mean for creative practice, and how should artists, designers, and thinkers respond?

In the summer of 2022, many philosophers, engineers, and cultural critics felt compelled to comment on the cultural and existential dimensions of creative AI. Things came to a head when an engineer at Google claimed that the LaMDA chatbot, a deep-learning language

processing AI, was in fact sentient. Similar to OpenAI’s better-known GPT-3, these models are more than capable of carrying on conversations with humans, in many ways far exceeding the canonical requirements of the “Turing Test”. Unsurprisingly, this engineer’s claims occasioned an outpouring of thinking and writing about the place of artificial intelligence in the tapestry of human life. But in all this fervor over the linguistic capacities of AI, and its challenge to traditional conceptions of the human, we think something important has been overlooked, namely, the incursion of new technologies into the similarly distinctively human endeavors of creating and appreciating art.

These developments raise questions that are at once pragmatic, economic, artistic, and existential. Is AI a surprising tool for artists and designers to incorporate into their workflow, or a doomsday replacement? What is an artist when not only production but ideation can be outsourced? At the moment, these text-to-image generators still link composition and production to the human voice, human art direction, but for how long? Can AI replace visual creativity—the “gift” of art?¹