brettworks

AI’s incursion into art-making raises the question, What defines the artist’s skill set? Artists are artisans and art is artisanal. Building on this, consider six qualities of art’s artisanal-ness and how AI fails to achieve them.

Art-making is by hand. Playing music, writing, painting, dancing, cooking—art contains and expresses traces of the body that made it. But there’s no hand behind AI, just code modeled on human-crafted output. No hand, no body, no human.

Art-making is small batch. Artists spend a lifetime producing, on average, relatively few novels, songs, or paintings. AI is large batch. Because there are no slow humans involved, AI can generate vast output in seconds.

Art-making is high quality and tasteful. The quality of light, the grain of sound, the structure of intricate plots—artists spend time getting the details right. They perform, think through, edit, refine, and finesse their works until they resound with excellence, sing with spirit. AI art, is, generally speaking, consistently mid to low quality. AI art lacks taste because its technologies are without the faculty of taste. AI art can’t see its own ugliness. This is why so much AI content is referred to as slop.

Art-making is dynamic. Artists are aesthetic specialists sensitive to life’s inputs. Their art-making is variable and whimsical, changing with the times, changing with one’s mood and tastes, adapting to one’s materials. AI art impersonates an artist’s dynamism, relying on ever-updated prompts to feign aliveness.

Art-making is process-oriented. Artists direct their energies to understanding their art’s expressive language, tools, techniques, and community of fellow practitioners. Thus, artistic process takes years to coalesce and grow ideas. AI replaces process with number crunching models that produce facsimile artistic results, bypassing the very embodied processes that are art’s generators.

Art-making is improvisatory. Artists build upon moment by moment trial and error tinking as the engine of invention. Improvising is conjecture, a dynamic What if? way of learning developed over a lifetime pursuit of know-how in the real (not virtual) world. An artist’s life is her only algorithm, while AI is a data center dummy waiting to be told what to do next, with the prospect of even a Turing-worthy imagination always out of reach.