Data Scraping Before AI: The Story of Modern Art
Cubists pioneered one of the most radical avant-garde movements in Europe during the early 20th century. However, these artists “scraped” information or visual elements from several sources to establish what we know today as Cubism. Artists like Picasso collected dozens of African masks and other figures which greatly influenced the artwork (see above).
Countless original masks and sculptures were widely smuggled and sold among Western societies… This rapid proliferation allowed European artists to come in contact with African art and its unprecedented aesthetics. — Carolina Sanmiguel
To add, it wasn’t just African masks/art that influenced Cubists. Several other art styles and artists’ works were studied and appropriated. I recently incorporated this information is a keynote presentation for art educators. Cubists took elements of Egyptian art, Impressionist art, collage and assemblage, Constructivism and Neo-Plasticism. Constructivist art is characterized by its industrial materials, utilitarian purpose, and rejection of decorative styles. Neo-Plasticism uses primary colors, geometric shapes, and straight lines to create a sense of visual harmony.
I show this slide (above) to compare Cubism with data scraping that focuses on extracting specific information from websites. Data scraping can involve collecting data from various sources, including databases, web documents, tagged images, and text files. It is an umbrella term under which both web and screen scraping fall. The practice of extracting or appropriating elements from publicly-accessible works did not begin with the invention of generative AI. What Cubists and other modern artists did was and still is an accepted practice, as a form of content appropriation; it is still acceptable to extract from and use existing creative works.
In the 1980s, Black artists’ creative music sampling was condemned as theft. 40 years later, AI’s steals all human creativity at scale, and they call it a genius breakthrough. — Ron Eglash
Content appropriation (ex. data scraping) became a kind of zeitgeist during the late 1900s/early 2000s and scholars like Lawrence Lessig, a proponent of remix culture presented this practice as a desirable cultural practice distinct from piracy or ‘stealing’. Lessig articulated remix culture as intrinsic to technology and the Internet, as an “amalgam of practice, creativity, “read/write” culture, and a hybrid economy.” Sampling in music production, which grew in popularity in the 1990s, is an example of reusing and remixing to produce new works. Some trace sampling to rock groups like Led Zeppelin who “interpolated” portions of songs by earlier musicians such as Memphis Minnie for “When the Levee Breaks.”
It should be noted that it wasn’t until the end (or well after the end) of Memphis Minnie’s life that she was given credit for her contribution to Led Zeppelin’s genre shifting version of her song. Her family is currently suing record companies and some artists for royalties and for using her music without permission. Just as Cubists did not provide attribution or compensation to African artisans and craftspeople, there is a long history of appropriation and re-appropriation in modern art. This includes digital art that is subject to replication and appropriation in ways that traditional art forms are usually not. Digital art can be easily reproduced and shared, making it susceptible to plagiarism and copyright infringement.
What is often ignored in discussions about digital art is the artist’s hand in the process. The same can be said for AI art.
All things considered, I find it interesting where and when artists and critics draw lines between what they deem is appropriate as far as content appropriation (ex. data scraping). Dadaism used appropriation as a way to challenge traditional art and society. Dada artists borrowed, altered, and copied existing objects, images, and ideas to create new works of art. The Dadaists’ “primitivist” performances often appropriated non-Western cultural artifacts without understanding or acknowledging their original values and purposes. Also, back in 1913, Dadaists described what would become the ‘latent space’ of generative AI, as a theory about the future of music production (i.e., noise aesthetics, diffusion). The noise aesthetic can be found in 1990s hip-hop when producers sampled snippets of sounds to create some called a “wall of noise”:
In an interview Hank Shocklee, one of the main producers of Public Enemy’s “Rebel Without a Pause,” likened this process to collage:
I had a ridiculous record collection. And I wanted to prove that it was the records that inspired me. Because … I understand scales and musical arrangements and that stuff but I didn’t have — I was not a player. I’m not going to pick up a bass or a guitar or keys and I’m going to, you know, put some virtuoso stuff down. That’s not going to happen. But what I do have is a turntable and records. And so I just want to create this collage, almost like a Romare Bearden kind of a painting. — Microphone Check
The person who put me on to the Dadaists was Vernon Reid, the British-American musician best known as the founder of the rock band Living Colour. Vernon agreed with me about the history of content appropriation now referred to as data or web scraping in the art world today. A line can definitely be drawn from early 20th century Futurist/Dadaist manifestos to Cubism, collage/assemblage, rock & roll, sampling and now AI art.
The variety of noises is infinite. If today, when we have perhaps a thousand different machines, we can distinguish a thousand different noises, tomorrow, as new machines multiply, we will be able to distinguish ten, twenty, or thirty thousand different noises, not merely in a simply imitative way, but to combine them according to our imagination. — Luigi Russolo, 1913
So here we are at another turning point in modern art history that is arguably the most impactful to practicing or professional artists than ever before. There are multiple quotes about history repeating itself and one of the most quoted ones is:
Those who cannot remember the past are condemned to repeat it. — George Santayana, The Life of Reason, 1905
We can keep spinning our wheels at every turn or attempt to throw the babies out with the bath water, so to speak, or we can acknowledge the important role that content appropriation (and re-appropriation) has in modern creative practices, as a way to better understand what is happening in the art world today. My point is that appropriation, data or web scraping (in GenAI) is a complex issue that touches on many things including sociopolitical issues, ethics, algorithmic bias and economics.
My question to ponder is: Is content appropriation simply an effect or result of late-stage capitalism?