AI-Generated: Humanity Is A Burden
I don’t think AI is bad, nor do I believe that it is good. It exists and needs inclusive design and to prioritize human wellbeing. Algorithms are capable of democratizing access to information and knowledge, but whose knowledge is promulgated? Algorithms define prison sentences, credit scores, mortgage lending decisions, and what you see from your friends and family. What could go wrong there? They remove human agency for convenience, supposed objectivity, or some notion of community “engagement.” Art is the latest career path for AI, and I wonder what art will become and whose art will be remembered?
If | Then
Algorithms are a framework for making decisions. You design algorithms anytime you think “if ‘X’ person shows up, then I’m leaving.” Machine learning (ML) algorithms take a set of instructions and process different outcomes at a scale no human or warehouse of humans could accomplish in a matter of seconds. Our slow brains and resource intensive bodies make for sub-optimal capital accumulation.
VALL-E, DALL-E, GPT
The arts were a place where humans endured, or perhaps art endured humans. Advancements in ML algorithms and increased accessibility of processing power plummeted the cost of artistic automation. With VALL-E, DALL-E, and ChatGPT art is cheaper, faster, and losing the need for inconvenient human labor. Aesthetic values are enshrined in complex computer code used to generate art; it removes the need for human struggle. The struggle to discover something so perfect it forces people to feel in broad daylight.
Without struggle, learning becomes optional. While I’m not about to gatekeep who has the right to be an artist, there’s merit in learning by doing. Learning a language versus Google translate has a very different outcome and is used for a different purpose. Becoming fluent allows for a level of freedom and agency not provided by the short-term utility of auto-translation, and not all scenarios require fluency. It’s a question of what AI-generated content is for. Currently there is no purpose. It’s innovation for the sake of innovation and people reflect their own values and intentions on to the machine.
Renoir Deconstructed
For years I observed artists create and critique their work to define something unique. Once the aesthetic had reached the height of innovation, the studio they worked for sought mass production. Artists were shipped out to third party studios to train armies of cheaper artists how to replicate their work. The result was never exact, but it was close enough for the original artists to polish and approve. Their labor used to create the art to train the machine to make the art. The artist was deconstructed and diluted—data.
An algorithm replicates what it is given, or more accurately, what developers steal from work posted online and input into a dataset. While MidJourney gives the illusion of creating original pieces of art, it does so by copying and pasting work from artists around the world—admittedly, this is a gross oversimplification. It replicates the lines, textures, and colors that collectively define an object or concept. The innovation is in the connection of these mechanical traits to specific styles or individual artists. Not only can you have an illustration of a sock, but you can have that sock in the style of a 19th century Renoir. A new sensation devoid of intention, narrative, and emotion producing mechanical skill masquerading as imagination.
Copying, referencing, and tracing are tried and true tactics for developing as an artist. We copy to learn. Creativity is essentially the synthesis of multiple references interpreted through our own experience. At our best these tactics are in pursuit of something unique, however that is not always the case. Artists have seen their work plagiarized or their style reproduced by others for years. However, AI appropriates the work of artists in a matter of seconds through an interface open to everyone. The abuse of existing harms becomes more widespread. Plagiarism is laundered as AI-generated content with no one person accountable and no one artist violated.
Automation Is Accessibility
Any good designer or engineer will say they are always trying to code themselves out of a job. With enough style guides and templates, I could minimize the need for a designer to be present and do more with less time. These automated processes I created improved brand consistency. In the world of design, consistency is a kind of quality with the many rogue actors who have a Photoshop license and just enough knowledge to be dangerous. However, speed and accessibility create the parameters for a ubiquitous aesthetic.
In the name of accessibility, technology assimilates our environment. Tools like Squarespace unify the web through templates; co-opting quality standards created through the labor of individual designers and regurgitated as a user friendly tool. This allows everyone to communicate with the world, share their thoughts, promote their work, and engage with customers—democratizes the web! Web design is a process of communication that is only made better through open and equitable access. Should art be an open and equitable process?
Our Human Affliction
Our unending cycle of innovation and efficiency de-personalizes people into data and humanizes technology. The future is not devoid of technology, at least you’ll be prying my phone out of the iron grip of my ancient bones. But our constant desire for more is driving society toward viewing humanity as an affliction. A world is manifesting where humans are replaced by machines, art is replaced by code, and life is a mere replica of those who lived before. There are no answers here, just a desire to contribute humanity to the world of machinery.
