In all the arts there is a physical component which can no longer be considered or treated as it used to be, which cannot remain unaffected by our modern knowledge and power
Walter Benjamin, 1935
In a few recent blog posts (What do you wish existed, Artificial Vanity for the masses), I described a series of technological advancements unlocking on-demand Artificial Intelligence image generation. As I continue to play with this technology, one question remains potent:
“Is this real Art? Is this real fantasy?”
Bohemian Rhapsody (and some creativity)
There are two good times to quote Freddie Mercury: Now and Forever. This instance is no different, but the paraphrase (changing the dichotomy of real and fantasy to separate questions) has a deeper meaning, as these are separate questions:
- Is AI art “Art”?
- Is this even a creative endeavor?
I understand how illustrators and artists can feel uneasy about this topic: These models ( the biggest ones being Stable Diffusion, Midjourney, and DALL-E ) were trained on the publicly accessible works of others. The computer seems to be producing some random blend of all available data guided by the text prompt without any attribution whatsoever.
One interesting angle to note is that with technological advancement, Humanity seems a little less remarkable: Copernicus dared to insinuate that the Earth is not the center of the Universe, and Darwin suggested that we are, in fact, part of the animal kingdom. AI “Art” is chipping away at our confidence in Art and Creativity being uniquely human.
Despite my techno-optimism, I recognize this is a very complex topic. In my research, I found a 1935 essay by Walter Benjamin titled:
Work of Art in the age of mechanical reproduction
In the essay, Walter Benjamin focuses on photography and film as new art forms. He argues that they are uniquely different since they only exist to be consumed by the public. No “art object” is attached to them, contrary to preexisting art like paintings, sculptures, etc.
Works of art are received and valued on different planes. Two polar types stand out:
1 – with one, the accent is on the cult value
2 – with the other, on the exhibition value of the work.
Artistic production begins with ceremonial objects destined to serve in a cult. One may assume that what mattered was their existence, not their being on view
He argues that after removing the “ritualistic magic” trapped in the art object, only its exhibition can be of value. That exhibition is subjective, and art’s reception can only be connected to the current political discourse:
photographic negative, for example, one can make any number of prints; to ask for the ‘authentic’ print makes no sense. But the instant the criterion of authenticity ceases to be applicable to artistic production, the total function of art is reversed. Instead of being based on ritual, it begins to be based on another practice—politics
I am going somewhere with this, I promise.
AI and Film
While reading this 1935 essay, it occurred to me that the questions regarding the “authenticity” of AI-generated Art are an extension of controversies surrounding photography and film:
But the difficulties which photography caused traditional aesthetics were mere child’s play as compared to those raised by the film.
One point of view Benjamin explores is that there is “substance” to film. The entire artwork is in operating the point of reference and frame used during the filming:
The camera that presents the performance of the film actor to the public need not respect the performance as an integral whole. Guided by the cameraman, the camera continually changes its position with respect to the performance. The sequence of positional views which the editor composes from the material supplied him constitutes the completed film. (…) There is no such place for the movie scene that is being shot. Its illusionary nature is that of the second degree, the result of cutting.
This analogy carries very well to AI-generated images: The artistic process in AI-Generated Art consists primarily of:
- Picking and changing the reference frame – that is, crafting the prompt
- Selecting, discarding, and curating the output
Producing AI-generated images is like owning a camera and a magic portal to the entirety of recorded existence.
Is the photographer less of an artist because he can only arrange preexisting elements?
Democratizing publishing
The mission of our company is to “Democratize Publishing”. I am happy to read that this was already accomplished in 1935:
For centuries a small number of writers were confronted by many thousands of readers. This changed toward the end of the last century. With the increasing extension of the press, which kept placing new political, religious, scientific, professional, and local organs before the readers, an increasing number of readers became writers—at first, occasional ones. It began with the daily press opening to its readers space for “letters to the editor.” And today there is hardly a gainfully employed European who could not, in principle, find an opportunity to publish somewhere or other comments on his work, grievances, documentary reports, or that sort of thing. Thus, the distinction between author and public is about to lose its basic character.
Coauthor – AI Writing Assistant and Artist
Last week, I (re)wrote a WordPress plugin to put the power of AI inside your WordPress editor. The plugin introduces two new blocks:
- Coauthor Paragraph will take your content above the block, automatically generate GPT-3 completion to the text, and insert it into the blog post.
- Coauthor Image will generate 4 images according to your prompt with DALL-E 2, upload the one you select and replace itself with an Image block.
You can find the Coauthor plugin in the WordPress plugin repository.