Having just watched a few minutes of the incredibly underwhelming Star Wars, one letter at a time by Brian Kim Stefans, I was expecting very little out of Dakota. To my surprise, experiencing Young-Hae Chang’s Dakota for the first time put me in a state of awe and shock. The drum beat and large, fast paced font put me in a sort of trance. My initial thoughts were that it was some kind of weird stream of consciousness piece but as I read/watched on, I realized that it was telling a story. While unfortunately my very short attention span made me click off the story roughly four minutes in, I thoroughly enjoyed listening to the drums and watching the flashing text tell a story over that period of time.
My immediate thoughts after reading Dakota, was why have I not seen this before and why is there not more of this. I proceeded to watch Young-Hae Chang’s other work, Bust Down the Doors!. This story is filled with even more action as it is told in first person which makes it seem like you as the reader are experiencing the story. Experiencing both these works of digital literature filled me with excitement for what the future holds in regards to modes of fast-paced, “speed reading” stories. This form of literature may be a solution to the new wave of millennials and their tendency to lose interest quickly as a result of this new era of social media. Especially for people who have short attention spans like me, this fast paced way of reading literature holds my interest longer than a typical book or article.
When I first opened the link to Young-Hae Chang Heavy Industries’ Dakota, my immediate reaction was confusion. Was the page broken? Why was it blank? Then, the thrumming drum beat began in sync with the page color lightening gradually to white, and the first words of introductory text, a countdown from ten to one, flashed onto the screen, disappearing as quickly as they’d materialized. No less confused, I focused all of my attention onto the words appearing in front of me and braced myself for a reading experience different from anything I’d read before. And in this aspect, Dakota definitely did not disappoint.
As I got further along into the story, I grew increasingly frustrated. Some of the text flashed by in less than a second, much too quickly for me, or any reader, to be able to comprehend. What’s the point of having this part of the text if you can’t read it? I wondered. But, as I learned from Jessica Pressman’s “Speed Reading,” removing interactivity from the equation was exactly the authors’ intention.
As we mentioned in class this past week, one of the five elements of digital literature is interaction, or a lack of it. There are no buttons in Dakota to allow the reader to stop, pause, or slow down the rate at which the text appears (or even to pause the music), and this is precisely the point. As quoted in Jessica Pressman’s “Speed Reading,” Young-Hae Chang has said, “My Web art tries to express the essence of the Internet: information. Strip away the interactivity, the graphics, the design, the photos, the banners, the colors, the fonts and the rest, and what’s left? The text” (81-82). This is a very bare-bones approach to digital literature as we know it.
The YouTube video above, though one could argue is convenient, blatantly defeats the purpose of the work by reintroducing interactivity into the work. This is because, due to the way YouTube as a platform functions, the audience is able to pause, rewind, and fast-forward through the text at will. To someone who hadn’t gone through the original text on Young-Hae Chang Heavy Industries’ website and had simply watched the above video instead, the experience and its impact on the reader would be dramatically reduced. This only further serves to emphasize how radically different Dakota truly is from other forms of digital literature.
The modernist electronic piece of literature Dakota is a unique piece of literature that captivates the reader in a very unusual way. From the action-adventure type music, and the words coming at by you gives off a cinematic experience that I enjoy. I was pretty captivated and, had I more free time, would have “watched” the whole thing.
Jessica Pressman offers some interesting points about the cinematic type literature that I find particularly interesting. She notes in Digital Modernism “Dakota is exemplary of digital modernism because it adapts a literary text and technique from modernism in order to challenge the status quo of electronic literature and our assumptions about it” (81). The technique she refers to is the fact that this text “evades reader-controlled-interactivity”, which is what happens with the more cinematic experience of the text: it happens as soon as you press play with you or without you. It challenges the status quo of conventional aspects of electronic literature into the remediation of electronic literature. It really is an example of an evolutionary feature of electronic literature even though electronic literature has only been around for a little period of time compared to physical literature.
This brings me to another point about the evolution of electronic literature. I’ve never seen any literature like Dakota, nor have I imagined a text appearing like a movie flashing before my eyes. It’s intriguing. And now that I’ve seen it once, I definitely want to explore more of this non-conventional piece of electronic literature.
This screenshot is taken from the screen in which “Dakota” displays the text fading out of the camera in a cinematic way
I found “Dakota” to be a really fascinating and engaging piece of literature, but the work itself, and our reading pertaining to it, left me with some questions about its deliberate inaccessibility.
I found our reading’s point about “Dakota’s” deliberate difficulty encouraging close reading to be well-argued, but I can’t say it had me entirely convinced. I enjoy reading poetry, and I always do it with a pencil in hand, marking various literary devices and making constant notes to myself. I read the poem a minimum of two, if not three times (and possibly many more, depending on its difficulty). “Dakota” took everything I knew about reading poetry and flipped it on its head. There was no slow, purposeful reading. No marking up the text. I found myself more focused on simply trying to keep up with the words flashing across the screen than making meaning. “Dakota,” at least for me personally, did not encourage close reading. However, I will be the first to admit that Pressman is clearly a lot more knowledgeable about poetry than I am, and that my failure at close reading may have been more of a function of my own shortcomings as a reader than of the text itself.
However, in my personal, subjective experience of “Dakota,” I think this inability to close read may have been the point. “Dakota,” to me, called into question our entire practice of reading poetry. “Dakota” suggested that the way we’ve been taught to read – slowly, deliberately, and carefully – is limiting. Not all works are meant to be experienced that way. For “Dakota,” the form is just as important as the content. The stark contrast of the black flashing text and white background, the frenetic drum beat, the anxiety a reader feels when she can’t keep up with the text are all as integral to the experience of reading “Dakota” as the text itself.
To me, the inaccessibility of the text had a purpose. Before I read the article and learned “Dakota” was a reworking of Ezra Pound’s “Cantos,” the poem to me seemed to be about youth. Youth is characterized, in many ways, by speed and urgency. Everything feels intense and immediate. Much of the experience of being young is a sort of confusion, trying, and often failing, to cobble together meaning and understanding from limited experience. Reading “Dakota” mirrored that experience. As a reader, I wanted the text to slow down. I wanted someone to tell me what was going on. I was trying and failing to keep up with the progression of the poem, plagued by a nagging feeling that I was missing something extremely important, that if I didn’t catch every word and detail, the whole point of the piece would be lost. None of that experience came from the text itself, but rather the delivery of it. Had I read “Dakota” the way I would traditionally read poetry, I would have lost that layer of meaning entirely.
Pressman identified parallels between “Dakota” and cinema. I experience this difficulty “reading” a text frequently with films and television. I get really frustrated when the lighting is dark and I can’t really see what’s going on, until I slow down and realize that maybe the experience of straining to see and not getting a clear picture is, in fact, the point. Source.
I think digital literature in general has a way of questioning our traditional reading practices and playing with our balance between form and content. While I do agree that there is a lot of close reading to be done in “Dakota,” plenty of allusions to tease out and the like, I think the work calls into question our entire conception of what it is to “read” a poem. I believe “Dakota” encourages us to read our experience far more than the text itself.
The intentional fallacy, as we learned in our last class, refers to the fallacy in literary theory of putting more worth in the author’s intention regarding a work than in the reader’s interpretation. But how is this idea of the author’s intention defined when a work has no clear author?
One such example is the case of the Twitter bots we examined in class. The authorship of these programs usually involved one person both writing the text the bot would work with and programming the bot to post the text, but as we know already, a computer program randomly arranges the text and posts it according to a set time increment. This muddies the waters when thinking about authorship or who the true author is, especially in cases where the text being modified was written by a third person. @poem_exe is an example that contains three possible “authors,” or even four if you add in the creator of the technique used to produce its poems.
Randomness, in the case of AIs or Twitter bots, is what upsets our seemingly static notions of authorship or owning rights to works of art. Inserting a computer program that automatically randomizes text into the mix only serves to blur the already-foggy definitions of originality, authorship, and who really owns a work. Robert Hart’s article in Quartz Media explains that currently, “AIs in the US can not be awarded copyright for something they have created.” Despite there being no formal requirement for human authorship in the US Copyright Act, “the courts have always assumed that authorship is a human phenomenon” (Hart). It remains to be seen whether AIs will be able to claim ownership of their creations in the future, but for now, humanity has not yet begun debating the question of AI rights.
A psychedelic art piece created with Google’s DeepDream algorithm (click image for source). Last year, two such pieces sold for $8,000, and the money went to the artists who claimed to own the images.
Anyone can author a novel, so long as there are loose definitions of what it means to “write” and what consists of a “novel.” The National Novel Generator Month is a month-long event where participants create 50,000 word-plus novels with the help of algorithms. Darius Kazemi, a programmer and creator of the event, gave participants a lot of freedom since there are very few rules to get involved. The rules are that the novel must have a title and contain at least 50,000 words. These events and what they produce really push the limitations of literature and raise many questions. One novel that was particularly interesting to me was “50,000 Meows” by Hugo van Kemenade. His code replaces every word in the classic novel The Jungle Book by Rudyard Kipling with “meow,” keeping the length and the punctuation of every sentence the same. This is not a book by most people’s standards since there is no plot and really just one word 50,000 times, but alas it meets all the requirements of the event, showing that the event really is pushing what it qualifies something as a novel. Since there is more than one writer involved in the authorship of the novel, it raises the question about the true authorship of these types of novels and how far electronic literature can go before it becomes problematic to society.
After reading Will’s blogpost about the worst possible thing that can come out of Artificial Intelligence is take-over or mass genocide of the human population, I started to think about other more likely consequences of this technological advancement. If technology advances more and more, machines will improve and novels will actual be able to create stories that make sense. Eventually there will be no need for people to make the stories because we’ll have machines that are much more time efficient. Something that is more likely and therefore more problematic that could come out of this is that though computers might be able to create stories on their own that make sense, new novels will not reflect the human experience or true human emotions because they are not being written by a computer. This is something that could easily be overlooked but how can a machine understand emotions of love or regret or revenge? Even Will pointed out that AI does not have emotions, which means no matter how much technology progresses, books generated by a computer will never be the same as books generated by people. If computer-created novels were eventually the only on the market, there would be questions of whether this technology was censoring society and eliminating some freedom of speech but again this is unlikely.
There are a lot of overlaps in this class with another one of my classes: Intro to Digital studies, which deals with issues and the history behind artificial intelligence and virtual reality. While reading about randomly computer-generated novels, I immediately thought about AI technology, and DIG 101 (Introduction to Digital Studies). Specifically, I started thinking about if AI technology could soon be writing these computer-generated novels, and all of the new questions that would rise about authorship and compensation.
I think it’s more than likely that this could happen, and although the randomly computer generated novels are not that coherent; they probably will be as technology progresses. In DIG 101 our conversations that focused on the dangers of AI stemmed from the science-fiction book Neuromancer. In the book, the line between AI and human blurs as well as the fact that humanity itself is starting to fall to the new age of artificial intelligence. This idea, although it seems not likely has been warned by Elon-Musk, Bill Gates, and many other important figures in today’s generation.
I found a video on YouTube that discussed the positives and negatives on Artificial Intelligence, and it came up with some good points. The positives were basically that AI with their machine learning capabilities have the potential for things like self driving cars and ultimately make our lives better with a variety of tasks that would either be to tedious or out of grasp of the human ability. And obviously the negatives would be the possibility of a take-over, enslavement, or mass genocide of the human population. There was a scary quote from the video from an AI researcher that says, “the AI does not hate you, nor does it love you. But you are made out of atoms which it can use for something else” (AI).
So back to our class, if an AI has the potential to do all of this, it will probably also be able to start generating novels because the two technologies are not that far off from one another. I also can see the randomness of these novels becoming more coherent as time passes because of the improvement in technology, akin to the improvement from the AI Eliza to Siri. In general, I find it interesting to see the progress in technology, and to ponder these different possibilities that can come out of these advancements.
(Please don’t press more videos for your own discretion, and you can stop video at 3:30)
Bots are permeated throughout the modern day Twittersphere, and they assume many various forms. Some bots manipulate words and instigate conversation, where as others take an artsy approach and produce random images. Others seem to be there just because they can be. For example, @everyword was a project, started in 2007, in attempt to tweet every single word in the English language. This task was completed seven years later and spurred a revolution of parody accounts such as @mispeleveryword.
These accounts are all in relation to the article written by Christopher Funkhouser on First-Generation Poetry generators. This article describes the first uses number generators to produce incoherent, yet methodical poetry. This article also discusses permutations in these generated texts. Permutation is a certain way a group or number of words can be arranged and rearranged. Randomly generated texts use this as they are coded w a set list of words for different parts of speech and then selected at random by a random number generator. The result is a confusing, yet readable collaboration of words that can be astonishingly thought provoking and in some cases human-like.
The famous Twitter bot Olivia Taters is one that pushes the limits of bots. Accidentally created by Rob Dubbin, Olivia taters has been especially notorious for closely resembling a teenage girl, even tricking people into fully believe it to be so.
Computer generated randomness has come a long way since its original uses in the Cold War, and continues to be developed in various ways. The more intricate the code the more believable and comprehensible. As technology advances, the question arises of who can be credited with authorship, the computer or the coder? Depending on the input given by either component, my answer could honestly go either way, however I feel there is no way to decisively give credit where credit is due.
Works Cited:
Funkhouser, Christopher. First Generation Poetry Generators. Mainframe Experimentalism: Early Computing and the Foundations of Digital Arts, 2012
Last semester I took a class on Posthumanism. It was one of the most thought-provoking classes I’ve ever taken, and forced me to question a lot of my deeply-held beliefs. The main questions of the class included “What does it mean to be human?” and “Is there even a universal definition of humanity?” I came out of the class with the conclusions of “I don’t know” and “Probably not,” which I think was likely the point. Nevertheless, I still believe there are a few characteristics that, while not entirely universal, are pretty emblematic of the human race. Among them is the ability to tell stories.
Apparently, I’m not alone in this belief. A quick Google search brings lots of results on this topic, among them, this pretty popular book (187 reviews!) that is likely a better authority than the opinions of some inexperienced undergrad.
When I heard we would be studying computer-generated novels, I felt a little uneasy. If some form of artificial intelligence was capable of telling a story, would we have to consider it human? Would it be deserving of rights? It would raise some complicated ethical questions.
I was relieved (albeit a little disappointed) to see that these computer generated novels were not original works and lacked any narrative structure. They were created by copying existing works in the “cut up” method from Monday’s reading. Furthermore, they weren’t really stories, but rather loosely connected sentences amounting to around 50,000 words.
It appears that while people can program Twitter bots or write code that creates a “novel,” there is no true, original, or organic generation of stories or ideas from these machines.
I found the most interesting part of the novels we explored to be some of the images within Generated Detective. I love creepy, eerie, unsettling things, and I’ve included a few of my favorite images.
A panel from Generated Detective, depicting a large-eyed doll that looks like something straight out of a horror movie.Another horrifying Generated Detective panel
I looked at Generated Detective before I finished reading the article explaining the various novels, so initially, I thought a computer had selected these images. That was exciting! Could a computer understand what constitutes a creepy aesthetic, or at least be programmed to select creepy images? Upon reading the article, however, I realized that the programmer himself was the one who selected the images. So the most interesting, unsettling part of the work was curated by a human.
Since the time the article was published, he’s updated the code so that the computer automatically selects the images, but because those images are from the first few issues, I’m assuming, hopefully not erroneously, that they were selected in the human curation phase. I also assume this because the most recent issue, which I’m assuming has computer-selected images, is far less interesting.
A very uninteresting (and I’m assuming computer-selected) image from Generated Detective
I’m by no means trying to downplay what a feat of coding it is to create programs that generate these novels. But while they’re interesting from a technological and theoretical standpoint, they, in my opinion, lack any true literary merit. Storytelling, it would appear, still belongs exclusively to humanity.
I’m curious as to when (if ever) we will be able to create technology that can truly generate original stories. That may be a science fiction dream that cannot ever be realized. I don’t know enough about computer science to say. If we do, however, it will likely force us to expand or alter our definition of what it is to be human.
People today utilize randomness to program bots in creating poems and literature based off fairly simple grammatical structures and a varying lexicon. However, most people do not know that over three thousand years ago, kings of the Shang Dynasty were implementing the technological equivalent of this at the time to create one of the first forms of writing in human history. Religious leaders would heat cow bones and turtle shells using fire. The heat of the flame would then crack the bone. The king would often ask the Diviner to ask the Heavens a specific question and the cracks would be viewed as the response. These cracks would then be interpreted as messages from the Heavens and relayed back to the king. The Shang would use cow bones and turtle shells because the bones were flat, smooth, and larger than the average animal. This practice became the first form of recorded writing in China. While the people at the time thought the bones cracked a certain way for a reason, these cracks were about as random as the poems created by bots and generators. The difference is people today are far less superstitious and understand that whatever is displayed can have no literal deeper meaning. Despite this fact, people enjoy trying to interpret these randomly generated poems the same way Shang Diviners would interpret cracks in the bone. It is interesting to think that in another three thousand years how people will view our forms of random generation now, especially with the accelerating advancement of technology.
Oracle bone with interpretations by Shang Diviner of cracks. (From Wikimedia, “Shang Dynasty Inscribed Scapula”.)
You must be logged in to post a comment.