3rd Party

Max Oginz

“…it was not so much that the dog and its owner come to look like each other as that they both by degrees take on a trim, unbidden resemblance to a third party…”

Gary Lutz, “The Dog and its Owner”, I Looked Alive



“Who is the you / of YouTube? / You / Tube’s you”

Arianna Reines, Coeur de Lion


 The application of algorithms dictates what we see online––what we watch and share, what we buy, etc. They make the world visible, and this visibility influences our interests, and perhaps even our identity. How much, then, is my interest in cinema, video production and editing, color correction, video games, French philosophers, designer footware––anything––a system of functions repeatedly solved by algorithms?

The author is watching animals sing the Twin Peaks theme song. The web suggests other videos of farm animal harmonies, but it also recommends videos of men shooting a crystal ball with an assault rifle. Who is this third author? If the author watching sees in the first person, and the algorithm solves for a “you” in the second person, can there be a 3rd person, climbing down between these two?

Repetition is important here. Virtual life is purposely habit forming. Even in an attempt to fool an application, unconscious repetition yields accurate results of what the user wants to see or buy, or, as so many videos are monetized, buy by seeing. Can these habits be something other than rote? Can they be refrains, emphasizing a passage between chorus and verse? Can this sense of repetition personify the same way it influences. If watching certain videos trigger the algorithm to present advertisements of a political candidate I don’t support, is it a mistake? Echoing Benjamin, what happens if we take these deviations as connections, rather than dismissing them as derivations?

If these deviations could form an architecture of virtual selfhood, can this architecture be looked through in the same way Sigfried Gideon described looking through the bars of the Tour D’Eiffel and Pont Transbordeur?

This project combines text and found video to confront the simultaneous, repetitive structures of virtual life. I will edit video culled from online sources, with an emphasis on what algorithms show me. The source material might include anything from streaming service recommendations to YouTube spirals. I may also choose to incorporate images or Wikipedia articles. Each (very) short film will stem from an input––the answer to a repetitive question that (hopefully) evokes the mixture of boredom and FOMO that can shape our online preferences, leading us to both sincere discovery (say, a new book, TV show, or film that speaks to us), but also leads us toward entertainment or commerce, a new pair of Doc Martens or a spiral of viral Vines or Tiktok videos. But are these deviations from good sense (or good taste), any less intellectual? I intend the effect to be aleatory, disorienting––both irreverent and perhaps also moving or disturbing. The point isn’t to come up with a thesis, or even a hypothesis, but instead to wander within the surreal collage that AI can inspire in the user.

They lose their distinctive shape, swirl into one another as we climb downward, merge simultaneously.” Sigfried Giedion, Bauen zn Franlcreich (Leipzig and Berlin), p. 7.


Dialogic, Not Dialectic

In a dialogic process, various approaches coexist and are comparatively existential and relativistic in their interaction. Here, each ideology can hold more salience in particular circumstances. Changes can be made within these ideologies if a strategy does not have the desired effect.

In a dialectic process describing the interaction and resolution between multiple paradigms or ideologies, one putative solution establishes primacy over the others. The goal of a dialectic process is to merge point and counterpoint (thesis and antithesis) into a compromise or other state of agreement via conflict and tension (synthesis). “Synthesis that evolves from the opposition between thesis and antithesis.”

-“Dialogic” Wiki


The concepts of database, data, algorithm, are therefore at the basis of every artistic and creative process in the era of computer. While historically every artwork constituted a sort of unique specimen in the context of a particular medium, now form and content are two entities which can quite completely be untied. (Christiano Poian 3)

2. Text

Each Video will be prefaced/accompanied by a brief, repetitive question, and an original, cryptic response, from which the corresponding video takes its material. Each response suggests a way to live, an ideology touted online, a kind of circumstance for the film to live inside.

Q: What is a refrain?

A: An Expression: Solve your browser history by combining like terms.

A: A minor setback: You have been logged off for security purposes, please login in again to access services.

A: A system: Buy this and become someone else.

A: A line: –––––––––––––––––––––––––––––––––––––––––––––––––––––––

A: A bargain: fit it into your schedule, your relationship, your budget.

A: A medicine: Wakening with the moon, working out before the sunrise. Teff porridge.

A: Time: Seconds, Minutes, Hours, Days, Weeks, Months, and Years.

Potential Formats

H5P Interactive Video

H5P Branching Scenario


  • What do you think of the potential formats? Do you prefer branching or interactive video, a mixture of both, or neither? I like the idea of branching, but in terms of design I think Divi builder gives me more flexibility…


  • There are other tools and modules I still want to explore, including video sliders and accordion modules. Anything else I should check out?


  • I’m still not positive what exact constraints I’ll stick to in terms of material. If you checked out the branching scenario, you’ll note that the two videos have the same clips, they’re just arranged differently. Is that a good constraint? Can you think of others that might fit with the theme of “refrains”.


  • Did the project’s concept remind you of anything you’ve watched, read, or listened to?


  • On a technical note, how many 30-60 second videos do you think these formats could support? How do these videos run on your computer? I’ve noticed a buffering problem, do you have any recommendations to make them run smoothly re: export formats, file size, etc?



  1. admin

    There are quite a few worthwhile questions here, so I’m not sure I can address all of them in a short span. You’re welcome to check in with me later about some of them.
    It is possible the create a branching scenario without using H5P simply by using links to other pages or sections of pages. I also recall some projects where we also randomized a set of links, so a click would lead to a different page or section each time. But pages are often a bit slow to load, and H5P does not usually have that problem.
    If anyone is doing multiple pages, though, I recommend putting the various pages inside of what WordPress calls a “Project,” which makes things easier to find (we had a case where a student made a lot of pages that weren’t organized under a Project, and it’s been a bit of a problem ever since).
    The Divi Slider is easy to use and will include videos, but it won’t randomize the order of slides. The Divi Gallery will randomize, but won’t, unless I have missed it, include video.
    Refrains does suggest repetitions of various kinds, and in terms of branching scenarios, it suggests the value of “loops,” which return one to previous points in the essay/presentation. Repeating videos in different orders can certainly evoke “refrains” as well. Even some repetition of quotes can give an echo-like feeling. Of course, there are obvious limitations: too much repetition can become obvious; so I’d argue for more sporadic and/or shorter types of repetition.
    Video buffering is certainly a worry in some cases for many projects, and if you put a lot of different videos on a single page, it can very much slow things down. But bear in mind, you may need to disable the visual builder to check how quickly the page and videos are loading. I’ve also been noticing that using “embed codes” (aka iframes in html)–rather pasting a share link in a video module– for YouTube and Vimeo seems to working pretty well, so in some cases, building a playlist on those platforms may be the way to go. I won’t deny, though, that it may take some trial-and-error testing with the use of video.

  2. A Smith

    You clearly have accumulated a great deal of interest and knowledge about this subject, which can be seen in your references and grounded theme. This is an ambitious project, but I foresee the available software tools to be flexible enough to achieve your goal. When you said “Each response suggests a way to live” that reminded me of tarot cards. I don’t have a lot of experience knowledge on tarot, but maybe there’s some connection to your project, especially with the randomization.

    I prefer the branching scenario because that feels more like an algorithmic association. Based off of Randy’s comments, he knows much more than I do about the various DIVI/ H5P tools, but it seems like there are limitations to all of them. The DIVI Slider tool doesn’t seem like the best option, as it won’t randomize and appears limited.

    Randy’s idea of creating a branching scenario using links to other pages seems like the best option in terms of flexibility, something with which I may experiment. I think using the same video in different scenarios works well for the “refrains” idea and will ask the user to come to their own conclusions. For export settings, I would go with WebM and MPEG-4 (MP4) and probably stick with H.264.

  3. Behrang Garakani

    >> On a technical note, how many 30-60 second videos do you think these formats could support? How do these videos run on your computer? I’ve noticed a buffering problem, do you have any recommendations to make them run smoothly re: export formats, file size, etc?

    Three [potential] technical bottlenecks:
    – The server-side: the server has to pull the videos and package them through WP to serve them up for the user, which is competing with other project pages serving up content (processor, memory, disk access)
    – The bandwidth: lots of data at once will tax the connection sending it to the user, again in competition with other pages/content
    – The client-side: Browsers have to render out all the videos simultaneously (processor, memory, some disk access for caching) – this is why social media only plays one video at a time on a feed.

    I like RL’s idea of YT and Vimeo serving the content as it would eliminate two of the three bottlenecks above. You may still have some client issues.

  4. A Robusti

    Hello Max!
    This is an interesting project and it clearly shows that you have put lot of work and thought on it! Good job!
    Both of these potential formats could work with your project; however, I agree that the branching scenario may be more appealing for the viewer especially considering your topic. Sorry that I can’t be of help with the technical aspects!
    Good luck!

  5. L Arango

    Hi Max,

    During the presentation I understood that that the algorithm you were going to use was taking algorithm work and compiling it in a way that allows me to see another dimension – sort of a collage made form collages expressing the same or a different idea.

    >>On a technical note, how many 30-60 second videos do you think these formats could support? How do these videos run on your computer? I’ve noticed a buffering problem, do you have any recommendations to make them run smoothly re: export formats, file size, etc?

    Certainly I think that the user experience is important, but I think that what is more important is the art creation. once you have a working prototype and the art has been crafted, then you can tailor whatever logic you need for speed, memory, storage, etc. I am looking forward to seeing this algorithmic expression.

Submit a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Share This