System Art: Couple of possible directions
Ultimate Direction: 코끼리와 장님. Gigantic hyperobject system. Invisible system dynamics – only the subsection of the whole system is visible, and multiple channels shows those ‘surface leve’ dynamics… Audience interaction altering the system. But the whole system can never be fully understood/comprehended. Only the existence of system is vaguely sensed.
However this is a very long-term ultimate direction - Rome is not built in a day. We should start from sub components of system arts… Including following early experiments
Direction 1: Generative-X, Generative Simulacra One of the most simple starting point for portraying system art Generative-X, Generative Simulacra: Simulacra System (Simulation?) auto-generated Generative-Junk Site: Commercial LLM-generatable Junk Site (Sort of a system) is massively generated & presented Generative-Art Biennales: Most art biennales & curations & artworks are massively simulacra/fake/merely combination of vectorised parameters, and thus could be massively generated & presented Mostly quoting on the ‘simulacra’ of our society (Junk Site, Biennale), which can be/should be easily replicated by AI, arguing & revealing the fakeness of those simulacra systems, urging that we should move forward rather than acting like AI creating these junk systems… Audience interaction two-folds Interaction A: New Simulacra/System Creation (New Junk Site Creation Interaction B: Navigating within the newly created system (Navigating through websites, navigating through biennale) Possible thanks to LLM/Generative AIs Yet not the perfect case of system art yet. Good solid starting point though
Direction 2: Abstractive/Conceptual System More conceptual starting point rather than descriptive (Generative-X is descriptive) Key: Clear & Intuitive. Repetition of the key abstractive theme. Beautiful, minimalistic, maximalism-via-minimalism. Sexy. Almost japanese like design. Example 1: Time / Clock Time Digital Time Clock shown in grid. Initially HH:MM synced with real time. As progresses, the time per each clock in the grid changes - and also digital clock shown in different visual states (refer to the System Art: Visual State and System State) Example 2: Time Series Graphs Multiple graphs. Also shown in grid, all starting the same line graph in the very start, starting to change/vary as the interaction progresses. Also shows other graphs than the time series graphs in visual states as the interaction progresses. No explicitly descriptive narrative Interaction mostly on state-based approach Multiple user interaction possible, creating an interference from one another
Direction 3: Multiple Agents/Profiles Simulated Multiple Agents/Avatars/Profiles (동물농장, Minecraft…) Referencing to Generative agents: Interactive simulacra of human behavior hci paper More like a world-building, an obvious/conventional approach when talking about system/world-building, but this also has lots of limitations by talking about humans/inter-human relationships This limitation is akin to what self-portrait/portraits have in a world of painting, versus the vast subject of other things to paint (nature, abstract…) Focusing only on multi agents/profiles is like focusing on portraits/human figure only, whilst focusing on the vast dynamics of system art is putting everything (abstract/descriptive…) into a realm of painting Therefore not my core focus, but good to experiment/test system art hypothesis Ranking Society: One example of non-world building yet multiple agents N people competitive society mimicing ranking system Useful to portrait the competitive society/zero-sum game of the modernity/quotation on educational system
Direction 4: HYPEROBJECT Quoting on whole society, complex society, real-world society Chaos System, Chaotic System, Uncertainty, Quotation on core system dynamics Spatial-Temporal Exploration Graph-Theory Based, State-based system based 코끼리와 장님: We are the blinds, and the system is the elephant. Portraying the world we’re living in - we’re just acting like the blinds!
Here let’s talk about the Generative-X, Generative Simulacra in a bit more detail:
Modern world is full of simulacra. Junk websites, junk conversations (das gerede), junk corporates, junk advertisements, junk competitions, junk biennales, junk reports, junk systems… Most of these merely act as a copy and paste of the already existing system, copying one another… A simulacra… Which can be vectorised! And vectorily generated!
Thanks to the AI: a useful tool to substitute all these simulacra. If there is a way how AI can be integrated in our society for the better of humanity, that is by substituting the simulacra within the world we’re living in, and let humans focus on the pure creative things not handling with simulacra.
Substituing and auto-generating simulacra with AI shows explicitly that all these simulacra systems we are surrounded by could be and should be substituted with AI, and thus their ideologies are no longer valid - We do not need to be obsessed with creating junk sites. We should not be obsessed with creating junk biennales… But people unforunately don’t get this message early (also there are so many scammers who benefit from creating these junk systems) - That’s the reason why we need Multi-Device Web Artwork approach to show these Generative Simulacra in a system art format.
Multiple channles in MDWA act as portraying the simulacra junk system - if it’s junk sites, it shows numerous junk sites created by ai. If it’s biennales, some screens might show the exhibited artworks, some curatorial text, other the binnenale’s website itself. User interaction: 1) Generating new junk-X, and/or 2) navigating through this newely created junk system.
JUNK SITE Interaction: Initially big group keywords dispalyed. User selects one. And then 2nd hierarchy keywords displayed: User selects second. So on and so forth… Going deeper into the keyword tree. And upon these keyword tree selection, the theme of junk sites get more concise and junk site is generated accordingly to match with this keyword… Multiple junk sites with different designs (but perhaps converging) generated in real-time, across multiple channels. Blink Interaction (TBD): Each blink, user navigating through the junk site in multiple channel
Eventually this interaction should be very smooth, a game-like, almost very phenomenological, like you’re just skating on top of the mobile interface with your fingers. Initially when you started to interact you care about the semantics (i.e. the meaning of the text written in every button), but after certain point, you are just clicking the button/selecting the button/or not even clicking the button, sliding through different sections and button is on the way of your finger sliding → And you enjoy the immediate effect it generates across the screens! Almost like a gaming, more adernaline! But this exactly portraits the world we’re living in, the speed in which new junk sites are generated, and the possibility of fast-paced algorithms where it can catch up with this speed generating all these junk sites!
JUNK BIENNALE Interaction: Mobile: Minimalistic Slider (0-360) just number → Screen: As interacted, unprecendelty the global map appears. Do this for lat/lan. And the biennale’s city is located Mobile: 1, 2, 3… Numbers just displayed → Screen: As number selected, XXX (City Name) Annuale/Biennale/Triannle…. Is generated. → User gets the meaning as they interact Mobile: Multiple Icons (With minimalistic indicator) → Screen: Icon is the topic of the biennale. It’s the buzzword most-used keywords that every biennale has: Minority/Postcolonialism/LGBT/Post-digital… All those 굉장히 식상하고 썩어빠진 keywords → And then the biennale starts to be auto-generated… Overall minimalistic interaction where on the mobile side no semantic/text is explicitly displayed - yet the user knows what it means posteriori as the correspoding result is displayed on the screen. And thus the biennale is slowly created. And after creation, it gradually automates & immerses user by creating an automatic/self-evolving system within multiple screens - Yet user still takes control, user can either navigate/edit the system/create new biennale (Need precise interaction design on this part)
So it’s the Generative-Automatic System! (Or generation of automatic system) - Automatically updated & editing aethetic, this being Generated in real time! A new kind of aeshtetic! A new way to portray the system!
This is revolutional akin to the Andy Warhol’s Campbell’s soup cans - Campbell’s soup cans can be interpreted as Generation (Differnt Flavour) of Automatic System (Soup Can) - Successfuly capturing the age of mass reproduction / flavour customisation in the early consumeristic era he’s living in. In the contemporary world we’re living in, it’s about the system - automatically evolving system, a sort of a field-based system… (Website-ecologies, biennale-art world) Which is totally simulacrum! And should be/could be substituted by AI! Generative-X/Generative-Simulacra - The Generative-Automatic System - is the new aesthetic! (Should be clearer in this metaphor yet)
Text written by Jeanyoon Choi
Ⓒ Jeanyoon Choi, 2025