Magnetic Bullshit: Hong Kong Police Edition
A Participatory Installation & Website
Overview
A fridge magnet kit, both physical and digital, featuring the 300+ most popular words gleaned from over 120 Hong Kong police press conference YouTube transcripts. Participants are invited to create their sentences starting with “Hong Kong Police”.
For months during the on-going Hong Kong
protests, the Hong Kong Police held daily
press conferences to address protest events.
Personally, it was difficult emotionally and
mentally to listen to the police’s rhetoric. Yet, I didn’t want to filter out their voices to create an echo chamber. Hence, I decided to use computational approaches to help me process pro-establishment narratives.
Part of a larger series documenting the Hong Kong protests
Public Showcase
Exhibited in CultureHub (New York). The piece, Tai Kwun Contemporary (Hong Kong), and squatted.online (part of The Wrong Biennale).
︎ Talk at Processing Community Day NYC
︎ Interview by Daniel Shiffman at ITP Spring Show
A fridge magnet kit, both physical and digital, featuring the 300+ most popular words gleaned from over 120 Hong Kong police press conference YouTube transcripts. Participants are invited to create their sentences starting with “Hong Kong Police”.
For months during the on-going Hong Kong
protests, the Hong Kong Police held daily
press conferences to address protest events.
Personally, it was difficult emotionally and
mentally to listen to the police’s rhetoric. Yet, I didn’t want to filter out their voices to create an echo chamber. Hence, I decided to use computational approaches to help me process pro-establishment narratives.
Part of a larger series documenting the Hong Kong protests
Public Showcase
Exhibited in CultureHub (New York). The piece, Tai Kwun Contemporary (Hong Kong), and squatted.online (part of The Wrong Biennale).
︎ Talk at Processing Community Day NYC
︎ Interview by Daniel Shiffman at ITP Spring Show
Role
Concept, Interaction & Visual Design, Development
Tools
HTML/CSS, JavaScript, Firebase, Python, YouTube-dl, Adobe InDesign
Concept, Interaction & Visual Design, Development
Tools
HTML/CSS, JavaScript, Firebase, Python, YouTube-dl, Adobe InDesign
Process
I used a command line program to scraped over 120 YouTube transcripts from Hong Kong police’s press conferences, and compile a 627-page source text. While the result of the compiled transcript is not perfect, there is a parallel between a machine-generated interpretation and an authoritarian narrative.
Initially, I created generative text using Markov chain generation and filtering keywords such as “women”, “protest”, “China”. The result was interesting but it was a one-person experience where I dictate the narrative. I wanted to create a participatory experience to include more voices. As part of my analysis, I searched for the most frequently used words, seeing a list of them inspired me to create fridge magnets.
Initially, I created generative text using Markov chain generation and filtering keywords such as “women”, “protest”, “China”. The result was interesting but it was a one-person experience where I dictate the narrative. I wanted to create a participatory experience to include more voices. As part of my analysis, I searched for the most frequently used words, seeing a list of them inspired me to create fridge magnets.
Outcome
The project was exhibited at the Art Book Fair at Tai Kwun Contemporary in January 2020. The location of the museum is not far from key protest sites.
Some of the sentences participants created included “Hong Kong Police target the innocent public”, “Hong Kong Police needs to say sorry”, “Hong Kong Police close to the ends”. Many viewers also took the liberty to re-arrange the magnets into protest slogans and symbols, such as “ACAB”, “Hong Kong Independence”, “721” (Date of an indiscriminate mobster attack on subway commuters).
Web Version
On the web version, participants can scroll through the word list and tap or click on a word to create their sentences. If selected, the word will disappear from the list on the right. Participants can tap or click again on a word in their sentence to deselect it. The sequence of words is shuffled on refresh. Once submitted, they can see all submissions received thus far. The submissions are stored on a database on Firebase.
︎ Live Site
Credits
This series of projects would not be possible without the generous support and advice from Clay Shirky, Genevieve Hoffman, Allison Parrish, Sukanya Aneja, Mark Lam, Rui An, Huiyi Chen, and Vince Shao.