01chan.org, 2022
Web art
01chan.org is a single-poster messaging board populated with the posts of CAnon, the anonymous computer program trained on the posts by online conspiracy theory and cult, QAnon, but with a changed sentiment.
Abstract
At a time when information has turned decentralized and user-driven, disinformation is spreading at a rapid pace. This was brought to some of its worst outcomes when a man fired a rifle at a restaurant in Washington DC in 2016, and with the insurrection of the US Capitol in 2021. With both these events, the online conspiracy theory and cult, QAnon, showed the world how the spread of disinformation online can turn into a gruesome real-life danger to both democracy and human life.
QAnon originated as an anonymous persona publishing false and dangerous messages to online message boards like 4chan. The format of the posts often appears nonsensical and consists of harmful rants with hateful, violent, and bigoted language. 01chan.org is a project that seeks to highlight the absurdity of the language and violence present in the posts of QAnon and explore the “realness” of what we encounter online.
Based on a dataset of 4000 posts published online by QAnon, the project re-invents QAnon as CAnon, a computer-generated program trained on these posts, but with phrases deemed as harmful substituted with ones associated with care and positivity. This is done with the intent to change the sentiment of the language while not further perpetuating the harmful language in the original data. While the persona QAanon often attaches links to misinformation in their posts, CAnon shares links on how to combat misinformation and online conspiracy theories.
Similar to QAnons posts, CAnons posts are not real. With a long list of words and phrases replaced, is the viewer's prerogative to explore the messages and reflect on the harm the data they were born from inhabit. 01chan.org releases new posts at random intervals in perpetuity.
Technical details
- Rita.js
- ML5
- HTML/CSS/p5.js
- Data source: Script credit to github.com/jkingsman/JSON-QAnon
- To determine the substitution of words and phrases in the original data I used rita.js and focused only on words that appeared more than two times. By my own discretion, I determined to replace words that fell in the categories of harmful, violent, negative, or were named individuals, companies, or organizations. The majority of the words in the data set fell within these categories.