Algorithm Jamplifying
Thesis project
Time: September 2018 - current
Individual Project
Exhibited at PRIMER19, New York, June 2019 / Die Digitale “Digital Overload”, Germany, Dec 2019 / NeurIPS 2019 AI Art Gallery
Responsibility and Workflow:
- Concept development : Re-imagined and re-designed users’ relationship with AI and social media in order to give user more agency and stimulates creativity in self-expression + culture production. Developed an organic system to achieve this goal.
- Research : Used Chinese social media and Chinese people as case study. Researched on topics including the current issue on AI algorithms embeded in social medias; user behaviors and tendency on social media useage; the relationship between social media and fabrication of pop culture
- Coding and prototyping : Designed and prototyped a website platform with 4 functional generative AI algorithm on it for users to generate social media contents and share them directly to social platforms.
- Workshop : Gave workshop on instructions of the website platform and facilitated the styling process, reflect on the insights and modify the platform accordingly.
- Post-production : Generated videos, collage, print materials, documentation,and built the installation to present the project.
INTRO
Algorithm Jamplifying imagines and develops new behavior and aesthetics evolving in Chinese pop culture. Users and cultural producers adopt A.I. algorithms as a new way to generate memes, videos, and aesthetics. By facilitating this collaboration with A.I., users are able to exert control over their presence on social platforms and create content that amplifies their desires, values, and creativity.
The work consists of 4 parts: the website platform for the public to generate jamplified contents, the direct output (social media meme, videos, aesthetic people generate) from the website platform, social media feed showing the way people consuming these AI content, and glossary explaining the spirit of the project.
BACKGROUND
AI is living in every corner of our lives, it’s embedded in CCTV, mobile device, social platforms, and more. Its ability for recognition, recommendation, prediction, and content generation is constituting and fabricating a hyperreality for Chinese people, which manifested through Chinese pop culture. We are living in a bubble that AI selected for us, our imagination is trapped. AI has pushed China to its next level of Ultra-Unreal, the reality of overloaded distorted information, the enormous growth of the Chinese economy, warp speed modernization, and political oppression.
Algorithm Jamplifying is a strategy for users to escape this flood and reclaim control over social platforms.
RESEARCH
My research questions sought to consider how Chinese social media and AI algorithms would change both functionally and aesthetically to accomodate for a more authentic self-expression platform.
︎How is the current AI algorithms embeded in Chinese social medias influencing the formation and expression of modern Chinese pop culture?
︎How can AI algorithms be more transparent and viewed as a creative inspiration rather than a algorithmic restriction to imgination?To begin, I did lots of visual and data investigation on the current Chinese pop culture and social media platforms including Tiktok,Weibo, Wechat.
I used web scraping technique and myself to collect videos, text memes from Tiktok and Weibo as a way to prepare for my data visualization analysis and initial observation on its algorithmic mechanisim. My observation mainly focused on the platform’s recommendation systems, ranking systems, tagging strategy, and catagorizing strategy.
For my video data visualization analysis, I used video collage, played with scale and number to gather insights from my data using large scale projection mapping. As for Weibo, a text-focused social platform, I used algorithm to calculate and analyze those text memes, then used AI algorithm to re-generate these memes to test my observation.
DESIGN / PROCESS
︎DIAGRAM
Jamplified content generator system diagram:
The diagram illustrates conclusion from my research and the whole design system concept: The first row shows the feedback loop between AI algorithms and user behavior and content. AI algorithms read user profile, record user behavior, tag user uploaded contents and use these data to create filter bubbles, give personalized recommendations, and guiding speicific trends.
The row below explains my design – Use AI to interfere with AI. It achieves this by initiating a feedback loop that starts by pulling content uploaded by users from exsiting social platforms, populating a database for a new neural network algorithm with that content, generating new content based on that algorithm, and reposting this content on social platforms. After this content is shared on social platforms, it becomes: 1) creative noise for the original AI algorithm to function differently, update itself, creating new filter bubbles; 2) the next round of source material for the AI; 3) creative inspiration for users to produce new content. In effect, the content becomes “jamplified”: Making, sharing, and consuming in this way amplifies the “noise” being injected on social platform and users’ agency.
︎WEBSITE PLATFORM
https://2w1djam.com
The website guide user through the spirit of the concept, and the usgae of the function, enabling users to generate social media content with AI and share them directly to their Weibo, Tiktok, and Wechat.
︎SOCIAL MEDIA FEED
Showing the evolving expression of Chinese pop culture and the new aesthetic of social media.
