art, creative tech, interactive design  

 
CREATIVE
Artificial Agency
   > Sour Gummies Don’t Cry

PROFESSIONAL
Exhibitions
Public Experiences
Programs

WRITING & CURATION



About
LinkedIn
Instagram
Artificial Agency:
SOUR GUMMIES DON’T CRY
Performance Lecture, MotionCapture, Web Development, System Design, Tech&Society
2025

An interactive lecture performance that blends code, body, and live audience participation to explore: where is our agency in the age of algorithms?

During the live performance, viewers are given the power to control my body (and my digital avatars) via a real-time participatory application on the web, and I execute their instructions like a humanoid or NPC while simultaneously giving a speech on agency. Through meaningful discomfort with a touch of playfulness, the work encourages the audience to critically reflect on data capitalism, digital body politics, and the latest developments of agentic AI. 


The performance unfolds in three chapters, taking the audience on a journey from simple interactions to layered audio-visuals – blending video essay, live speech, synth sounds, and MoCap animations – culminating in “explosion” of my digital avatars and myself.

As I start to share labor and agency with my avatars as well as the audience, the performance becomes a social experiment that questions who holds power: is it the artist, the audience, or the system itself? Unlike traditional HCI (Human-Computer Interaction), I aim to reconfigure a more collective and fluid type of agency through constant interactions. 

The work was first developed as my M.A. thesis at IMA Low Res, NYU, and later iterated in various academic, artistic, and public contects in Shanghai, New York, Helsinki, and online.
Performance at NYC PIT Pop-up, the Oculus, New York, 2026

concept & research
  
Agency, the fantasy of autonomy, the lucid dream of control. In human-machine interactions and digital interactive experiences, people crave agency and a sense of control. When we press a button or wave at a screen, we expect something to happen right away. This is what technology promises us: immediate and continuous feedback loops that reinforce our sense of control.

However, does that mean we truly have agency? In the age of algorithms, action becomes reaction, shaped by data, technologies, and social media that are meticulously designed to control our cognition and behaviors.

My research centers around three questions:
  1. How is the illusion of agency created and then used as a means of control?
  2. How does the controlling system impact our bodies, resulting in splintered selves and digital dissonance?
  3. What can we do to reclaim our agency, bodies, and stories?



My research include literature and artworks focusing on data surveillance, digital identities and bodies, Cyberfeminism and posthumanism, game theories, and the latest developments of AI agents. It resulted in a scripted lecture that I recited during the performance. 
In the early stage of my research, I created text-based designs to illustrate my findings and thoughts, which ended up in a web-based experience of falling through infinite browser windows to evoke the feeling of getting down a digital rabbit hole.


system design
a system of reconfigured agency
Cyberfeminist scholars pointed out that agency is never solo or binary, especially in the human-technology reconfigurations. In Lucy Suchman’s 2016 essay “Agencies in Technology Design: Feminist Reconfigurations,” she quoted Karen Barad, “(a)gency is not an attribute but the ongoing reconfigurings of the world.” 

If we take into account all the players in a human-machine system, no matter human or nonhuman, “master” or “servant,” agency is relational and contextual. Think about anesthesia, the patient’s agency is not simply transferred from patient to doctor but is distributed across an interconnected system of machines, medicines, nurses, doctors, and the patient. Agency is not a private property that someone can simply take away; it is always changing and becoming in the process of reconfigurings.

Therefore, I designed a system that allows the audience to control my bodily actions, but only through algorithm. It can be seen as an experiment of the reconfigured agency in Suchman and Barad’s writings. I wouldn’t be doing what I’m told to do without the audience participation, and they wouldn’t have the power to control me without the app and the system I designed. A more fluid and collective agency emerges from these ongoing interactions.



work-in-progress system map

technical details

The main interaction relies on a real-time participatory application on the web (via web-socket), which receives audience input from their mobile devices and send to the main screen, processed through a voting algorithm. 

The audience is provided with a QR code, which leads them to the custom web-based application that can receive real-time input from the audience.
Their selections are then processed through a voting system, and the top choice will be announced every 15 seconds unless outvoted by another option.
The digital avatars and animations are made with 3D scanning, MoCap, Blender, and Mixamo. 


Scanning
MoCap
Rigging the scanned model
Animation
The 3D model and its animations are then loaded to the main website using ThreeJS.



iterations and reflection

The performance has taken place both online and in person several times with different audience sizes, ranging from under 10 participants to a full house of nearly 100 people.

Based on post-performance conversations and casual interviews, different patterns of interaction were identified: some people showed a competitive trait of wanting to control the voting results and therefore interacting with the app constantly, while others felt uncomfortable directing the artist’s body especially when there are only a few participants; the more participants there were, the more anonymous it felt to certain people; more tech-oriented viewers spent most of the time trying to figure out technical details, while some people focused more on the lecture.

For future iterations, I hope to optimize the audio-visual performance with custom sound and lights and conduct more formal surveys to collect feedback. I also intend to experiment with different modes of interaction – for example, what if there is only one device so that the participants would send instructions one at a time and therefore take on more responsibility for their choice? Or, instead of a voting system, is there another algorithmic system that illustrates the idea of reconfigured agency? Online performance on Twitch in April, 2025


In-person performance at NYU Shanghai in June, 2025

© Mengna Da 2025

Index   Next →