Skip to content

Generative Living Pixels

An Interactive Bioart Project Using Slime Mold’s Memristive Properties to Transform Human-Drawn Pixel Art Into Dynamic Visual Output

TEAM: Ruyan Chen

MY JOB: Grow Slime Mold, Circuitry, Coding, Interaction Design, 3D Modeling & Printing

TOOLS: Slime Mold, Microcontroller, CircuitPython, P5.js, Houdini, 3D Printer

PROJECT OVERVIEW

Inspired by the BioComputer Music work by scholars Eduardo Miranda and Edward Braund, Generative Living Pixels involves mapping RGB values of human-drawn pixelated paintings into voltage, applying voltage to slime mold which function as a bio-memristor that introduces randomness, reading and mapping output current back to RGB values to create an output that cannot be predicted.

CONCEPT & INSPIRATION

The concept for this project is inspired by the groundbreaking research of Professor Eduardo Miranda and his colleagues, as detailed in papers such as “A Nonlinear Approach to Generate Creative Data using Physarum polycephalum-based Memristors” and “BioComputer Music: Generating Musical Responses with Physarum polycephalum-Based Memristors.” These works detail that the memristive properties of slime mold enables it to remember the history of charge it is being applied and alter its resistance as a function of the previous charge that has flown through it. They explore the use of this property to create a bio music composer, highlighting the potential of integrating biological systems into creative processes.

We are in an era of burgeoning AI-generated content (AIGC) and generative media, which are expanding our creative potential by enabling new forms of artistic expression. I’m interested to find out how the memristive properties of slime mold can offer another avenue for exploration in this space.

OBJECTIVES

  • To utilize the memristive properties of slime mold in generating creative visual output.
  • To reconsider the intrinsic roles that living organisms, in this case, the slime mold, play in our technological and creative expressions.
  • To explore the potential of the collaboration between human creators and non-human living organisms in creative processes.

PROCESS & METHODOLOGY

Fabrication of 3D Printed Receptacles
I used the paper “A Method for Growing Bio-memristors from Slime Mold” by Professor Miranda and his colleagues as a reference for fabricating 3D receptacles. These receptacles are essential for containing the slime mold and integrating it into the circuit.
Fabricating components in 3D printer.
A fully assembled receptacle.
Growing Slime Mold
I cultivated slime mold on filter paper and agar in petri dishes, regularly feeding them water and oats to sustain their growth. Due to their susceptibility to contamination, I transferred healthy slime mold to new petri dishes weekly to ensure a consistent supply of viable samples.
I used the paper by Professor Miranda and his colleagues as a reference for cultivating slime mold in the receptacle. Agar and oat flakes (food for slime mold) were added to each of the receptacles’s chambers. Slime mold was placed in one of the two chambers so that it would grow towards the other chamber and form a protoplasmic tube that would function as a living wire.
It took numerous attempts to consistently produce healthy protoplasmic tubes. Due to the slime mold’s susceptibility to contamination, I created new protoplasmic tubes weekly to maintain a reliable supply.
Circuitry
With guidance from Professor Mark Olson and by referencing Professor Miranda and his colleagues’ paper, “A Nonlinear Approach to Generate Creative Data using Physarum polycephalum-based Memristors”, I built a circuit and wrote the CircuitPython script that would map the RGB values of pixelated paintings into voltage, apply the voltage to slime mold, and read the output current to map it back into RGB values.The script can be found HERE.
The script performs a bio-computational task. By adjusting the voltage based on input RGB value, the slime mold’s resistance changes dynamically. This dynamic behavior reflects its memristive nature and influences the output current, which is then mapped back to RGB values to create unpredictable result.
Building Interaction Interface

To achieve the objective of exploring the potential of the collaboration between human creators and non-human living organisms in creative processes, I created an INTERFACE using P5.js on which the users would be able to create pixel art. Using serial communication, the RGB values of pixel drawings will be sent from P5 to the microcontroller running the CircuitPython script. Mapping procedure will then be done, which will eventually send a new set of RGB values back to P5. New pixel drawings are displayed below the original ones.

OUTCOMES

This project was exhibited in Duke University’s BASS CONNECTIONS Laboratory Art in Practice Exhibition on 23 April 2024.

REFLECTIONS

Slime mold exhibits biological intelligence. It’s adaptive and natural. AI systems embodies artificial intelligence. It is often rule-based and constrained by human-designed logic and human-chosen data. It will be fascinating to explore the unique opportunities that emerge from leveraging and combining the capabilities of living organisms and machines.

Through co-creating with slime mold and designing circuits and scripts to harness its memristive properties, I discovered that Generative Living Pixels can serve as a platform to expand the boundaries of co-creation with living organisms. This parallels how contemporary AIGC tools are redefining co-creation with machines. By enabling users to modify and shape the generative system—determining how its memristive properties are utilized—anyone can become a creator.

GALLERY

ACKNOWLEDGEMENT

This project was made possible thanks to the guidance of Professor Mark Olson. Special thanks as well to Professor Nina Sherwood, Lab Manager Kathleen Miglia, Teaching Assistant Ze-Yi and Hugo for their invaluable support.