Event “Grow with the Glow” at MIT Future Fest — Cambridge Science Carnival, Boston · Sun Oct 4, 12–4 PM · come say hi →
FunguyLab
Bio-art & research · FunguyLab

Drawn by light.
Grown by life.

A photophobic fungus. A low-power laser. A neural model that learned how living tissue spreads. Together they print designed forms — in a living organism, with no ink and no contact.

Scroll

Mycelian Micro is a living-image project. The medium is not paint or pixels — it is Mucor, a fast-growing fungus that recoils from light.

We give it a boundary made of light, and it grows the rest itself. Every piece is drawn by a machine and finished by an organism — a collaboration between an engineered system and a living one.

Selected work

From a neural simulation
to a form grown in living mycelium.

Grid showing four designed forms progressing from neural-cellular-automaton simulation to physical Mucor fungal prints on agar plates
Fig. 01Four target forms, left to right: neural growth simulation → laser-contained boundary → the finished print, grown in living Mucor on agar.
The method

Light that draws.
A fungus that fills in the rest.

Mucor oligosporus is strongly photophobic — it retreats from light at the cellular level. A low-power 405 nm laser scans continuously across an agar plate, tracing the boundary of a target shape. Wherever the beam lands, growth is inhibited; wherever it does not, the fungus is free to expand. Over hours, the mycelium fills every unlit region and stops cleanly at every lit edge — printing the design in living tissue.

01 / Observe

See it grow

Time-lapse frames of real Mucor are segmented moment by moment, tracking the branching white network as it spreads across the agar.

02 / Predict

Learn its rules

A neural cellular automaton — a grid of cells learning local spreading rules — reproduces Mucor's morphology with no hand-coded biology.

03 / Contain

Steer with light

A moving laser boundary guides real growth into the designed shape. Simulation and organism converge on the same form.

Fig. 02 · in silicoNeural morphology simulation — the model growing a fungus from learned rules alone, before a single spore is placed.
Simulation and reality comparison: a laser boundary steering Mucor growth into an S-curve on a petri dish
Fig. 03Dynamic light containment — simulation (top) and reality (bottom). A live laser edge holds the growing boundary in place.
The instrument & the interface
The light-printer — a low-power laser scanning a living plate
The interface — a shape is drawn, then handed to the laser as a path
Plates & process
Rendered simulation next to a printed living fungal artwork
Rendered simulation beside the living print it predicted
White Mucor fungus growing across an agar plate
Mucor mycelium across the agar surface
Real Mucor morphology used as training data
Real Mucor morphology — training data for the model
The light bio-printer beside inoculated agar plates ready for growth
The light-printer and inoculated plates, ready to grow
Temporal segmentation frames tracking fungal spread
Temporal segmentation — the organism, tracked in time
Wavelength and power comparison of fungal inhibition response
Wavelength × power — mapping the fungus's response to light
Fig. 04 · motionThe printer at work — a full growth cycle, condensed. Living tissue resolving into a designed image.

“A deeper question than making a picture: it is about interacting with a living system — how an organism responds to, moves with, and is guided by light over time.”

In the lab now
Living record

Photography, built into the process

We are bringing time-lapse capture directly into the control platform — monitoring a print as it grows and building a cinematic record automatically. Alongside it, a compact motorized gimbal rig designed specifically for filming fungal prints: preset angles, smooth moves, repeatable shots.

Beyond one fungus

New species & the behavior of light

The method began with Mucor; we are now testing it on other species, and studying how light shapes fungal behavior — not only its visual form — as organisms respond to and are guided by light over time.

Play

A physical bio-game

Out of that research, a prototype installation: use light to interact with living fungus in real time. Part game, part experiment, part art — a new kind of interface to a living medium.

Fig. 05 · new mediumThe method, moved to a slime mould (Physarum) — a different organism, guided by the same light.
Yellow Physarum slime mould guided by light into a clean triangle on a white substrate
Physarum steered into a triangle by light
Light-controlled slime mould growth on an agar plate
Light-guided containment — a second organism, the same principle
Presentation of the research at SIGGRAPH Asia 2024 in Tokyo
Presented at SIGGRAPH Asia 2024, Tokyo
Presentation at the BioFabricate summit
BioFabricate summit
Origin & recognition

The project grew out of published research reframing fungal growth as a graphics-generation problem — treating morphology as a two-dimensional time-series to be learned, simulated, and then physically steered with light. It was developed across MIT CSAIL, Brown University, the Rhode Island School of Design, and the Royal College of Art, and first presented at SIGGRAPH Asia 2024. A follow-up, Biocompiler, extends the same light-guided method to living computation and was published at NeurIPS 2026 (Creative AI Track).

Discipline
Bio-art · living media · computational morphology
Organism
Mucor oligosporus — food-safe, non-pathogenic, fully compostable
Lead researcher
Kexin Wang — FunguyLab / MIT CSAIL · Brown University · RISD
Recognition
reEarth International Art Prize — Jury Choice · SIGGRAPH Asia 2024 · NeurIPS 2026 · BioFabricate
Kexin Wang, Ivy He, Jinke Li, Ali Asadipour, Yitong Sun. “Exploring Fungal Morphology Simulation and Dynamic Light Containment from a Graphics Generation Perspective.” ACM SIGGRAPH Asia 2024, Art Papers. Tokyo, Dec 3–6, 2024.
dl.acm.org/doi/10.1145/3680530.3695440 ↗

Kexin Wang, Marwa AlAlawi, Minglu Fang, Marcello Tania, Jiaming Liu, Ruipeng Wang, Qingfeng Zhou, Stefanie Mueller. “Biocompiler: Compiling Spatial Network Optimization Problems into Living Solvers with Digital Fabrication System.” NeurIPS 2026, Creative AI Track.
openreview.net/forum?id=lQFpP0YF2l ↗