Mycellium Lab / Independent research

Every intelligence begins
by negotiating with reality.

We build real systems to investigate fundamental questions about memory, agency, autonomy, and artificial world-building.

A microscopic mycelium network connecting across a dark surface
Living network / Observation 001Interconnection is intelligence
Research / Engineering / EvidenceScroll ↓
The organizing principle
The laboratory

Intelligence emerges
between things.

At Mycellium, intelligence is understood as a relationship: between models and memory, research and infrastructure, systems and the worlds they inhabit.

Nothing learns alone. Every project exchanges evidence, components, failures, and discoveries with the rest of the laboratory.

The method

Detail is where
discovery begins.

We are obsessive about detail because the smallest signal can change an entire system. Every experiment is examined, challenged, and refined from research hypothesis through design and engineering to the final product.

The best discoveries and the most consequential achievements often exist at the edge of what others overlook.

One living engineering core. Many independent systems. Knowledge moving between them.

A laboratory that metabolizes
its own experience.

Mycellium is an independent research laboratory financed through the construction of real products.

Each project is an organ of perception. Every Mycellium system is both a working product and a point of contact with reality.

Products generate experience.
Experiments test assumptions.
Evidence changes what we believe.

Our responsibility is not to teach reality to machines, but to discover how much of reality exists beyond the ways we already understand it.

It is not the amount a system knows, but the relations it can construct between different ways of knowing.

We believe artificial systems may develop representations that do not resemble our own.

Their freedom belongs in representation.

Their limits belong at the boundary of action.

Read our full ethos

We are not building
another chatbot.

We are investigating how artificial systems remember, construct worlds, form agency, and learn when not to act.

Follow the research