Can Fungi Compute?

Introduction

The Computer Beneath Our Feet

Beneath soil, fallen leaves and decaying wood, microscopic fungal threads branch, encounter obstacles and sometimes fuse into interconnected pathways. Their architecture emerges through growth rather than deliberate assembly.

The network has no central processor. No engineer designed its connections. No factory assembled its components. It is alive. It is fungal mycelium.

Could something like this become part of a computer?

The suggestion sounds almost absurd. Computers are products of precision engineering, built from carefully manufactured electronic components. Fungi are living organisms shaped by evolution, growth and environmental conditions. Yet the distinction is not quite as absolute as it first appears.

Fungal cells maintain electrical potentials. Their tissues can produce measurable electrical fluctuations. Their networks respond to environmental changes, and some recorded electrical events show spatial timing patterns consistent with propagation.

Researchers have identified Boolean-like input-output relationships in physical fungal-colonised material. Others have investigated memory-related electrical behaviour in cultivated fungal materials. Processed mycelium has also been incorporated into experimental physical reservoir-computing devices.

These are genuine research developments, but they are not evidence that mushrooms possess miniature brains, communicate through human-like language or can replace modern computer processors.

Three meanings must remain distinct. Biological fungal computing concerns information processing within living fungal organisms. Fungal-derived electronics concerns components made from materials produced by fungi, whether living or preserved. Hybrid fungal computing concerns systems combining fungal materials with conventional electronics and software.

The title therefore contains two related questions: what can a living fungus do, and what can material grown by fungi contribute to a machine? The answers may differ. Discovering that a useful component need not remain alive is part of the investigation, rather than a reason to abandon it.

Throughout this investigation, we will distinguish what researchers observed from what they inferred, what they demonstrated from what they proposed, and what remains uncertain.

The possibilities are fascinating enough without exaggeration. The journey begins not with a computer, but with a microscopic thread extending through the darkness. A hypha. A network. A living organism whose electrical possibilities we are only beginning to understand.

Chapter 1

The Network Beneath Our Feet

The architecture of a living network

The other landscape

A mushroom is not a computer. That seems an obvious place to begin a book asking whether fungi can compute. Yet the familiar mushroom is not the most interesting part of this investigation. The living architecture lies beneath it: microscopic fungal threads called hyphae, collectively forming the mycelium.

Hyphae extend from their tips, branch, and sometimes fuse. They explore soil, wood and leaf litter without a central plan. There is no fungal CPU waiting somewhere beneath an oak tree. The network develops through local growth, not instructions issued by a hidden command centre.

This is the world of fungal mycelium. To understand why scientists have begun investigating fungi as possible computational materials, we must first understand the structures they produce—and resist the temptation to call every network a brain.

The mushroom and the organism

In many filamentous species, the extensive vegetative body consists of hyphae: narrow tubular structures that grow through or across their surroundings. The general organisation of filamentous fungal networks is described in the review by Fricker and colleagues. [1]

A hypha extends chiefly through growth at its tip. Behind that tip, branches can form. Some species produce cross-walls, called septa, that divide the hypha into compartments while still allowing certain materials to pass between them. These are established features of filamentous fungal biology, synthesised in the network literature. [1]

Fungal anatomy varies widely. Yeasts, moulds, mushroom-forming fungi and other fungal groups cannot all be described as though they shared one identical body plan.

Our technological investigation concentrates mainly on filamentous systems because their branching architecture and electrical properties have attracted unconventional-computing researchers.

How a network emerges

Imagine a germinating spore producing a short hypha. As it grows, the hypha branches. Some branches encounter rich substrate; others encounter barriers or poor resources. Local conditions influence the rate and direction of growth. This is a conceptual illustration of established growth responses, not a description of one particular experiment. [1]

Compatible hyphae in many filamentous fungi can fuse, a process called anastomosis. Direct microscopy in Neurospora crassa has demonstrated specialised conidial anastomosis tubes during colony establishment. Separately, Olsson and Gray (1998) observed labelled material moving around basidiomycete colony margins in patterns consistent with connected routes; that tracer study did not directly image each fusion event. [2; 3]

The network is neither a fixed wiring diagram nor a random pile of threads. It is an evolving biological structure shaped by metabolism, developmental regulation, physical constraints and environmental feedback. Fricker and colleagues review the relationship between network architecture and resource movement, including the limits of present knowledge about larger-scale coordination. [1]

A computer network is manufactured. A fungal network is grown. And growth changes the network itself: new branches appear, some connections become important, and resources are redirected as circumstances change. The hardware—if we dare use that word this early—is continually rebuilding itself. That is intriguing architecture, not proof of computation.

A network without a headquarters

A large fungal colony may experience several environments simultaneously. One region reaches nutrient-rich material. Another dries. Elsewhere, hyphae encounter a competitor or a physical obstruction.

Growth patterns and resource allocation can change across the organism. The resulting organisation may appear purposeful, yet it can arise through distributed physiological processes without conscious planning.

Calling this a decision may be useful shorthand in some biological models. It does not demonstrate deliberation, awareness or a central controller.

The lesson is not that fungi secretly possess a brain. It is that complex coordinated patterns can emerge from many interacting local processes.

Transport and connection

Fungal networks are not merely shapes. They can redistribute materials across connected regions. In a 1998 tracer experiment, labelled phosphorus and an amino-acid analogue travelled along distinct routes through intact Pleurotus ostreatus and Schizophyllum commune mycelia; a third species showed much slower or more restricted movement. The authors found that simple diffusion alone did not explain the observed patterns. Different mechanisms of transport may contribute in different species and circumstances. [1; 3]

Not every visible branch carries resources equally. A connection may change its activity as the colony develops, and the details differ by species and environmental conditions. [1]

These transport properties matter because a network's geometry can influence how quickly or effectively resources reach particular regions.

However, a network optimised for biological transport is not automatically an electrical information network. The two functions require separate experiments and separate evidence.

The electrical boundary

Fungal cells are surrounded by membranes that maintain differences in ion concentration and electrical potential. Ion transport supports nutrient uptake, pH regulation and growth. The subsequent historical electrophysiology chapter examines direct measurements of those gradients. [4; 5]

These electrochemical gradients provide a reason to investigate fungal electrical behaviour. They do not by themselves show that messages travel through the mycelium.

A voltage across one cell membrane is different from an extracellular spike recorded by electrodes inserted into a substrate. It is different again from a signal intentionally applied by an engineer to a fungal-derived material.

Keeping those phenomena distinct is essential if we want the story to remain scientifically trustworthy.

Why the network interests engineers

Engineers studying unconventional computing sometimes seek physical systems whose internal dynamics transform inputs into complicated outputs. A heterogeneous branching network may provide interesting electrical pathways, nonlinearities or history dependence.

Fungal growth also suggests an alternative manufacturing route. Rather than fabricating every microscopic feature, perhaps a biological organism could produce a useful material that is subsequently processed and integrated into electronics.

These are hypotheses about technological value, not achievements guaranteed by the mere existence of mycelium.

A useful fungal device would need reproducible properties, controlled interfaces, suitable lifetime and a measurable advantage over simpler alternatives.

Evidence in brief

Established: many fungi grow through hyphal extension and branching; compatible hyphae can fuse in certain species; fungal systems transport resources and respond to environmental conditions.

Plausible: specific network geometries or growth conditions may produce reliably useful electronic functions.

Unsupported: the claim that a branching fungal network is equivalent to a nervous system or that its architecture alone demonstrates computation.

To understand what such a network might transmit, we must first understand the electrical gradients maintained by its cells.