Mycelium as the Intelligence of Occupancy (A Thought Experiment)

Intelligence as Occupancy

In myceliation, what is somewhat obvious, but still fascinating to think about explicitly, is that mycelium’s depth of physical decision-making is entirely directed toward occupying the negative space of its substrate. It expands through structural matrices by filling the interstices; the voids and gaps that define the structure of its environment. Mycelium first takes the shape of absence: the negative space of a static substrate and, simultaneously, the negative space of an ephemeral, shifting environment; the dark matter of the matrix environment (temperature, humidity, competition, ecological cascades). This act of initial occupancy is the primary gesture through which the organism establishes a path toward long-term resource capture, and as such the growth mechanics of efficient spatial occupancy are of singular importance. 

As mycelium engineers we focus on this occupancy, but that focus is only realizable because evolution has produced an organism with nested layers of physical intelligence as a solution to achieving this occupancy. This layered intelligence, not neuronal, is distributed across structures and processes that collectively form a recursive and decentralized decision-making system. In myceliation, the organism actualizes the latent intelligence of environmental complexity into an embodied intelligence through the physical act of occupying that complexity. In a sense, the fungus uses intelligent assembly to grow an inverted echo of its environment: a dynamic counter-shape generated by following, testing, and exploiting the gradients and geometry of possibility around it (ultimately using this counter-shape to overtake the positive space). Something like a casting of intelligence from a physical and ecological solution space.

This way of thinking about mycelium rhymes with the definition of intelligence proposed by Chis-Ciure and Michael Levin in Cognition all the way down 2.0: neuroscience beyond neurons in the diverse intelligence era (Chis-Ciure & Levin, 2025). In their framework, intelligence is not tied to neurons, human cognition, or mental representation. Instead, intelligence is defined operationally as search efficiency; the degree to which a system’s behavior outperforms a maximal-entropy random walk when navigating a problem space. Formally, biological intelligence is the capacity to reach preferred states despite constraints, uncertainty, and limited predictive horizons. A system is intelligent to the extent that it prunes combinatorial explosion, modulates constraints, adapts its available operators, evaluates options under changing conditions, and uses whatever predictive window it has to consistently move toward viable goals.

I use that framework here as an interpretive scaffold rather than as a claim that mycelial intelligence has been formally measured in these terms. Mycelial intelligence certainly has not been formally measured within these terms, and the goal here is not to claim so, but rather to use the framework to explore how mycelial vocabulary may map to this framework. 

This definition, centered on goal-directed search rather than thought, provides a natural language for describing mycelium. Nearly every facet of mycelial growth, from branching patterns and translocation optimization, to damage responses, tip steering, wall remodeling, foraging memory, and exploration–exploitation strategies, functions as a way of intelligently refining search through a high-dimensional environment. In this sense, the behaviors we work with in fungal R&D are expressions of exactly the kind of intelligence this framework is designed to formalize.

Intelligence as Search Efficiency

In the framework developed by Chis-Ciure and Michael Levin (2025), intelligence is defined as the ability of a system to navigate its environment more efficiently than chance. Every biological system inhabits a structured problem space and must move through it under uncertainty, constraint, and limited foresight. Intelligence, in this view, is the measurable extent to which an organism compresses that search, reducing the combinatorial overload inherent to its physical and ecological context.

This problem space is formalized as P = ⟨S, O, C, E, H⟩. The state space (S) encompasses all configurations the system can occupy, from morphological arrangements to biochemical or spatial states. Operators (O) are the transformations available to the system (growth, branching, fusion, contraction, transport). Constraints (C) outline what is limited by structure, energetics, or environment. Evaluation (E) identifies which states are adaptive or preferred without implying intention. Horizon (H) captures the temporal depth of the system’s predictive or integrative capacity. Together, these components define what the organism is, what it can do, and the boundaries within which it must act.

Intelligence becomes a scalar measure of search efficiency, expressed as K. K compares the expected cost of a maximal-entropy random walk through the problem space to the cost of the organism’s actual behavior. Cost inherits physical units (energy, time, or material investment) based on the operators used. A positive K indicates that the system outperforms randomness; the magnitude reflects how many orders of magnitude of work the organism avoids. Intelligence, then, becomes a continuum rather than a binary property.

Systems that avoid futile pathways and collapse vast branching possibilities into productive trajectories demonstrate intelligence. So do systems that actively reshape their constraints, reorganize their available operators, or extend their effective horizon by integrating signals or anticipating environmental change. Robustness under perturbation also signals efficient search. Flexibility matters as well: intelligent systems reliably reach adaptive states through multiple pathways when conditions shift.

This perspective aligns directly with fungal behavior and physicality. Mycelium continuously explores and restructures high-dimensional problem spaces. Its branching patterns, transport networks, wall remodeling, fusion behaviors, damage responses, and exploration–exploitation dynamics all serve to refine search under uncertainty. Fungi prune possibilities quickly, modulate structural and physiological constraints, integrate information across space and time, and maintain adaptive trajectories. The features we rely on in fungal R&D—patterns of occupancy, redistribution, allocation, memory, and repair—are the same behaviors that, in this framework, constitute intelligence.

Intelligence as Mycelial Physicality

If intelligence emerges through how organisms negotiate and prune their problem spaces, then fungal physiology represents a highly coherent, multi-modal stack for compressing search. The following sections explore, as an interpretive exercise, how major fungal structures and behaviors can be mapped onto components of the Chis-Ciure–Levin framework.

1. Clamp Connections

Clamp connections in higher basidiomycetes are an example of micro-scale problem solving that maps cleanly onto intelligence as efficient search. In dikaryotic hyphae, clamp cells form during cell division and orchestrate a brief separation and re-fusion of the two haploid nuclei, ensuring that each daughter hyphal compartment recovers the dikaryotic state. Classic work describes clamps as devices for maintaining a one-to-one nuclei ratio along the hypha, preventing drift away from the dikaryotic condition. More recently, Aanen and colleagues have argued that clamp connections also function as nuclear quality checkpoints: during each division, nuclei alternately enter the clamp cell, and successful fusion back into the subapical cell serves as a “test” that nuclei with loss-of-fusion (LOF) mutations are likely to fail. This constant, local screening reduces the accumulation of selfish or deleterious nuclear lineages and maintains genotypic fidelity, despite ongoing mutation and nucleus-level selection. From an intelligence perspective, clamp connections implement a highly efficient search and pruning process: they continuously remove bad trajectories with minimal energetic and material cost, dramatically reducing the amount of trial-and-error required to maintain a functional genomic configuration. This represents a cellular-scale intelligence mechanism for sustaining a preferred genomic state under persistent evolutionary pressure (Aanen et al., 2023).

2. Dynamic Hyphal Differentiation

The differentiated hyphal system of higher basidiomycetes (generative, skeletal, and binding hyphae) embodies a continuously adaptive architectural strategy that aligns closely with the definition of intelligence as efficient search in a structured problem space. As described by David Moore, generative hyphae function as the metabolically active, exploratory, and highly plastic elements of the colony: they branch frequently, form septa, fuse (anastomose), and respond rapidly to local cues (Moore, n.d.). These hyphae expand into newly encountered voids, track gradients, and repair or reorganize the network when perturbed by insects, animals, or microbial competitors. Skeletal and binding hyphae, by contrast, are mechanically specialized; skeletal hyphae forming long, thick-walled, low-branching reinforcements, and binding hyphae forming intricately interwoven tissues that increase toughness and cohesion. Moore emphasizes that hyphal types emerge relationally, as phenotype expressions tuned to context. A generative hypha may thicken and transition toward a skeletal role under mechanical stress, while a rhizomorph composed of all three types may form in response to the need for long-distance exploration. These dynamic physical decisions can occur simultaneously across the colony, producing a mosaic of architectural decisions within a single contiguous mycelium driven by local needs for search efficiency.

This plasticity represents a sophisticated modulation of operators (O) and constraints (C) within the fungal problem space. Generative hyphae act as high-resolution search operators optimized for local uncertainty, while skeletal and binding hyphae lower the energetic cost of maintaining established routes by increasing mechanical stability and transport efficiency. And these decisions are made based on continuous energetic cost-benefit analysis; where and when is it advantageous to invest in physical durability vs. metabolic efficiency? Rhizomorph formation, which Moore describes as a coordinated developmental response integrating multiple hyphal types into a composite organ, further increases search efficiency by enabling rapid, directed exploration without the full metabolic cost of diffuse branching. Each of these architectural responses prunes vast regions of the combinatorial landscape: the colony minimizes futile search where certainty is high (focused subapical branching and local resource exploitation) and concentrates adaptive plasticity where uncertainty is greatest (dynamic and efficient foraging behavior). This dynamic partitioning of labor—exploration at the margins, consolidation in the core, reinforcement under stress, and modular organogenesis when long-distance foraging becomes advantageous—constitutes precisely the kind of efficiency gain over random growth that defines intelligence in this framework. In the terms of this thought experiment, mycelium would increase K through distributed, concurrent phenotypic adjustments, allowing the colony to reshape its problem space in real time and navigate it with remarkable economy.

3. Cell Walls

Fungal cell walls are continuously remodeled composites of β-glucans, chitin, mannoproteins, and accessory polymers whose proportions shift in response to nutrient availability, osmotic gradients, mechanical load, and local pressures (Moore et al., 2020; Haneef et al., 2017). This remodeling is proactive rather than purely reactive: hyphae adjust wall elasticity, porosity, and branching potential before catastrophic stress occurs, using integrated biochemical and biophysical signals to anticipate the demands of their immediate microenvironment. Hydrophobins extend this adaptive repertoire by allowing the fungus to dynamically tune surface wettability; coating walls with proteins that modulate gas exchange, repel water, promote aerial growth, or facilitate substrate attachment. Together, these mechanisms operate as an anticipatory system that modulates constraints (C) and operators (O) within the problem space: the wall becomes a medium through which the colony reshapes its own mechanical landscape, enabling efficient foraging, stable growth, and resilience against perturbation. In Levin’s framework, this predictive adjustment of material properties represents an expression of intelligence as search efficiency, reducing the expected cost of maintaining viable states relative to the high energy losses associated with passive, damage-driven repair. By continuously optimizing wall composition and surface behavior, the fungus prunes vast regions of potential failure and ensures that growth remains aligned with the preferred states defined by evaluation (E).

4. Directional Memory

The cord-forming basidiomycete Phanerochaete velutina provides one of the clearest demonstrations that fungal mycelia retain internal state information about prior resource encounters, and subsequently use that history to guide future decisions. Fukasawa, Savoury, and Boddy (2020) showed that when P.velutina networks colonized a new wood bait resource and were then transferred to fresh soil, the regrowing mycelium frequently biased its growth toward the side that had previously faced the bait, even though the physical connection to that resource had been removed. This directional bias indicates a form of resource memory: the colony stores information about past resource position within its physiological or architectural state and deploys that information when navigating new terrain. The study further revealed resource valuation: when the bait block was large enough the mycelium abandoned its original inoculum; when small, it remained. In the language of the Chis-Ciure–Levin framework, this can be interpreted as an increase in search efficiency (K) over space and time based on previous experience. Rather than reopening the full combinatorial space of possible growth directions, the colony prunes that space using historical information, biasing exploration toward previously rewarded regions and thereby lowering the energetic and material cost of search relative to a random foraging strategy. Not neuronal but instead encoded in the topology, internal tensioning, or physiological state of the network; what Boddy describes as a structurally instantiated record of past interactions.

5. The Spitzenkörper 

The Spitzenkörper (SPK) is a highly organized, mobile vesicle supply center positioned just behind the apical membrane and is one of the clearest examples of localized decision-making in fungal cells. As decades of work have shown (Steinberg, 2007; Virag & Harris, 2006; Riquelme & Sánchez-León, 2014; Zheng et al., 2020), the SPK coordinates vesicle delivery, polarity machinery, and apical growth dynamics, making it central to how hyphae steer and extend through changing local conditions. It acts as a vectorial controller; exocytic vesicles within the SPK predict the next increment of wall expansion with striking accuracy, effectively computing the local tangent of the growth trajectory. When the hyphal tip encounters a mechanical obstacle, changes in surface topology, gradients of nutrients, or electrical fields, the SPK shifts position accordingly producing immediate steering responses (thigmotropism, chemotropism, and galvanotropism). Rather than reacting reflexively, the SPK continuously balances competing constraints: minimizing curvature to maintain transport efficiency, maximizing contact with informative gradients, and preserving tip integrity. The SPK modulates operators (O) by redirecting vesicle delivery, adjusts constraints (C) through localized wall synthesis, and updates its evaluation (E) based on incoming signals that redefine the preferred direction of travel. Through this dynamic recomputation, the SPK prunes vast regions of geometrically possible trajectories, selecting only those that preserve goal attainment under uncertainty using relevant incoming information. As a result, the hypha navigates complex three-dimensional substrates with an efficiency far beyond random extension.

6. Bi-Directional Flow and Signaling

Fungal networks are not simple source-to-sink plumbing systems but dynamic transport architectures capable of bidirectional signal and nutrient flow, a discovery made explicit in the study by Schmieder and colleagues (2019). Using microfluidic devices and live-cell imaging, they showed that hyphal cords in Neurospora crassa can simultaneously conduct nutrients and signals in opposing directions, with flow reversal triggered by local changes in resource availability, hydraulic gradients, and colony demand. This behavior reveals a system in which no single hyphal segment permanently functions as a source or sink; instead, each node participates in a distributed arbitration process, allocating resources where they are most needed and routing information across the network through coordinated pressure-driven mass flow. Combined with earlier modeling from Heaton et al. (2010) showing that growth-induced mass flows can propagate over long distances in large mycelia, these observations underscore that fungal networks continually update internal evaluations and rebalance transport pathways to maintain global function. Reversible transport and context-sensitive redistribution demonstrates a form of decentralized cognition: subunits modulate operators (O) and constraints (C) in real time to maintain preferred states under changing conditions. In this interpretive frame, bidirectional flow would increase search efficiency K by reducing uncertainty, coordinating distant regions of the colony, and preventing costly overinvestment in any single trajectory. Rather than relying on fixed pathways, the mycelium dynamically reconfigures transport topology, embodying a form of collective computation in which information and resource flows converge on system-level optimization.

7. Growth Scale: Longevity & Stability

The Armillaria gallica genet known as C1 in Michigan is a compelling example of large-scale biological stability expressed through space and time (Anderson et al., 2018). In this case, a single individual at least 2,500 years old extends across roughly 75 hectares, while exhibiting exceptionally low somatic mutation accumulation. Across multiple isolates sequenced from around the genet, only a modest number of variants were detected, most of them singletons, indicating that new mutations rarely spread or fix within the organism. From these data an extraordinary growth-to-division ratio is suggested: the genet appears to have expanded nearly a kilometer with only a small number of cell divisions. They propose several mechanisms that could underlie this genomic fidelity, including highly efficient DNA repair, asymmetric segregation of older template strands, and the possibility that long-lived rhizomorph tips behave as meristem-like structures with infrequent divisions.

This combination of spatial vastness and mutational restraint reflects a system that maintains preferred states with remarkably little dissipation. Over millennia, the colony has preserved its genomic identity while adapting its architecture to shifting forest conditions, suggesting a long-term strategy that minimizes unnecessary search, avoids destructive drift, and channels growth along stable trajectories. In this sense, Armillaria C1 exemplifies intelligence as temporal search efficiency: an ability to preserve functional coherence across deep time and space through mechanisms that stabilize Operators (O), protect Constraints (C), and maintain system-level goals at minimal cumulative cost.

Mycelium as Distributed Intelligence Architecture (A Thought Experiment)

Defining the Agent

In this thought experiment lets approach the agent is the mycelium as a whole: a physically contiguous, dynamically reorganizing fungal network composed of branching, fusing, differentiating hyphae coordinated across centimeters to kilometers. The agent is the entire distributed architecture that integrates these subsystems into coherent action.

The mycelium qualifies as an agent in the Levin–Chis-Ciure sense because:

• it persistently maintains itself in a structured state despite perturbations;

• it performs goal-seeking behavior (resource acquisition, territorial expansion, competitive positioning or partitioning) without a centralized controller;

• it modulates its operators based on environmental and internal signals;

• it adapts to shifting constraints through structural and physiological reconfiguration;

• it carries and reuses information (directional memory, resource valuation, genomic stabilization);

• it generates large-scale coherent outcomes from local rules.

The “agent” here is thus the integrated colony, regardless of spatial extent or genotype, functioning as a coherent decision-making system occupying a physical substrate.

Defining the Problem Space

To understand how a mycelium might be evaluated within the Levin–Chis-Ciure framework, we must first define the problem space in which it acts. The mycelium does not simply inhabit its environment; it occupies a high-dimensional landscape of possibilities shaped by its own architecture, physiology, and history. Its problem space, P = ⟨S, O, C, E, H⟩, captures the full set of conditions under which the colony must navigate uncertainty and maintain adaptive trajectories.

The state space (S) encompasses all the configurations a fungal network can assume: the shifting topology of its hyphal network, the locations and densities of its branches and anastomoses, osmotic gradients and nutrient flow, the composition and elasticity of its cell walls, the physiological status of its nuclei, and even its embodied memory traces and directional biases. At any moment, the colony’s state is a spatially distributed pattern of tensions and possibilities; its current form and its accumulated history.

The operators (O) available to the mycelium are the transformations through which it reshapes this state space. These include extending or vacuolizing hyphae, branching or fusing them, thickening or reinforcing cords, changing Spitzenkörper position, altering wall composition, rerouting internal flows, forming rhizomorphs, or reallocating metabolic investment. Each operator carries a cost and opens or closes pathways in the state space; together they constitute the organism’s repertoire for navigating complexity.

The constraints (C) the fungus faces are equally multifaceted. Some are imposed externally by the substrate’s geometry, its pores and mechanical resistances, the distribution of nutrients, or the presence of competitors and grazers. Others arise internally from energetic limitations, hydraulic thresholds, genomic vulnerabilities, or structural commitments already made. But, a key feature of fungal intelligence is that constraints are not fixed but malleable: hyphae soften or stiffen their walls, redistribute flow, or reorganize their network to renegotiate the limits within which they operate.

Within this landscape, the evaluation function (E) identifies the states the fungus tends to preserve or seek out: states that sustain integrity, maintain transport continuity, secure nutrients, reinforce territories, stabilize the genome, or reduce dissipation. These preferred states guide the organism’s movement through its state space and anchor the colony’s long-term coherence.

The horizon (H), the temporal window over which it integrates information and adjusts behavior, spans an extraordinary range. It responds to mechanical cues and chemical gradients over milliseconds to seconds, reorganizes flow and branching patterns over hours, remodels its architecture over days, consolidates territories over seasons, and in some species maintains genomic stability and navigates vast expanses over centuries or millennia. Few biological systems integrate such a wide range of temporal scales into coherent, adaptive behavior.

Thought Experiment 1: Spatial Navigation

To estimate how a mycelium might score on search efficiency K in the spatial-nutritional domain, imagine a colony placed at the center of a heterogeneous substrate where pore size, mechanical resistance, and nutrient quality vary unpredictably. A maximal-entropy random walker would fan outward in a sphere, testing every direction with equal probability, expending enormous energy probing inaccessible or unproductive regions, and relying on chemical gradients only after they accumulate.

A mycelium instead deploys a tiered spatial strategy.

At the colony margin, leading hyphae adopt high-efficiency exploratory morphologies: the hyphal growth unit lengthens, cell diameters widen, branching frequency drops, and negative autotropism keeps tips apart to maximize the volume of new substrate interrogated per unit biomass. This produces sparse, fast, low-cost exploratory growth that preserves translocation efficiency. When local resources remain absent, these marginal hyphae may transition into rhizomorphic growth, a strategy optimized for long-distance foraging in which the colony invests in a small number of reinforced, high-efficiency conduits that traverse unproductive terrain while burning minimal internal reserves.

Just behind this exploratory front lies a zone where the colony begins subapical commitment. Here, branching density increases only where resources are newly encountered. Rather than distributing branching arbitrarily, the fungus commits biomass to exactly those regions where discovery has already been validated, converting weak signals of opportunity into tissue-level investment. This local commitment exploits immediate resources and reduces future uncertainty: the colony now “knows” where productive directions lie.

Further back, in the interior of the colony, the network reorganizes itself around the outcomes of earlier exploration. Anastomosis, densification, and cord formation optimize translocation pathways; redundant routes collapse into efficient networks; mechanical stresses trigger reinforcement through wall thickening, hyphal differentiation, or cellular vacuolization; and animal foraging pressure induces local branching or fusion to preserve flow through formation of redundant pathways. This part of the colony becomes an adaptive logistics network, tuned to sustain the exploratory margin while stabilizing the territory already acquired.

Every stage of this spatial hierarchy compresses search relative to randomness:

• Sparse margin exploration prunes vast swaths of dead space with minimal cost.

• Rhizomorph deployment under resource absence extends exploration range without proportional metabolic burn.

• Subapical branching only after resource confirmation collapses the combinatorial explosion of possible directions into a manageable set of validated trajectories.

• Interior densification and route optimization reduce long-term transport cost and uncertainty.

• Directional memory encoded in cords biases future exploration toward historically rewarding regions.

• Mechanical sensing and SPK steering identify futility (blind alleys, blocked pores) before resource expenditure.

A random walker must sample countless possibilities to locate resources, while mycelium uses a layered strategy that systematically prevents wasted work. Each operator reduces expected search cost. Each zone of the colony plays a distinct computational role.

In this spatial-navigation thought experiment, mycelium would achieve high K because it is incapable of searching the world blindly.

Thought Experiment 2: Spatio-Temporal Persistence

To evaluate mycelial intelligence in the temporal domain, we can compare two very different strategies by which fungi preserve identity, coherence, and functional continuity over time. Temporal persistence is itself a search problem: a system must maintain preferred states across varying lengths of time while contending with mutational pressure, environmental fluctuation, resource depletion, and structural degradation. A maximal-entropy process would drift genetically, collapse structurally, and lose informational integrity. Fungi, however, have evolved multiple strategies for dramatically reducing this temporal search burden.

The Armillaria gallica C1 genet provides an extreme example of long-term temporal intelligence. Over an estimated 2,500 years and across more than 70 hectares, the colony has accumulated remarkably few somatic mutations. This suggests the fungus has discovered a solution to the temporal search problem that minimizes the number of genomic trajectories explored. Mechanisms such as infrequent apical cell divisions, high-fidelity DNA repair, asymmetric DNA strand segregation, and nuclear-quality policing via clamp connections may collaborate to maintain a narrow corridor of viable genomic states. As a result, across centuries the network reorganizes spatially yet its identity persists with coherence. This represents a system that prunes the combinatorial space of genomic drift over extreme temporal spans. In the terms of this temporal thought experiment, its K would be high: Armillaria reaches preferred future states (continued growth, ecological dominance, genomic stability) while sampling only a tiny fraction of the mutational landscape. This is intelligence expressed as deep-time constraint management, a capacity to remain coherent with a long horizon.

A second, fundamentally different temporal persistence strategy occurs in many white-rot polypores, with Ganoderma species as an example, through intercalary blastoconidiation and chlamydospore formation (Loyd et al., 2019). Here, persistence is not achieved by minimizing mutation, but by reformatting the mycelium’s body and colonial footprint into a durable, low-metabolic, resource-preserving architecture.

When conditions decline (desiccation, nutrient depletion, host decline, temperature stress), the fungus triggers an intercalary transformation in which individual generative hyphal compartments swell, thicken their walls, accumulate storage compounds, and differentiate into chlamydospores. Instead of maintaining an active, metabolically expensive network, the colony converts its existing spatial lattice into a vast constellation of dormant but geographically faithful propagules.

This transformation is a temporal bookmarking mechanism. If this process occurs broadly across an occupied substrate, the resulting chlamydospore distribution can be read, in this thought experiment, as a partial preservation of the spatial solution the fungus previously discovered. When favorable conditions return, germination occurs across this entire footprint, allowing the colony to instantly recapture the same resource volume without having to re-explore or re-map the substrate.

From a search-efficiency perspective, this strategy compresses temporal search by:

• storing spatially distributed memory of successful occupancy,

• avoiding the cost of rediscovery after disturbance,

• minimizing metabolic expenditure during inhospitable periods, and

• ensuring immediate reactivation of the full resource domain.

Where Armillaria maintains continuity through genomic stability, Ganoderma maintains continuity through structural serialization: they turn their present form into a stored representation of their past successes. This is a different kind of temporal intelligence oriented on medium-term resilience and rapid re-entry into preferred states after disruption.

These two cases reveal that the depth of fungal physicality can solve the temporal search problem in multiple, orthogonally intelligent ways:

• Armillaria compresses temporal search by minimizing mutational drift and maintaining developmental operators across millennia.

• Chlamydospore-forming polypores compress temporal search by encoding spatial success into durable cytological packets, preserving the colony’s solution space as a recoverable template.

Both strategies drastically reduce the need to re-explore, re-optimize, or re-stabilize their environment. Both can be read as demonstrating high K in the temporal domain; one through deep-time genomic discipline, the other through medium-term structural serialization. Taken together, they show that intelligence is a direct extension of their foundational imperative: to occupy space efficiently and persistently. Whether by freezing a spatial solution into millions of dormant cells or by preserving a unified genomic lineage across millennia, fungi maintain coherence by continually compressing the search required to remain in place. Their mastery of time, like their mastery of substrate, emerges from the same principle: occupancy demands a layered, nested capacity to prune possibility. Mycelium achieves this by turning the physical act of occupancy into a multi-scale computation across space and time.

This recognition orients mycelium engineering. We are designing problem spaces whose optimal solutions, arrived at through fungal intelligence, satisfy both fungal and human objectives. When we specify substrate composition, inoculation density, or environmental conditions, we are defining P; the state space, operators, constraints, evaluation criteria, and horizon within which the mycelium must navigate. Our design succeeds when the fungus's search for preferred states (efficient occupancy, resource security, structural integrity) naturally converges with our material requirements (mechanical properties, growth kinetics, biological efficiency). In this framework, the fungus can be seen as an intelligent collaborator whose search efficiency we strategically channel. In this framing, high K for the organism becomes high performance for the material. Engineering becomes a practice of mutual optimization, of finding the win-win. The value of this thought experiment is that it offers a disciplined way to read fungal growth as efficient, embodied problem solving.

References

Aanen, D. K., van ’t Padje, A., & Auxier, B. (2023). Longevity of fungal mycelia and nuclear quality checks: A new hypothesis for the role of clamp connections in dikaryons. Microbiology and Molecular Biology Reviews, 87(3), Article e00022-21. https://doi.org/10.1128/mmbr.00022-21

Anderson, J. B., Bruhn, J. N., Kasimer, D., Wang, H., Rodrigue, N., & Smith, M. L. (2018). Clonal evolution and genome stability in a 2500-year-old fungal individual. Proceedings of the Royal Society B: Biological Sciences, 285(1893), Article 20182233. https://doi.org/10.1098/rspb.2018.2233

Chis-Ciure, R., & Levin, M. (2025). Cognition all the way down 2.0: Neuroscience beyond neurons in the diverse intelligence era. Synthese, 206, Article 257. https://doi.org/10.1007/s11229-025-05319-6

Fukasawa, Y., Savoury, M., & Boddy, L. (2020). Ecological memory and relocation decisions in fungal mycelial networks: Responses to quantity and location of new resources. The ISME Journal, 14(2), 380–388. https://doi.org/10.1038/s41396-019-0536-3

Haneef, M., Ceseracciu, L., Canale, C., Bayer, I. S., Heredia-Guerrero, J. A., & Athanassiou, A. (2017). Advanced materials from fungal mycelium: Fabrication and tuning of physical properties. Scientific Reports, 7, Article 41292. https://doi.org/10.1038/srep41292

Heaton, L. L. M., López, E., Maini, P. K., Fricker, M. D., & Jones, N. S. (2010). Growth-induced mass flows in fungal networks. Proceedings of the Royal Society B: Biological Sciences, 277(1698), 3265–3274. https://doi.org/10.1098/rspb.2010.0735

Loyd, A. L., Linder, E. R., Smith, M. E., Blanchette, R. A., & Smith, J. A. (2019). Cultural characterization and chlamydospore function of the Ganodermataceae present in the eastern United States. Mycologia, 111(1), 1–12. https://doi.org/10.1080/00275514.2018.1543509

Moore, D. (n.d.). Hyphal differentiation. In 21st century guidebook to fungi (Platinum ed.). Retrieved June 2, 2026, from https://www.davidmoore.org.uk/21st_century_guidebook_to_fungi_platinum/Assets/Hyphal_differentiation_Resources.html

Moore, D., Robson, G. D., & Trinci, A. P. J. (2020). 21st century guidebook to fungi (2nd ed.). Cambridge University Press. https://doi.org/10.1017/9781108776387

Riquelme, M., & Sánchez-León, E. (2014). The Spitzenkörper: A choreographer of fungal growth and morphogenesis. Current Opinion in Microbiology, 20, 27–33. https://doi.org/10.1016/j.mib.2014.04.003

Schmieder, S. S., Stanley, C. E., Rzepiela, A., van Swaay, D., Sabotič, J., Nørrelykke, S. F., deMello, A. J., Aebi, M., & Künzler, M. (2019). Bidirectional propagation of signals and nutrients in fungal networks via specialized hyphae. Current Biology, 29(2), 217–228.e4. https://doi.org/10.1016/j.cub.2018.11.058

Steinberg, G. (2007). Hyphal growth: A tale of motors, lipids, and the Spitzenkörper. Eukaryotic Cell, 6(3), 351–360. https://doi.org/10.1128/EC.00381-06

Virag, A., & Harris, S. D. (2006). The Spitzenkörper: A molecular perspective. Mycological Research, 110(1), 4–13. https://doi.org/10.1016/j.mycres.2005.09.005

Zheng, P., Nguyen, T. A., Wong, J. Y., Lee, M., Nguyen, T.-A., Fan, J.-S., Yang, D., & Jedd, G. (2020). Spitzenkörper assembly mechanisms reveal conserved features of fungal and metazoan polarity scaffolds. Nature Communications, 11, Article 2830. https://doi.org/10.1038/s41467-020-16712-9