Stability as an Emergent Property in MycoManufacturing
Aligning human psychology with the dispassion of fungal responsiveness can create a disorienting gap between what we want to be true and what the organism is actually doing. Scaling whole-thallus mycelium cultivation can feel less like a clean technical progression and more like trying to steer the historically rooted, historical and self-reinforcing nature of fungal behavior through the turbulence of human systems built on capital, ego, and intuition. Those systems do not align with the ideal of rigorous, linear scaling. Not even close.
I’ve said for years (and anyone who has worked with me has heard this mantra many times) that “if you tell the fungus to do the same thing every time, it will do the same thing every time.” But contrary to this mantra, when instability shows up in a myceliation or mycomanufacturing system, human psychology almost reflexively labels it as biological unpredictability. But more often than not, the first interpretive stance should be that the fungus is responding coherently to the conditions it has been given. That does not rule out strain drift, physiological history, stochastic biological variation, or genetic and epigenetic instability, but it does force us to examine the system designed around the organism before assigning blame to biological unpredictability. It’s the system designed around the organism that isn’t drafted right. In which case, instability becomes a mirror for our own inconsistency, and unfortunately it can be hard to look at yourself in the mirror at times.
In practice, this space can become a technical house of mirrors. The true shape of the problem warps; reality becomes obscured. What seems rational or irrational, even solvable, can drift further from what is ultimately true because fungal behavior is dispassionate about our human plights along the path of “making a thing that does a thing such that I may survive.” It doesn’t care about urgency or the story we’ve told ourselves about how scaling works. It simply responds to the problem space our process design has posed according to its own logic.
This mismatch between how mycelium behaves and how humans behave is one of the central challenges in scaling in mycomanufacturing.
The concepts developed here should be understood as practitioner-facing interpretive constructs, not formal taxonomies of organizational behavior, fungal development, or process control. I am using them to organize a recurring pattern encountered in real-world mycelium R&D: biological responsiveness, process instability, technical debt, human psychology, and scale pressure becoming entangled. Their value is operational. They help name the ways stability is built or eroded across nested biological and human systems.
Mycelial Organization as an Emergent Property
To talk about stability in mycomanufacturing we have to be clear-eyed about what mycelium is actually doing. It is not a collection of traits or fixed morphological parts, but a hierarchy of behaviors expressed across scales, each nested inside the next (Moore et al., 2011). Growth begins hyper-locally: a hyphal tip responding to gradients in nutrients, moisture, stiffness, neighbors. But what emerges from those small, local decisions are network-level behaviors that look coordinated (Fricker et al., 2009). None of this comes from a central plan. It emerges because enough local rules, applied consistently, propagate upward.
In practice fungal structure is a transient expression of the growth context. A hypha differentiates, thickens, branches, fuses, or regresses depending on what it encounters. The tissue states we see are programs that rewrite themselves constantly. And in a very real sense, this scaling from single hyphal compartment to complete thallus is how mycelium organizes everything. The fungus scales its behaviors upward: from tip-level growth physics and branching, to inter-hyphal interactions and differentiation, to macro-morphological responses such as organ formation, mass flow decisions, damage repair, and colony-scale coordination. Each level inherits the logic of the one below it, and constrains the one above. The total phenotype is, then, the emergent outcome of this nested process telegraphing through scale.
This matters because the physicality of mycelium is always emergent. The form we get is the form the fungus determines is appropriate given the conditions we built around it, where our process design poses a problem space to be solved by fungal form. Local decisions shape meso-structure, meso-structure shapes the global phenotype, and the global phenotype feeds back into local decisions. In this context it is easy, and understandable, that fungal systems can appear inconsistent when viewed through human stress and wants. What we call variability is often the fungus behaving with perfect consistency relative to its own logic. It is our systems, our designs, our untracked changes, our misunderstanding of its nested behavior that drift, not the organism.
Process Stability as an Emergent Property
If fungal structure is an emergent property of nested behaviors, then, in the operational framing I am developing here, stability in mycomanufacturing can be understood as an emergent property of nested scaling decisions. At the bench, scale-up, and production, each inherits the logic of the stage before it and bounds the stage after. Stability is cumulative, the downstream effect of upstream decisions, upstream habits, and upstream philosophies.
It’s a mistake to believe that stability is something you ‘arrive at,’ a milestone achieved through linear progression: run the screens, then scale the system, then lock the parameters, then validate the process, thus stability. That linear narrative is itself a model, and like all models, useful only when its limits are recognized (Sterman, 2002). This framing is too narrow for a biological system whose defining feature is dimensional and relational responsiveness. The fungus continuously interprets, evaluates, and acts within the full dimensionality of its context. What we call stability is the momentary alignment between that responsiveness and the structures we’ve built around it. In this sense, stability is a running measure of compatibility, our systems alignment with the recursive logic the organism already uses to maintain coherence across its own scales.
Stability is a discipline we have to keep exercising, because the moment we stop, it begins to erode. At the bench, it shows up as disciplined experimental design that helps reveal the range of fungal responsiveness rather than collapse it prematurely. In the scale train, stability shows up in how we interpret eccentricity and reconcile scale-dependent effects as information about the organism’s relationship to volume and physical environment. In manufacturing, stability shows up in how we monitor performance drift and respond to it. The tools themselves are secondary; what matters is what they enforce: an attentiveness to dimensionality that keeps us from collapsing the system into something simpler than it really is.
At the bench, this practice begins with understanding the physical response space of the strain independent of scale; the global phenotype that defines how the fungus behaves across its potential range. Structured experimental design helps reveal input–response relationships rather than merely confirm assumptions. Dimensionality-reduction tools clarify which variables matter and how they interact. Perturbation studies expose limits and transitions instead of chasing premature “optimal points,” while longitudinal tracking across passages or batches makes slow phenotypic drift visible before it becomes a structural problem.
As we move into the scale train, stability depends on treating scale itself as an independent parameter, as an ecological condition the organism responds to, and not a simple geometric expansion of the bench. Each step reveals its own emergent eccentricities: temperature gradients, airflow patterns, moisture stratification, and other scale-dependent shifts that shape biological outcomes. Scale-step experimentation allows these behaviors to surface rather than be smoothed over. Comparative analyses clarify which responses remain stable, which become fragile, and which newly emerge. Incorporating scale directly into modeling ensures that it is represented as its own dimension within a model-centric learning philosophy, rather than an afterthought added once problems arise. This aligns with bioprocess scale-up literature noting that biological phenomena within bioreactors are difficult to scale, that rule-of-thumb approaches have limits, and that digital or model-based methods can help compare and predict process indicators across scales (Alavijeh et al., 2022).
At manufacturing scale, stability depends on understanding both the effectors within our field of view and the uncertainty outside it. Statistical process control helps define what “normal” looks like and detect deviations early, while repeated, intentionally consistent runs establish operational baselines. Feature-importance tools clarify which controllable variables exert the greatest influence. Residual analyses help surface unknown variation rather than bury it, and error-budgeting exercises allocate uncertainty across stages, revealing where epistemic blind spots reside. Drift monitoring, forecasting tools, and stability mapping across time and passage then turn temporal behavior into actionable knowledge rather than retrospective explanation.
Despite every method, every analytic framework, every attempt to build systems that mirror and value dimensionality, stability still depends on how we behave as the human component of the system. The tools only work if we use them, the structure only holds if we resist the urge to bypass it, and the philosophy only coheres if we remain disciplined enough to honor it. Because the pressures of scaling, the noise of urgency, and the distortions of ego and intuition all exert forces that run directly counter to the recursive logic of the fungus. If we want stability to emerge downstream, we have to confront the upstream tendencies in ourselves that break it before it has a chance to form.
The ‘Move Slow to Move Fast’ Paradox
Real-world pressure can produce the illusion that success comes from speed, and speed comes from skipping fundamentals.
Scaling happens inside a pressure system; capital burn feels urgent, leadership expectations feel urgent, production schedules feel urgent. And under that pressure, when the organism refuses to operate on the same timeline we do, it becomes surprisingly easy for speed to start masquerading as competence. Organizational literature on speed cautions against assuming speed is inherently good, emphasizing that speed can produce pathologies and should be evaluated in relation to its broader social, organizational, and practical consequences (Baird, 2021). Mycelium does not care about these pressures, but we do, and the mismatch can distort our reasoning.
This is the environment in which a very seductive perspective becomes seemingly reasonable:
We can move faster by ignoring complexity.
It’s not entirely false. You can take on technical debt. You can skip steps, make fast decisions, and leverage that debt to meaningfully push progress. Technical debt, in its original formulation, was not inherently bad. Cunningham framed it as a debt metaphor in which ‘a little debt’ can speed development if it is paid back promptly through later consolidation or rewriting (Cunningham, 1992). Contemporary reviews treat technical debt management and prioritization as an active research and practice problem, especially because unmanaged debt can accumulate into maintainability and long-term viability concerns (Lenarduzzi et al., 2021; Terrón-Macias et al., 2025). There are moments when this is exactly the right move. In early discovery, in prototyping, in exploring the boundaries of a new strain’s behavior, in scaling rapidly to explicitly derisk the process and back propagate along the scale train; debt can be a catalyst rather than a liability.
But in the interpretive extension I am making here, unchecked technical debt can become epistemic debt: a slow erosion of our ability to understand what is actually happening. And epistemic debt becomes operational collapse, the erosion of our ability to control what is happening; entry into failure mode as an emergent property of technical debt. Shortcuts that once felt efficient begin to accumulate into blind spots and unexamined assumptions. What felt like acceleration reveals itself, in hindsight, to have been drift.
And the hardest part is that, in the moment, drift feels like progress. It feels like speed. It feels like competence.
Paradoxically, moving fast often produces the sensation of momentum while dismantling the foundations that operational and systemic stability depends on. Moving slowly feels frustrating and even counter-cultural, but often in retrospect is demonstrably faster than the alternative when you define speed in the most relevant terms: pace of learning, performance, and economic improvement.
But there is an additional risk when we consider the formative and multi-disciplinary nature of mycelium technology at scale: it is nearly impossible to enter a process development role in this context with all required competencies to address the full breadth of scaling challenges. So, while mired in the layers of urgency, the technical and psychological stress, there is an additional and uncomfortable question we have to ask ourselves: am I taking on technical debt because I simply lack the competency to make an alternative decision?
The Ego & Intuition Trap
I am constantly aware of the limits of my original training. I didn’t come into this field through a traditional scientific or engineering pathway, and in high-pressure moments, when the technical landscape expands faster than the boundaries of my competency, I feel that gap immediately. It’s an uncomfortable place to sit, and that discomfort is not unique to me; it is a universal condition of doing real process development inside a system defined by urgency, ambiguity, and the demand for answers. When the technical problem outpaces our skillset, or even just our confidence, ego and intuition rationally fills that gap. And in the context of scaling, that gap can become enormous.
I use the ego and intuition trap as a practitioner-facing diagnostic construct, not a formal psychological category. It names a recurring pattern in high-pressure mycelium R&D: when technical uncertainty exceeds available structure, intuition and ego can begin to substitute for disciplined inquiry.
So we arrive at a largely unspoken dynamic: ego-driven decision making is not always rooted in overconfidence or the myth of singular genius. Often it is rooted in the opposite; insecurity, a quiet fear that we may not have the tools required to resolve the complexity of the moment. Under stress, those insecurities can invert. They can transform into conviction and the seductive feeling that intuition is enough. That a good story and a confident answer is enough.
Scaling pressure tempts us into behaviors that look like leadership but function more like self-preservation. Intuition starts to masquerade as insight, limited data gets overinterpreted to confirm existing convictions, and the idea that a single “genius move” will unlock the system becomes strangely appealing. Conversations drift from open-ended analysis toward closed-room pontification, and the work of knowledge-making gives way to the comfort of narrative-making.
These behaviors are not failures of character; they are predictable human responses to environments that demand certainty where none exists. But mycelium is not a system that yields to intuition, and scaling is not a context where ego can compensate for missing understanding.
No individual, no matter how brilliant, experienced, or instinctively talented, can resolve the complexity of scale-dependent mycelial systems through intuition alone. At scale, complexity quickly compounds beyond intuition’s capacity. Ego and intuition left to their own devices are not merely unreliable, they are structurally incompatible with the organism’s high-dimensional logic. They collapse precisely where mycelium expands: in dimensionality, in nested responsiveness, in the generative tension between local behavior and global outcome.
And yet, under pressure, ego functions as a stabilizing force while intuition feels efficient. Certainty, even if false, feels like leadership. When the team is stressed the smallest dose of confident narrative can feel like relief. It is dangerously comforting to believe someone “just knows” the answer. It protects us from confronting that the problem requires more than any one person in the room currently holds.
This is the ego and intuition trap. Not a cartoonish arrogance, but a nuanced psychological pivot: from “I don’t know enough” to “I don’t need to know.” From “the system is complex” as a starting point to an answer.
Risk of failure is often not because people are careless or reckless, but because they are committed to success and afraid; afraid of their own limitations, and afraid that admitting those limitations will slow the momentum the organization believes it needs. And the organism, indifferent to our psychology, may still be behaving coherently within a history and context we have failed to understand, even as we drift further from the evidence needed to distinguish process instability from organismal change. The tragedy and opportunity is that this pattern is predictable, but only if we’re willing to see ego and intuition not as personal failings, but as structural vulnerabilities that emerge exactly when the system is under its greatest strain.
Scaling as a Human Amplifier
In the practitioner framing developed here, scaling acts as a human amplifier. It amplifies whatever is already present in the system: rigor or drift, humility or ego, skepticism or wishful thinking. If each layer leans even slightly toward convenience over discomfort, or toward story over structure, the effect magnifies as the process moves downstream.
Scaling creates its own psychological weather system; capital burn accelerates and data streams proliferate and simultaneously may become noisier. Failures start to look novel because the contexts in which the fungus is being asked to perform have multiplied. Each new passage step, or step-change in volume, adds another layer of complexity. This swirl of scale produces real noise and real ambiguity in which signals can become harder to parse. The demand for clarity intensifies at the exact moment when clarity is hardest to obtain.
And scaling does something else that is easy to underestimate: it inserts layers of humans at each step of the scale train. The bench scientist, the pilot engineer, the scale-up lead, the operations manager, the production supervisor; each occupies a different locality in the system. Each works with a unique view of shared information as well as unique information, and a different tolerance for ambiguity. These are orthogonal human systems, where each has the capacity to introduce its own distortions into how reality is perceived and explained.
Under pressure, those distortions compound. The shortcuts taken at the bench propagate into assumptions at pilot. The interpretations at pilot ossify into "truths" at intermediate scale. The stories told at intermediate scale become operating doctrine in production. By the time instability shows up in manufacturing, it is no longer traceable to a single decision or a single person; it is the emergent product of a long chain of small interpretive decisions, each made by someone trying to do the right thing.
Confirmation bias becomes structural when technical debt introduces blind spots, and ego and intuition move in to fill those blind spots with confident explanations. Confirmation bias then does the work of locking those explanations in place. Data that supports the emerging story is elevated, and conflicting data is reframed as noise, one-off error, or otherwise "not representative." The more the system is stressed, the more valuable a simple, coherent narrative feels.
And this is the core danger: the behaviors that feel stabilizing from the inside are often the same behaviors that accelerate epistemic collapse. People who insist on slowing down to interrogate anomalies can feel like obstacles. People who keep asking for more data or more structure can feel misaligned with "the urgency of the moment." The system starts to reward those who can keep the story intact and marginalize those who keep noticing the cracks. All the while, the fungus continues responding to gradients, history, and its own biological state, whether or not our narrative can account for it. Sometimes that response reflects process inconsistency; sometimes it may reflect organismal history, drift, or instability. The danger is that our story becomes simpler just as the system demands more careful discrimination.
But the fact that scaling amplifies human failure modes also implies an inverse principle: it can amplify human discipline.
The same recursive structure that propagates epistemic wandering can propagate rigor, if each layer of the scale train treats stability as a practice rather than an entitlement. The toolkit described earlier—structured experimentation, dimensionality reduction, longitudinal tracking, drift analysis, process control, model-centric learning—is about managing our own tendency to fill gaps. These tools force us to confront what is actually there instead of what we hope is there.
At this point I feel I am in danger of minimizing the value of intuition. If all this is to drive to a core argument it is not to diminish the value of human intuition in the scaling process, but rather to amplify it. Empowered within a framework of technical rigor, intuition becomes an asset rather than a risk. It stops functioning as a replacement for missing understanding and instead becomes a power for orchestrating the full force of the technical scaling toolkit. The analytic structure provides the guardrails within which intuition can operate creatively to drive to reality and stabilize operation. Rigor doesn't extinguish intuitive insight; it grounds it, aligns it with reality, and channels it toward high-impact questions. In this configuration, intuition is a fast and internalized synthesis of knowledge. The toolkit empowers intuition by ensuring that the feelings we follow are anchored in something more than stress and urgency.
If we choose, each scale step can function as a kind of epistemic checkpoint. At the bench, the work is to map the breadth of responsiveness honestly, not to prematurely collapse it into a single convenient operating window. At pilot, the work is to treat scale as an independent variable and to surface, not smooth over, the eccentricities that emerge. At intermediate and production scale, the work is to treat drift and deviation as information, not personal affronts or threats to the storyline. At every stage, the question shifts from "How do we keep the story intact?" to "What new structure is the system revealing, and what does that force us to update in our understanding?"
In this way, disciplined practice at each layer begins to do for us something analogous to the hypothesized role of clamp connections in dikaryotic fungi. Aanen et al. (2023) propose that clamp connections may function as part of a nuclear quality-checking system that helps maintain long-term mycelial stability. We can borrow that principle metaphorically: before a conclusion, a practice, or a narrative is passed from one scale layer to the next, we can ask whether it has been stress-tested against the dimensionality of the system, checked against contradictory data, and tested against the full complexity of what the organism is doing.
When that happens consistently, scaling still acts as an amplifier, but what it amplifies is coherence. The cumulative effect of small, disciplined choices at each stage is a process that becomes more stable as it grows, not less. The story becomes simpler only in the sense that it becomes more accurate, more tightly coupled to the organism's logic, more resistant to wishful thinking. Stability, in this configuration, is the emergent property of a nested set of human practices that mirror the nested behaviors of the fungus itself.
Scaling will always amplify something. The question is whether it amplifies our unexamined anxieties, shortcuts, and stories, or our willingness to align our decisions with the reality of a system that does not need us to believe in it to behave the way it behaves.
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
Alavijeh, M. K., Baker, I., Lee, Y. Y., & Gras, S. L. (2022). Digitally enabled approaches for the scale up of mammalian cell bioreactors. Digital Chemical Engineering, 4, Article 100040. https://doi.org/10.1016/j.dche.2022.100040
Baird, C. H. F. (2021). Problematizing speed in and around organizations: Struggles over the temporal commons in the British artificial intelligence field [Doctoral dissertation, University of Edinburgh]. Edinburgh Research Archive. https://doi.org/10.7488/era/2080
Cunningham, W. (1992, March 26). The WyCash portfolio management system. OOPSLA ’92 Experience Report. https://c2.com/doc/oopsla92.html
Fricker, M. D., Boddy, L., Nakagaki, T., & Bebber, D. P. (2009). Adaptive biological networks. In T. Gross & H. Sayama (Eds.), Adaptive networks: Theory, models and applications (pp. 51–70). Springer. https://doi.org/10.1007/978-3-642-01284-6_4
Lenarduzzi, V., Besker, T., Taibi, D., Martini, A., & Arcelli Fontana, F. (2021). A systematic literature review on technical debt prioritization: Strategies, processes, factors, and tools. Journal of Systems and Software, 171, Article 110827. https://doi.org/10.1016/j.jss.2020.110827
Moore, D., Robson, G. D., & Trinci, A. P. J. (2011). 21st century guidebook to fungi. Cambridge University Press. https://doi.org/10.1017/CBO9780511977022
Sterman, J. D. (2002). All models are wrong: Reflections on becoming a systems scientist. System Dynamics Review, 18(4), 501–531. https://doi.org/10.1002/sdr.261
Terrón-Macias, V., Mejia, J., & Terrón-Hernández, M. (2025). A systematic literature review on technical debt in software development: Types, tools, and concerns. International Journal of Combinatorial Optimization Problems and Informatics, 16(4), 288–302. https://doi.org/10.61467/2007.1558.2025.v16i4.1150