Why the most dangerous business advice of the AI age is “pick a niche and go deep”:

Part 1: The Climate Change of Business
If you’ve ever attended a business conference, read a startup book, or spent more than eleven minutes on LinkedIn, you’ve encountered the adage: “Pick a niche. Go deep. Become the go-to expert. Plant your flag. Grow tall. Be the oak tree.“
The metaphor is always arboreal. You’re supposed to put down deep roots in a single vertical, say, CRM implementation for mid-market dental chains, and grow upward through that specific canopy until you tower above the competition. The roots represent your expertise. The trunk is your brand. The canopy is your market dominance. It’s a beautiful metaphor. It’s also, increasingly, wrong.
Not wrong in the way that business advice is usually wrong (which is to say: vaguely inapplicable). Wrong in the way that building a stone fortress is wrong in an activating earthquake zone. The engineering is flawless for static ground; but the ground is no longer static. The environment has changed.
Here’s the thing about trees that the metaphor conveniently ignores: trees are spectacularly fragile when the climate shifts. A tree that took fifty years to grow can burn in an afternoon. A drought can kill it in a season. A new pest can hollow it from the inside. And because all of its investment is locked into a single trunk: a single, rigid, vertical structure. It can’t move, can’t adapt, can’t redistribute its resources to where the opportunity now lives.
The metaphor holds. Just not in the way its advocates intend.
So if the tree is the wrong organism to model your business on in 2026, what’s the right one?
Allow me to introduce you to the organism that has been quietly winning for approximately one billion years longer than trees have existed.
Part 2: The Organism That Ate the World (Quietly, Underground)
Fungi arrived on land roughly 1.3 billion years ago. Trees showed up about 350 million years ago. This is not a small head start. By the time the first tree put down its first root, fungi had already been running the planet’s nutrient cycling for almost a billion years.
But here’s the wild part: trees didn’t outcompete fungi. Trees couldn’t exist without fungi. The first land plants formed symbiotic partnerships with mycorrhizal fungi to survive on barren rock. The fungi could already break down minerals and transport nutrients through soil. They essentially terraformed the land surface, creating the conditions in which roots could function.
Trees are, in a meaningful sense, a product of the fungal network. They are the tenants acting like they own the building.
Merlin Sheldrake, the biologist whose book Entangled Life (2020) brought fungi into mainstream consciousness, puts it this way: fungi are “metabolic masters, earth makers, and key players in most of life’s processes.” He describes mycelium (the thread-like network that forms the main body of most fungi) as “not a thing but a process: an exploratory, irregular tendency.”
That phrase is eye-catching: not a thing but a process. Not a fixed structure. An ongoing exploration.
This is where the business metaphor gets interesting. Because the default model for business treats a company as a thing. A fixed structure with defined boundaries, a clear trunk, identifiable branches. You can draw it on an org chart. “We do X for Y.”
But a mycelium network is fundamentally different. A single network can span kilometres underground, connecting hundreds of trees, constantly reallocating nutrients, exploring new territory, and – crucially – fruiting in completely different locations depending on where the conditions are right. The mushrooms you see poking above ground? Those are just the temporary, visible expression of a vastly larger, invisible system. They appear in days when conditions align. Then they vanish. The network persists.

This is not a coincidence.
Part 3: The Books That Were Each Looking at One Part of the Elephant
Some of the most influential business and systems thinking of the last twenty years has been circling this same insight from different angles. If you lay their ideas side by side, the shape that emerges is unmistakably mycelial.
Nassim Nicholas Taleb. In Antifragile (2012), Taleb introduced the barbell strategy: combine extreme safety with extreme risk, and avoid the middle. Put 90% in ultra-safe positions and 10% in wildly speculative ones. This is what mycelium does naturally; maintaining a vast, resilient network (safe) while constantly sending exploratory hyphae into unknown territory (speculative). Most explorations fail. The ones that find food become new growth fronts.
Donella Meadows. In Thinking in Systems (2008), Meadows identified self-organisation as one of the highest leverage points in any system. A system that can reorganise its own structure in response to changing conditions will outperform one that can’t, regardless of how optimised the rigid system is for current conditions. A tree cannot reorganise its structure. A mycelium network is continuously self-organising — redirecting resources, abandoning unproductive pathways, thickening transport channels toward productive ones.
Geoffrey West. In Scale (2017), West showed that biological organisms, cities, and companies all follow mathematical scaling laws. Biological organisms (and companies) scale sublinearly: as they grow, they slow down and eventually die. Cities scale superlinearly: as they grow, their innovation increases. Companies that resist death, West argues, must somehow maintain the superlinear scaling of cities. A mycelial network behaves more like a superlinear city; it gets smarter and more capable with every new connection.
Eric Ries and W. Brian Arthur. The Lean Startup’s rapid prototyping is just mycelial rapid fruiting by another name—sending out cheap probes to see what finds traction rather than committing the whole organism. Meanwhile, Arthur’s theory of combinatorial innovation—that new technologies are just combinations of existing ones—perfectly describes how mycelium networks connect disparate domains to create new value.
Now: what does that model actually look like when you build a business around it?
Part 4: Seven Properties of a Mycelial Business
To see this in practice, we just need to look at what happened when the enterprise landscape experienced a seismic shock in 2024–2025.
Consider the customer experience (CX) sector. “Apex CX” (The Tree) had spent five years deeply niching into building custom offshore call centres for e-commerce. Their trunk was rigid and highly legible. When conversational AI agents hit the market, Apex tried to build a proprietary AI wrapper around their massive, legacy human-capital system. It took 18 months, cost millions, and was obsolete the day it launched. Their niche collapsed, and the tree fell.
Contrast that with “Fluid Operations” (The Mycelium), a generalised ops consultancy. They had a broad underground network of process mappers, API integrators, and systems architects. When the AI shock hit, they didn’t try to grow a massive new trunk. They simply “fruited”—rapidly deploying conversational AI modules using off-the-shelf LLMs into whatever fragmented legacy systems their clients already had. Apex died; Fluid digested Apex’s former clients and expanded their network into legal and HR triage.
Here are the seven biological properties that allow the mycelium to survive while the tree shatters:

1. Substrate-Sensing: Continuous Environmental Diagnosis Mycelium constantly reads its chemical environment. It doesn’t wait for a food source to announce itself. The business equivalent is a persistent diagnostic capability. You’re actively sensing problems across your network and identifying opportunities before they become explicit demand. Palantir’s Foundry platform does this by building a living digital model of a business, effectively becoming its central nervous system. The organism that senses best, wins.
2. Rapid Fruiting: Fast Deployment of Context-Specific Solutions Mushrooms appear in days when conditions are right. They’re the visible expression of a capability that was already there, waiting. When fruiting was expensive (building a custom product took years), the tree strategy made sense: commit to one thing, protect your investment. When fruiting is cheap, the mycelial strategy dominates: maintain broad capability underground, fruit rapidly wherever conditions are right.
3. Resource Swarming and Network Intelligence When mycelium detects a new food source, it actively swarms—redirecting growth and thickening transport channels without any central hierarchy telling it where to go. In a conventional organisation, swarming means approvals, resourcing meetings, and six-month hiring plans. In a mycelial business, the organisational layer is soft. Resources flow toward emerging opportunities without hitting structural bottlenecks.
4. Redundancy and Damage Tolerance: Antifragility in Practice Cut a mycelium network in half, and both halves keep functioning. A mycelial business has no single point of failure—no single client, product, or niche whose loss is catastrophic. Companies that niched into “AI chatbots for real estate” in 2023 were commoditised by 2025. The niche was the vulnerability. Firms with broader capability networks just lost a single mushroom. The mycelium was fine.
5. Decomposition as Value Creation: Turning Dead Wood into Nutrients Fungi break down dead organic matter and convert it into usable nutrients. A mycelial business takes broken processes and outdated workflows—the dead wood accumulating inside organisations—and metabolises them. The old consulting model (the tree approach) plants a new massive system next to the old broken one. The mycelial approach is continuous decomposition: come in, digest what isn’t working, convert it to usable nutrients, keep going.
6. Symbiotic Embedding: Mutual Benefit, Not Extraction Mycorrhizal fungi form partnerships with plant roots where both organisms benefit. A mycelial business embeds with clients in a way that makes both organisms stronger. The client gets capability they couldn’t build alone; the provider gets domain intelligence that enriches their platform for everyone. Extractive vendor lock-in is parasitic. A mycelial business leaves the host organism stronger than it found it.
7. Underground Before Above-Ground: Invisible Capability Most of a mycelium network is invisible. What people see—the mushrooms—is just today’s temporary expression. The diagnostic frameworks, the accumulated pattern recognition, the composable solution library—that’s all underground. A tree business looks at a mycelial competitor and sees a confused generalist. What they’re actually seeing is an organism whose underground network is so vast that they can fruit in any direction.

Part 5: Why Now?
If the mycelial model is so superior, why hasn’t it been dominant all along? Because until recently, the cost of maintaining a broad underground network exceeded the cost of committing to a single trunk.
Three things changed: First, AI collapsed the cost of expertise. When your generalists are augmented by AI systems that carry domain-specific depth, the underground network becomes radically cheaper to maintain. Second, AI collapsed the cost of fruiting. Building a prototype used to take months. Now, the “vibe coding” movement (generating functional software entirely through natural language prompts) means a small team can fruit a working application in days. Third, the environment became volatile enough to punish trees. Company lifespans are shrinking because the environment they committed to keeps shifting under them.

Part 6: The Uncomfortable Implication
The mycelial model has an implication that most business advice actively suppresses: you don’t need to know what you are.
A tree knows what it is. It occupies a defined niche. This legibility makes for good elevator pitches. A mycelium network doesn’t know what it is in that sense. Its identity is in its capabilities (sensing, decomposing, connecting) not in its current expressions.
This is deeply uncomfortable for people trained to have a crisp positioning statement. In the mycelial model, the honest answer to “What do you do?” is: “I maintain a network of capabilities that can express as different solutions depending on where the conditions are right.”
To tree-trained ears, this sounds like you don’t have a strategy. To systems theorists, it’s the ultimate strategy: one adapted for volatility rather than stability.
Part 7: How to Become Mycelial
If the analysis is persuasive, how do you actually build this way?
- Invest in sensing before solving. Before you prescribe or build anything, understand the chemical substrate of your client’s actual environment. Mycelium that fruits without sensing first produces mushrooms in the wrong place.
- Keep your organisational layer soft. Hierarchical departments and rigid reporting lines act as hardened tissue that prevents nutrients from flowing to new opportunities. Design your organisation as a fluid transport network, not a containment system.
- Build composable, not custom. Every solution you deliver should generate reusable components that enrich your underlying network. Extract generalisable capability from every specific engagement rather than growing a single, unrepeatable branch.
- Embed, don’t extract. Design your client relationships as genuine symbioses where your network gains intelligence while their system gains capability. Parasites get evicted, but symbionts get invited deeper.
- Let mushrooms die. When a product or service line stops being the right expression for current conditions, let it go immediately. Don’t pour network resources into propping up a mushroom that the environment no longer supports.
- Maintain more capability than you express. The invisible business (your tools, patterns, and frameworks) must always be larger than your visible offerings. If everything you can do is currently deployed, you have no reserve capacity to survive the next shift.
Epilogue: A Billion Years of Evidence
Here is the most striking fact in all of biology, and the one that should inform every business decision made in the age of AI:
Fungi have survived every mass extinction event in Earth’s history. All five of them. The Permian-Triassic extinction wiped out 96% of marine species and 70% of terrestrial vertebrates. The fungi were fine. They ate the dead trees.
Trees, meanwhile, keep getting knocked back. They grow tall, build impressive canopies, dominate landscapes. Then the climate shifts, or the fire comes, or the pest arrives, and they fall. And when they fall, it’s the fungi that digest them and turn their remains into soil for whatever comes next.
The metaphor resolves itself. In a stable world, be a tree. Grow tall. Go deep. Dominate your niche. It works brilliantly… until it doesn’t.
In a volatile world – and we are now, unambiguously, in a volatile world – be a mushroom: Be the organism that senses its environment continuously, fruits solutions rapidly wherever conditions are right, shares intelligence across its entire network, survives damage to any node, creates value by decomposing what’s broken, embeds symbiotically with its hosts, and keeps its capability primed and ready to fruit into action.
The tallest trees in the forest are not the survivors. The widest network underground is.

Matt Stone is the founder of Re-Thrive, an operational effectiveness consultancy that helps £1M-£50M businesses build adaptive, AI-augmented operating systems. He has been a mushroom for longer than he realised.