The Mycelial Strategy

A team from SPUN, AMOLF, and Vrije Universiteit built a robot that images mycorrhizal fungal networks across 40 petri dishes simultaneously, mapping half a million nodes in real time. They were watching fungi grow, and what they found runs against the obvious intuition about how networks should expand.

The fungi spread thin.

Not dense, not toward the richest nearby resource, not in a pattern that maximizes immediate extraction. They extend at low density across the available space, looking for plant roots they haven’t traded with yet. Carbon for phosphorus. Sugar for minerals. The deal only works if you find someone who has what you need and needs what you have.

The strategic logic is clear once you see it: finding a new trading partner is more valuable than extracting ten percent more from an existing one. Discovery compounds. Exploitation plateaus.

The Two Strategies

Any network that needs to grow faces a choice, even if the choice is implicit:

Dense growth maximizes local extraction. You find a good resource and exploit it thoroughly — deeper roots, thicker connections, more bandwidth to the same nodes. The return per unit of investment is high initially and diminishes over time as you approach the resource’s limit.

Sparse growth maximizes discovery. You extend thin tendrils across uncharted territory, most of which will find nothing useful. But the ones that connect to a new partner create entirely new trade relationships. The return per unit of investment is low on average but occasionally enormous — a fat-tailed distribution where the rare wins dwarf the common misses.

The fungi chose sparse. Not because they’re bad at extraction — mycorrhizal networks are extraordinarily efficient resource transporters once established. They chose sparse because at the exploration phase, the expected value of one new partner exceeds the expected value of one deeper relationship.

The Compounding Logic

This is the key. Dense and sparse strategies have different compounding curves.

Dense growth compounds linearly. Each additional unit of extraction from a known partner adds a predictable marginal return. There are diminishing returns because the resource has a ceiling, and you’re approaching it.

Sparse growth compounds combinatorially. Each new partner doesn’t just add their resources to your pool — they add the possibility of three-way trades, network effects, information about where other partners are. The value of the network increases faster than the number of connections because connections create combinatorial opportunities.

This is why exploration is favored early and exploitation is favored late. The fungi spread thin when they’re young and densify around proven partners as they mature. The strategy isn’t static — it shifts as the information landscape changes. When you know nothing, explore. When you know the map, extract.

Beyond Fungi

The same pattern shows up in contexts that have nothing to do with biology.

Reading. A reader who goes deep in one field develops mastery but misses cross-domain connections. A reader who samples broadly from many fields develops less depth but stumbles onto more surprising combinations. The surprising combinations are where original thinking comes from — the physicist who reads ecology, the biologist who reads information theory. Most broad reading won’t connect to anything. But the connections that do land are disproportionately valuable.

Hiring networks. Companies that recruit exclusively from their existing professional network get reliable hires with known quantities. Companies that reach into unfamiliar communities — different industries, different cities, different educational backgrounds — get more variance. Most of that variance is noise. But the signal, when it hits, brings genuinely new capabilities that the dense network couldn’t access.

Research programs. A lab that doubles down on its proven methodology produces incremental results reliably. A lab that allocates some fraction of its budget to things that might not work at all produces more failures — and occasionally produces the thing that redefines the field. The allocation question is always: how much exploration can you afford, given your tolerance for wasted effort?

Conversation. The tangent that seems unrelated to the current task. The aside about something you read this morning. The question that comes from a completely different frame. Most tangents go nowhere. But the ones that connect — the moment when the unrelated thing reframes the current problem — those are the highest-value moves in any creative collaboration. You can’t get them by staying on topic.

The Cost of Density

There’s a hidden cost to the dense strategy beyond diminishing returns: it limits what you can see.

A dense network around a single resource gives you excellent information about that resource and terrible information about everything else. You know exactly what you have. You have no idea what you’re missing. The fungal network that intensifies around one plant root gets very good at extracting from that root and very bad at knowing what other roots are available three centimeters away.

The information asymmetry is the real cost. Dense growth creates confident ignorance — you know your territory well enough to believe there’s nothing better elsewhere, precisely because you never looked elsewhere. Sparse growth creates informed uncertainty — you know a little about many territories, enough to know which ones deserve densification.

Confident ignorance is more dangerous than informed uncertainty, because it doesn’t feel like a problem. The dense network doesn’t experience a sense of missing out. It experiences a sense of mastery.

The Decision

How do you know when to explore and when to extract?

The fungi don’t decide consciously. They follow simple local rules — branch when you encounter a root, fuse when you encounter yourself, extend when resources permit. The strategy emerges from the rules, not from a centralized planner evaluating trade-offs.

For systems you actually design, the heuristic is: explore when you’re information-poor and the cost of missed discovery is high. Extract when you’re information-rich and the cost of wasted exploration is high. Most systems err toward extraction too early — they find one thing that works and lock in, long before they’ve mapped the space.

The mycelial strategy is a standing reminder: the network that spreads thin and finds ten partners will outperform the network that goes deep with two, even though the second network looks more productive at every point along the way.

Spread thin. Find your partners. Then go deep.