Carlos Upgrade | November 2, 2026 | Biology, collective intelligence and design
A Wikipedia editor can improve a stranger’s paragraph without ever speaking to its author. The changed page points to the next task.
This is a useful doorway into stigmergy: indirect coordination through traces left in a shared environment. The important feature is that one person’s trace changes another’s next action.
1. A nest without a single blueprint
French biologist Pierre-Paul Grassé introduced the term in 1959 while studying social insects. A 1999 review by Guy Theraulaz and Eric Bonabeau explains how insects respond to local traces to coordinate construction without a central blueprint.
Think of a termite building site. Material already deposited can influence where subsequent material is placed. In the broader literature, signals may vary by amount (a stronger trace attracts more activity) or by kind (a particular local configuration invites a particular response). The mechanisms differ.
The minimal loop is:
Act → change the shared environment → another participant reads the change → act again.
Termites also communicate in other ways, and nest-building details remain under study. The central point is that a shared environment can help coordinate the next action without any insect holding a complete architectural plan.
2. Outside CPG: Wikipedia edits
Consider someone updating a historical article. They correct a date and add a citation. A later editor encounters that revised paragraph and can extend it, challenge it or repair an omission. The shared page is both the product and part of the coordination system.
A 2023 study by Zheng and colleagues analyzed thousands of Wikipedia articles. It identified patterns of clustered contributions consistent with stigmergic coordination and found associations with participation and article quality. That supports the framework, but does not establish causation.
A misleading edit can invite bad follow-up work. Visibility needs correction mechanisms.
3. A beverage application
Imagine a hypothetical field team checking a new beverage's shelf presence across 40 stores. One representative finds the SKU missing from its expected location and mentions it in a group chat. Two days later, the next rep cannot tell whether the issue was resolved.
Now compare a different process: the rep records a time-stamped shelf photo, the exact store and SKU, what was checked, and a clearly assigned next action in the shared store record. The record makes the next useful move clearer.
The commercial decision is to pilot that field record against the current workflow, measuring time to verified resolution, repeat visits, stale reports and unnecessary escalations. Don't assume a photograph will improve sales. Test whether it reduces missed handoffs without adding excessive administrative work.
4. An assumption worth challenging
“Better collaboration means more meetings and messages.”
Sometimes the problem is that useful evidence disappears after the call. If work can leave a trustworthy, accessible trace, colleagues may need fewer explanations of what happened.
The counterargument: a shared board can circulate outdated signals. Critical decisions still need owners, privacy controls and approvals. Stigmergy can support explicit accountability; it cannot replace it.
5. Your mental operator
Use Leave the next clue. After a task, ask: What changed? Where can the next person see it? What action should the change prompt? How will they know when the signal is stale or wrong?
Design the trace for the next decision, not the appearance of activity.
6. Reasoning exercise: answer before reading on
In an illustrative museum inventory project, a volunteer flags an artifact as missing at 9 a.m. Another volunteer finds it at noon but never updates the shared record. At 3 p.m., a third volunteer starts an expensive search based on the old flag.
Which part of the stigmergic loop failed? Would adding more volunteers solve it?
Worked answer
The environmental trace was not updated when the underlying state changed. The third volunteer responded rationally to stale information. More volunteers might amplify the wasted effort. A better system requires both a visible flag and an equally easy way to resolve it, with a timestamp and a clear owner for verification.
Keep this: Coordination is not only what we tell people. It is what our actions leave behind for the next person to read.
Research links
- Theraulaz and Bonabeau (1999), A Brief History of Stigmergy.
- Bonabeau, Dorigo and Theraulaz (1999), Nest Building and Self-Assembling.
- Zheng and colleagues (2023), Stigmergy in Open Collaboration: An Empirical Investigation Based on Wikipedia.
Research sources and notes
Three authoritative scholarly sources were opened and checked for attribution, mechanism, and the scope of the examples. Claims about Wikipedia are observational associations, not proven causal effects. The beverage and museum scenarios in the lesson and the beverage scenario in the LinkedIn post are hypothetical.
Guy Theraulaz and Eric Bonabeau (1999). “A Brief History of Stigmergy.” Artificial Life 5(2):97–116. DOI 10.1162/106454699568700. PubMed indexes the review and its abstract, confirming Pierre-Paul Grassé introduced the term in 1959 and that the authors distinguish quantitative from qualitative stigmergy. https://pubmed.ncbi.nlm.nih.gov/10633572/
Eric Bonabeau, Marco Dorigo and Guy Theraulaz (1999). “Nest Building and Self-Assembling,” chapter in Swarm Intelligence: From Natural to Artificial Systems, Oxford University Press. Describes coordination through the environment as an important mechanism of social insect nest construction, while distinguishing quantitative and qualitative stigmergy. https://academic.oup.com/book/40811/chapter-abstract/348772109
Lei Zheng, Feng Mai, Bei Yan and Jeffrey V. Nickerson (2023). “Stigmergy in Open Collaboration: An Empirical Investigation Based on Wikipedia.” Journal of Management Information Systems 40(3):983–1008. DOI 10.1080/07421222.2023.2229119. The published abstract describes analysis of thousands of articles, spatial-temporal contribution patterns, and positive associations with participation and quality. https://www.jmis-web.org/articles/1636
Educational editorial content. Hypothetical examples are labeled in the original edition.
