First thought is that probably the concept explains why "Self-Discipline" is mostly a scam. We don't act based on what we "want." We act based on the signals in our immediate environment.
If your phone is on the desk, the "trace" says: "Scroll."
If your running shoes are by the door, the "trace" says: "Run."
We are engineers of our own stupidity because we design environments that trigger the wrong algorithms.
Could stigmergy be a solution to the Prisoner’s Dilemma? I don’t see how. It works with ants and termites because all the “players” have the same motivation (reproductive success), but does it work in a human society when each individual have their own goals, some that conflict?
Read my papers (I & II) for a discussion of that problem. Agents don’t need to have the same goals to collaborate efficiently via stigmergy. I may want to write a Wikipedia paper on physics, while you write one on the history of football competition: different goals, but both add to the encyclopedia. The only thing to avoid are conflicting goals: I want to claim that country X started the war, while you claim it was country Y. In Wikipedia, this can lead to “edit wars”, where people undo each other’s contributions. This is relatively uncommon, and usually resolved by putting both views side-to-side: according to A, X started the war, while B claims it was Y. This cannot resolve all prisoners’ dilemma type of situation, but can take the sting out of many conflicts…
I think the Prisoner’s Dilemma is quite a consequential problem. We experience it in climate change, nuclear non-proliferation, ensuring that governments do not rely excessively on tax revenues from labour income while leaving capital free of tax, etc. I suspect all these challenges cannot be addressed without planning and conscious coordination.
This more or less *is* natural selection, right? Or perhaps natural selection and stigmergy both orbit a more general concept of recursive iteration which builds on top of an iteration which already worked
In a sense, stigmergy is the mirror image of natural selection. With natural selection, an agent (typically an organism) undergoes variation. The environment then selects and reinforces good variations. With stigmergy, the environment undergoes variation. An agent then decides whether the variation (trace) is good as it is, or must be replaced by another variation (action).
Hold up brah, the chemical markers are the language of insects, there are at least 20 known vocab words ants express via chemicals. Its not like they're just acting on their enviroment, their chemical markers are like getting an email.
I'm most interested in our shared growing knowledge of ways nature works, en route to practical respect for stewardship of these ways as they are, and as they may offer appropriate engagement with them.
What a beautiful concept! I wrote a couple days ago about how bees "intelligently" decide where to find the richest source of nectar, just like the ant example you provide. It's just a pity that 'stigmergy' is quite an ugly word; if I had to suggest another word 'tracecraft' sounds nice to my ears.
Exciting concept and a great article. It would be good to also explore the limitations. Termites know only how to build one kind of structure. They cannot build anything else. The knowledge of how to build it based on the cues in the substrate is embedded in their biology. This concept breaks for more universal or complex tasks spanning across different domains. It's worth studying, but the applications are limited.
The only limits are the limits of the skills distributed among the agents. What stigmergy does is aggregate those skills so that the whole becomes much larger and more complex than any single agent could have produced. This is what Wikipedia demonstrates: no individual author could have that much knowledge, but all together can cover the knowledge of humanity as a whole.
Good point. Application is limited and doesn’t worth it in many scenarios.
This approach solves bottlenecks of communication and coordination. Small scale usually doesn’t have this bottlenecks so not much gain in that.
This is very similar to pull in the information flow, someone’s post something in the environment (i.e slack, kanban board) and I pull the work. Unprompted.
The “normal” way is someone pushing i.e calling action by someone else and coordinating if the person has availability.
Christopher Alexander (The Nature of Order, A Pattern Language, etc.) developed a concept where any existing center or field of centers at a point in time, can be strengthened, elaborated, increased in function and beauty--by adding centers to the field at both smaller and larger levels. The real key to his architectural design thinking was to begin with what exists now (T 1) and make stepwise improvements (T 2, T 3, T 4, etc.)
Christopher Alexander’s ideas are a revelation in this field. A good example: If you want to know where to put hardened walkways among buildings on a campus, wait until you see where students leave their traces of footsteps on the earth or in the snow.
Where do the questions come from? The questions that everyone somehow knows to ask when confronted with the materiality of the trace left in the system? Why is it that those who approach the wikipedia article ask themselves "is it correct? what is missing?" out of all the questions they might ask about it?
This explainer you've written seems to imply that the questions emerge somehow naturally from the trace, but I don't think that's true. I say this having been a member of at least three of these self-organizing systems. It almost seemed magical, the way a person in the system could pick up where another left off, but it wasn't.
In all three cases, it was person-to-person education that gave us the shared mindset. I use the word "education" broadly. In one case, it was actual education at a university in a particular methodology that united our purpose. In another case, it was a single leader with a clear vision and huge amounts of buy-in. In the third, it was on-the-job training through constant interaction with people already in the system who could convey the meaning of the environmental inscriptions, which were supplied until you could operate on your own. One of these systems was filled with volunteers, which made a difference, but the organization was creating the interpretations of the inscriptions internally somehow.
These are the little magics that adminstrators try to create through ice-breakers, onboarding, and top-down trainings and they always fail, I agree. In fact, the work of these people often destroys self-organization and then sits there wondering why nothing works under their regime.
So, what I hope you can tell me is, where to the shared approaches to the environmental inscriptions come from? How do people in the system learn to read them and respond appropriately? Only part of the incitement can come from the inscriptions themselves - if we find a chair, most of us will assume it's for sitting, sure. But the inscriptions won't always be that clear. If the question in wikipedia editing is "is it accurate" instead of "what is the best way to make this into believable propaganda" or "how can we use this to make people believe fantasies," where did that shared purpose emerge from?
The ant analogy breaks down here. Their pheremone tracks are entirely about biology. We are more complex and may interpret inscriptions differently among us. If we want a self-organizing entity to do more than simply look for food items in a single terrain and address complex or cognitive issues, we need to take education into account.
I would add to this. Not only are we biased toward intentional explanation, we are conditioned for it. Ever since kindergarten, until high school, followed by college, most of what we know comes from the "lecture" of more knowledgeable other. A system of efficiency, mimicking the successful assembly line, to mass produce what we call "intellectual", which only have "obedience" as its redeeming quality, such is the state of our education system. A child, then full of curiosity, is forced to follow a meticulously prepared curriculum of increasing difficulty, where inquiry outside of what is taught deemed as unnecessary. Not only do their curiosity dwindle as they age, but their mental model of the world, lacking in complexity, failed to see the world with all its nuances. This is because school water down complex topic so much that even children can digest it. This is by no means the best. But it's efficient. And thus, in our education system today, children sacrifices their curiosity for obedience. All to meet industry demand.
Stigmergy, as I see it, has three hidden qualification:
- You need to be part of the project enough to notice room for improvement.
- You need to be skilled enough to see the traces left behind by other.
- You need to be even more skilled to contribute to the project.
I fear that future generation won't be able to handle this. Youth mental health crisis, capitalisation of attention, leveraging thinking with AI. Will we have enough youth to crowd all the crucial stigmergy projects?
What a great concept.
Didn't know it.
First thought is that probably the concept explains why "Self-Discipline" is mostly a scam. We don't act based on what we "want." We act based on the signals in our immediate environment.
If your phone is on the desk, the "trace" says: "Scroll."
If your running shoes are by the door, the "trace" says: "Run."
We are engineers of our own stupidity because we design environments that trigger the wrong algorithms.
Great article.
Bravo!
Ahh this last line in your comment is gold:
“We are engineers of our own stupidity because we design environments that trigger the wrong algorithms.”
Appreciate it! Thanks
Could stigmergy be a solution to the Prisoner’s Dilemma? I don’t see how. It works with ants and termites because all the “players” have the same motivation (reproductive success), but does it work in a human society when each individual have their own goals, some that conflict?
Read my papers (I & II) for a discussion of that problem. Agents don’t need to have the same goals to collaborate efficiently via stigmergy. I may want to write a Wikipedia paper on physics, while you write one on the history of football competition: different goals, but both add to the encyclopedia. The only thing to avoid are conflicting goals: I want to claim that country X started the war, while you claim it was country Y. In Wikipedia, this can lead to “edit wars”, where people undo each other’s contributions. This is relatively uncommon, and usually resolved by putting both views side-to-side: according to A, X started the war, while B claims it was Y. This cannot resolve all prisoners’ dilemma type of situation, but can take the sting out of many conflicts…
I think the Prisoner’s Dilemma is quite a consequential problem. We experience it in climate change, nuclear non-proliferation, ensuring that governments do not rely excessively on tax revenues from labour income while leaving capital free of tax, etc. I suspect all these challenges cannot be addressed without planning and conscious coordination.
This more or less *is* natural selection, right? Or perhaps natural selection and stigmergy both orbit a more general concept of recursive iteration which builds on top of an iteration which already worked
In a sense, stigmergy is the mirror image of natural selection. With natural selection, an agent (typically an organism) undergoes variation. The environment then selects and reinforces good variations. With stigmergy, the environment undergoes variation. An agent then decides whether the variation (trace) is good as it is, or must be replaced by another variation (action).
Hold up brah, the chemical markers are the language of insects, there are at least 20 known vocab words ants express via chemicals. Its not like they're just acting on their enviroment, their chemical markers are like getting an email.
Barlow, thanks, that’s a constructive comment. Could you please list the 20 or so words? Or suggest links we might look to to find them?
E.O. Wilson; Chemical Language of Ants https://vimeo.com/34924361
Thanks Barlow. I'll go look.
I'm most interested in our shared growing knowledge of ways nature works, en route to practical respect for stewardship of these ways as they are, and as they may offer appropriate engagement with them.
What a beautiful concept! I wrote a couple days ago about how bees "intelligently" decide where to find the richest source of nectar, just like the ant example you provide. It's just a pity that 'stigmergy' is quite an ugly word; if I had to suggest another word 'tracecraft' sounds nice to my ears.
Agreed on the aesthetics of both the word and the concept. Trancecraft is fun, sounds like an event I’d like to participate in.
To me it seems another word for stimergy is simply co-creation.
That works too!
This is a nice representation of what us Austrian Economists call “Spontaneous Order” thanks to the brilliant work of Hayek.
Exciting concept and a great article. It would be good to also explore the limitations. Termites know only how to build one kind of structure. They cannot build anything else. The knowledge of how to build it based on the cues in the substrate is embedded in their biology. This concept breaks for more universal or complex tasks spanning across different domains. It's worth studying, but the applications are limited.
The only limits are the limits of the skills distributed among the agents. What stigmergy does is aggregate those skills so that the whole becomes much larger and more complex than any single agent could have produced. This is what Wikipedia demonstrates: no individual author could have that much knowledge, but all together can cover the knowledge of humanity as a whole.
Good point. Application is limited and doesn’t worth it in many scenarios.
This approach solves bottlenecks of communication and coordination. Small scale usually doesn’t have this bottlenecks so not much gain in that.
This is very similar to pull in the information flow, someone’s post something in the environment (i.e slack, kanban board) and I pull the work. Unprompted.
The “normal” way is someone pushing i.e calling action by someone else and coordinating if the person has availability.
good constructive point, thanks.
good constructive answer, thanks.
Such a clear and beautiful explanation! Another helpful resource for thinking about how the noosphere operates.
So inspired by the article and concept, aligned directly to work I'm doing now on human cognition and biolinguistic creativity that I had to write a response: https://open.substack.com/pub/miloidea/p/stigmergy-and-the-myth-of-control?r=3eq816&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true
This is gold. Stigmergy means admitting we don’t truly have control, or know what we are doing. Acknowledging it means accepting our limitations.
Christopher Alexander (The Nature of Order, A Pattern Language, etc.) developed a concept where any existing center or field of centers at a point in time, can be strengthened, elaborated, increased in function and beauty--by adding centers to the field at both smaller and larger levels. The real key to his architectural design thinking was to begin with what exists now (T 1) and make stepwise improvements (T 2, T 3, T 4, etc.)
Christopher Alexander’s ideas are a revelation in this field. A good example: If you want to know where to put hardened walkways among buildings on a campus, wait until you see where students leave their traces of footsteps on the earth or in the snow.
Exactly.
Alexander did not believe in master-strokes but rather intensification over time (and collaboration with others).
This is beautiful.
Also, "coordination of action based on past activity" literally describes a block chain. We need blockchain-based governance.
Where do the questions come from? The questions that everyone somehow knows to ask when confronted with the materiality of the trace left in the system? Why is it that those who approach the wikipedia article ask themselves "is it correct? what is missing?" out of all the questions they might ask about it?
This explainer you've written seems to imply that the questions emerge somehow naturally from the trace, but I don't think that's true. I say this having been a member of at least three of these self-organizing systems. It almost seemed magical, the way a person in the system could pick up where another left off, but it wasn't.
In all three cases, it was person-to-person education that gave us the shared mindset. I use the word "education" broadly. In one case, it was actual education at a university in a particular methodology that united our purpose. In another case, it was a single leader with a clear vision and huge amounts of buy-in. In the third, it was on-the-job training through constant interaction with people already in the system who could convey the meaning of the environmental inscriptions, which were supplied until you could operate on your own. One of these systems was filled with volunteers, which made a difference, but the organization was creating the interpretations of the inscriptions internally somehow.
These are the little magics that adminstrators try to create through ice-breakers, onboarding, and top-down trainings and they always fail, I agree. In fact, the work of these people often destroys self-organization and then sits there wondering why nothing works under their regime.
So, what I hope you can tell me is, where to the shared approaches to the environmental inscriptions come from? How do people in the system learn to read them and respond appropriately? Only part of the incitement can come from the inscriptions themselves - if we find a chair, most of us will assume it's for sitting, sure. But the inscriptions won't always be that clear. If the question in wikipedia editing is "is it accurate" instead of "what is the best way to make this into believable propaganda" or "how can we use this to make people believe fantasies," where did that shared purpose emerge from?
The ant analogy breaks down here. Their pheremone tracks are entirely about biology. We are more complex and may interpret inscriptions differently among us. If we want a self-organizing entity to do more than simply look for food items in a single terrain and address complex or cognitive issues, we need to take education into account.
Great work, Francis.
If interested, I've designed a school around this concept.
> we are biased toward intentional explanations.
I would add to this. Not only are we biased toward intentional explanation, we are conditioned for it. Ever since kindergarten, until high school, followed by college, most of what we know comes from the "lecture" of more knowledgeable other. A system of efficiency, mimicking the successful assembly line, to mass produce what we call "intellectual", which only have "obedience" as its redeeming quality, such is the state of our education system. A child, then full of curiosity, is forced to follow a meticulously prepared curriculum of increasing difficulty, where inquiry outside of what is taught deemed as unnecessary. Not only do their curiosity dwindle as they age, but their mental model of the world, lacking in complexity, failed to see the world with all its nuances. This is because school water down complex topic so much that even children can digest it. This is by no means the best. But it's efficient. And thus, in our education system today, children sacrifices their curiosity for obedience. All to meet industry demand.
Stigmergy, as I see it, has three hidden qualification:
- You need to be part of the project enough to notice room for improvement.
- You need to be skilled enough to see the traces left behind by other.
- You need to be even more skilled to contribute to the project.
I fear that future generation won't be able to handle this. Youth mental health crisis, capitalisation of attention, leveraging thinking with AI. Will we have enough youth to crowd all the crucial stigmergy projects?
This is but my own speculation.
Stigmergy makes leadership look less like direction and more like architecture.