Robin, this is a powerful and thought-provoking piece. The metaphor of "high-dimensional foraging" is a compelling way to frame the immense challenge of navigating cultural evolution.
From my perspective as a systems engineer, your "high-dimensional space" looks like what I'd call a Constraint Landscape—a complex terrain of possibilities shaped by physical, economic, and informational realities. In this view, a culture isn't so much a path we find through foraging, but an emergent pattern that successfully satisfies thousands of these constraints all at once.
This reframing helps diagnose why the "monoculture" is so dangerous. It becomes a system that powerfully reinforces its own rules, getting stuck in one part of the landscape—like a species trapped in a nutrient-rich but dead-end canyon. It optimizes for its own short-term stability at the expense of the diversity needed for long-term adaptation.
This leads me to think the real task isn't to find one final, stable culture, but rather a resilient capacity generating new, viable cultures—an ecology of them, constantly adapting to a shifting landscape.
This raises a final question. You frame this as a search for an adaptive culture. But perhaps values like justice, dignity, and human flourishing aren't the goal of the search, but are themselves fundamental constraints to emerging cultures? Any cultural pattern that fails to satisfy these deep human needs will ultimately prove unviable, no matter how "adaptive" it appears in the short term. They are a key part of the terrain we're all navigating.
Thanks for mapping this territory. It's a conversation we desperately need to be having.
It seems that both sides of the analogy take for granted the willingness to subsidize “scouting” of new part of dimensional or cultural space.
In practice, I think this scouting mechanic has been very hard to subsidize and implement. Some combination of risk aversion, time discounting, and in-group bias, means that usually excess resources are not put towards meaningful exploration. There is also a problem that current cultural hierarchies tend to be built on assumptions (technologically or culturally) that would be undermined by meaningfully new discovery, further deincentivizing discovery (e.g. Copernicus and Church, U.S. gov’t and lab leak theory).
Have you considered mechanisms that can solve this in practice?
Re. "Thus when physically searching a high dimensional space for rare hiding food, there are big scale and coordination gains to collecting a network of ropes that connect the space, allowing newly discovered food opportunities to be shared by all.":
This is how ants forage on a 2D surface. But they have the advantage that any food they find will eventually run out after many ants have switched to that rope. That's what allows ants to continue adapting, rather than get stuck on a single food source.
Cultures don't necessarily "run out" like food does. There is no reason to think that every culture contains the seed of its own destruction. So the cultural foraging problem is actually 2 problems: how to find new things, and how to abandon old things when better alternatives arise.
Cultures benefit from having adherents. That is, the value to individuals of a culture is an increasing function of the number of individuals already adhering to that culture. So it may be very difficult to change a culture to take advantage of new opportunities. The problem of abandoning one "food source" (value? goal? norm? tool?) may be more-difficult than the problem of foraging.
We do have existing examples of evolutionary lock-ins. In DNA evolution, Hox genes evolved early, had a lot built on top of them and now can't easily change. In cultural evolution the QWERTY keyboard is a famous example of early lock-in.
Could something similar happen to global culture? Absolutely. However what we see with the examples above is that such features often do not exhibit "drifts into maladaption". If they lasted long enough to wipe out their competition, they are often good enough to survive for extended periods of time.
What about when QWERTY keyboards are replaced by man-machine telepathy? Or when mammals are completely replaced by robots - who no longer use Hox genes? I would say this illustrates the importance of takeovers in evolution. Even if the QWERTY keyboard totally wiped out all its competitors, it doesn't mean there's no evolutionary path to replacing it. Nor does it mean we should have better attempted to preserve keyboard pattern diversity.
not a bad pitch for the importance of cultural diversity on multiple scales. not just monocrop "diversity". "diversity of cultures" > "cultures of diversity" in the long-run.
Hey @Robin Hanson what a fun coincidence — I was prepping for a talk on precisely this when I found you'd posted about it the day prior! Included a nod to your article. You might enjoy my take, and if you're up for going deeper on how to address this I think we'd offer people a lot to chew on in a follow-up dialogue.
Nature does have a force resisting monocultures. It is parasites. If any type becomes too common, nature attacks it with parasites, until other types gain advantage over it again. Many organisms use sexual recombination to avoid such attacks. This generates variety and makes life more difficult for the parasites - and it also works against monocultures. The red queen - as it is known as.
Culture can use similar tricks. Printer manufacturers face competition from cheap knock-off ink manufacturers. To avoid this they can twist and turn with their cartridge interfaces - resulting in a diverse plethora of printer cartridge designs that is more difficult for the parasitic firms to track and attack.
However, it should be said that culture has a great potential for producing monocultures. Look at all the US dollars in circulation for example. These may not be identical - but they are close.
Very interesting analogy and a concept I hope you continue to develop.
This trend to monoculture follows in the footsteps of our becoming a monoculture in the biological sense over the last 200k years. Homo Sapiens went from relatively minor species – one of many types of humans – to dominant species in every ecological niche accessible to us. We are like the Cavendish banana, the winner that took all. Is this outcome good? Bad? Or just the way biology works? Sometimes in the high-D search you hit on a food source so good that every worker ant gets drawn to it, for a while. The mathematics of explore versus exploit.
Also there is the challenge of *understanding* in high-D. Just as surface area dominates the bulk in high-D, the information content of the environment quickly overwhelms the ability of any localized organism to understand it. What does this imply about our ability (or inability) to understand the high-D environment and direct our search in useful ways?
In high dimensions, ropes are difficult to constrain, and therefore difficult to follow or use for anything much. Anything we would recognise as a knot for example can't exist. There are surely some sorts of analogues of knots, but they are probably between D-2 dimensional things, not between 1 dimensional things.
Our culture is high dimensional, but it exists in 3D physical space and so can make use of ropes, communications, and all the rest of what we have. This means that it doesn't behave the same as high dimension space, even though "cultural space" is probably infinitely dimensioned. For example communications within and between cultures can probably go from any point to any point with equal ease, whereas communications energy dropping 1/d^(D-1) is going to be problematic for communications in physical high D space.
I doubt that high dimension physical space is a good analogy for culture for a number of reasons, but the primary one being that high dimensional physical space is probably even stranger and more difficult for people to visualise accurately than cultures.
To grasp or attach to a rope in D dimensions as we commonly do in 3 dimensions, you need a high dimensional thing to do it. Even a D-2 dimensional constraint will leave a way for the rope to fall out of the grasp. Some sort of attractive force is needed to attach, like magnetism or something, but it needs to be switchable.
A lot of things we think are basic just won't work the same in high dimensions. Even 4 dimensions is really hard to think about.
Robin, this is a powerful and thought-provoking piece. The metaphor of "high-dimensional foraging" is a compelling way to frame the immense challenge of navigating cultural evolution.
From my perspective as a systems engineer, your "high-dimensional space" looks like what I'd call a Constraint Landscape—a complex terrain of possibilities shaped by physical, economic, and informational realities. In this view, a culture isn't so much a path we find through foraging, but an emergent pattern that successfully satisfies thousands of these constraints all at once.
This reframing helps diagnose why the "monoculture" is so dangerous. It becomes a system that powerfully reinforces its own rules, getting stuck in one part of the landscape—like a species trapped in a nutrient-rich but dead-end canyon. It optimizes for its own short-term stability at the expense of the diversity needed for long-term adaptation.
This leads me to think the real task isn't to find one final, stable culture, but rather a resilient capacity generating new, viable cultures—an ecology of them, constantly adapting to a shifting landscape.
This raises a final question. You frame this as a search for an adaptive culture. But perhaps values like justice, dignity, and human flourishing aren't the goal of the search, but are themselves fundamental constraints to emerging cultures? Any cultural pattern that fails to satisfy these deep human needs will ultimately prove unviable, no matter how "adaptive" it appears in the short term. They are a key part of the terrain we're all navigating.
Thanks for mapping this territory. It's a conversation we desperately need to be having.
It seems that both sides of the analogy take for granted the willingness to subsidize “scouting” of new part of dimensional or cultural space.
In practice, I think this scouting mechanic has been very hard to subsidize and implement. Some combination of risk aversion, time discounting, and in-group bias, means that usually excess resources are not put towards meaningful exploration. There is also a problem that current cultural hierarchies tend to be built on assumptions (technologically or culturally) that would be undermined by meaningfully new discovery, further deincentivizing discovery (e.g. Copernicus and Church, U.S. gov’t and lab leak theory).
Have you considered mechanisms that can solve this in practice?
I don't have specific mechanisms in mind yet, but my thinking on all this is still in early stages.
One of your geekiest posts ever. I love it.
Intriguing.
Re. "Thus when physically searching a high dimensional space for rare hiding food, there are big scale and coordination gains to collecting a network of ropes that connect the space, allowing newly discovered food opportunities to be shared by all.":
This is how ants forage on a 2D surface. But they have the advantage that any food they find will eventually run out after many ants have switched to that rope. That's what allows ants to continue adapting, rather than get stuck on a single food source.
Cultures don't necessarily "run out" like food does. There is no reason to think that every culture contains the seed of its own destruction. So the cultural foraging problem is actually 2 problems: how to find new things, and how to abandon old things when better alternatives arise.
Cultures benefit from having adherents. That is, the value to individuals of a culture is an increasing function of the number of individuals already adhering to that culture. So it may be very difficult to change a culture to take advantage of new opportunities. The problem of abandoning one "food source" (value? goal? norm? tool?) may be more-difficult than the problem of foraging.
We do have existing examples of evolutionary lock-ins. In DNA evolution, Hox genes evolved early, had a lot built on top of them and now can't easily change. In cultural evolution the QWERTY keyboard is a famous example of early lock-in.
Could something similar happen to global culture? Absolutely. However what we see with the examples above is that such features often do not exhibit "drifts into maladaption". If they lasted long enough to wipe out their competition, they are often good enough to survive for extended periods of time.
What about when QWERTY keyboards are replaced by man-machine telepathy? Or when mammals are completely replaced by robots - who no longer use Hox genes? I would say this illustrates the importance of takeovers in evolution. Even if the QWERTY keyboard totally wiped out all its competitors, it doesn't mean there's no evolutionary path to replacing it. Nor does it mean we should have better attempted to preserve keyboard pattern diversity.
not a bad pitch for the importance of cultural diversity on multiple scales. not just monocrop "diversity". "diversity of cultures" > "cultures of diversity" in the long-run.
Hey @Robin Hanson what a fun coincidence — I was prepping for a talk on precisely this when I found you'd posted about it the day prior! Included a nod to your article. You might enjoy my take, and if you're up for going deeper on how to address this I think we'd offer people a lot to chew on in a follow-up dialogue.
https://michaelgarfield.substack.com/p/foraging
Nature does have a force resisting monocultures. It is parasites. If any type becomes too common, nature attacks it with parasites, until other types gain advantage over it again. Many organisms use sexual recombination to avoid such attacks. This generates variety and makes life more difficult for the parasites - and it also works against monocultures. The red queen - as it is known as.
Culture can use similar tricks. Printer manufacturers face competition from cheap knock-off ink manufacturers. To avoid this they can twist and turn with their cartridge interfaces - resulting in a diverse plethora of printer cartridge designs that is more difficult for the parasitic firms to track and attack.
However, it should be said that culture has a great potential for producing monocultures. Look at all the US dollars in circulation for example. These may not be identical - but they are close.
Very interesting analogy and a concept I hope you continue to develop.
This trend to monoculture follows in the footsteps of our becoming a monoculture in the biological sense over the last 200k years. Homo Sapiens went from relatively minor species – one of many types of humans – to dominant species in every ecological niche accessible to us. We are like the Cavendish banana, the winner that took all. Is this outcome good? Bad? Or just the way biology works? Sometimes in the high-D search you hit on a food source so good that every worker ant gets drawn to it, for a while. The mathematics of explore versus exploit.
Also there is the challenge of *understanding* in high-D. Just as surface area dominates the bulk in high-D, the information content of the environment quickly overwhelms the ability of any localized organism to understand it. What does this imply about our ability (or inability) to understand the high-D environment and direct our search in useful ways?
In high dimensions, ropes are difficult to constrain, and therefore difficult to follow or use for anything much. Anything we would recognise as a knot for example can't exist. There are surely some sorts of analogues of knots, but they are probably between D-2 dimensional things, not between 1 dimensional things.
Our culture is high dimensional, but it exists in 3D physical space and so can make use of ropes, communications, and all the rest of what we have. This means that it doesn't behave the same as high dimension space, even though "cultural space" is probably infinitely dimensioned. For example communications within and between cultures can probably go from any point to any point with equal ease, whereas communications energy dropping 1/d^(D-1) is going to be problematic for communications in physical high D space.
I doubt that high dimension physical space is a good analogy for culture for a number of reasons, but the primary one being that high dimensional physical space is probably even stranger and more difficult for people to visualise accurately than cultures.
I was imagining climbing along a rope to travel from one end to the other, and vibrations in the rope telling about good vs bad food at the other end.
To grasp or attach to a rope in D dimensions as we commonly do in 3 dimensions, you need a high dimensional thing to do it. Even a D-2 dimensional constraint will leave a way for the rope to fall out of the grasp. Some sort of attractive force is needed to attach, like magnetism or something, but it needs to be switchable.
A lot of things we think are basic just won't work the same in high dimensions. Even 4 dimensions is really hard to think about.
Attractive forces are available in this scenario.
An argument for Network States if ever I've read one. The centralisation of human culture will lead to stagnation & ruin.
do I remember reading you may have a book on cultural drift in the works?