5 ms·

The Simple Algorithm That Ants Use to Build Bridges (2018)

120 points by silverkite 2y ago

wonderwonder 2y ago
Always found colony based insects but especially ants fascinating. Its as if the ant is not the organism but rather the colony is.
- senthil_rajasek 2y ago
  If you find Ants interesting you would probably like E.O. Wilson's works. That's all he did his entire life, study ants. https://en.m.wikipedia.org/wiki/Journey_to_the_Ants https://en.m.wikipedia.org/wiki/Journey_to_the_Ants
  - SeanAnderson 2y ago
    E.O. Wilson's published work was read by Will Wright, creator of SimCity / The Sims / SimAnt, and the material is what inspired him to create SimAnt :) Will has also stated that the pheromone system designed for SimAnt was repurposed to be used in The Sims! Very interesting stuff if you ask me :)
  - withants 2y ago
    Thank you. What a life!
- jsenn 2y ago
  Deborah Gordon (ant biologist) makes this argument in her book Ant Encounters. Because the ants in a colony are all sisters, the unit of reproduction is actually the colony rather than the individual ant. Great book!
  - kevinventullo 2y ago
    The unit of reproduction, and thus the unit of Darwinian evolution!
  - wonderwonder 2y ago
    Thanks for the rec, just ordered it. Its part of a series on complex systems, pretty cool
- EdwardCoffin 2y ago
  "A colony of ants is more than just an aggregate of insects that are living together. One ant is no ant." - E. O. Wilson
- munificent 2y ago
  An interesting follow-up thought is to what degree that is also true of humans and their tribes.
- thowawatp302 2y ago
  Have you extended that idea to the perspective of your cells and you, and thought up any major differences? (Genuine question, the biggest difference I can think of is most of our cells are in permanent physical contact.)
  - wonderwonder 2y ago
    Actually yes, its an interesting thought. Not only are we a cooperative group of independent cells but those cells are fully replaced as we age so not even the original cells we start with. We are essentially walking cities of cells and somehow we gain consciousness. Pretty wild.
    - KineticLensman 2y ago
      > but those cells are fully replaced as we age Not nerve cells. Our post-childhood adult brains persist until we die, so the city of cells (nice image BTW) has a permanent and persistent ruling council, as it were, even if the citizens come and go.
digging 2y ago
It's a pretty cool finding, but a shallow article and arguably clickbait. > ... it’s very possible there’s more governing army ant behavior than two simple rules. Then don't say otherwise in the title!
- EdwardCoffin 2y ago
  I read the article. I think that though it is relatively short, it isn't shallow, and certainly isn't clickbait. They also point to the paper the article draws from if you want more. >> ... it’s very possible there’s more governing army ant behavior than two simple rules. >Then don't say otherwise in the title! The title constrains the rules to those dictating bridge building, whereas the quoted caveat above is more general.
- beerandt 2y ago
  Yea they don't really define stampede, let alone the subsidence/ threshold to break-off. That's not a simple rule. Is it traffic, weight, vibrations? An absolute threshold for decrease, or just a rate drop? Which made me think- maybe a second rule exists, but it probably isn't necessary from an 'algorithm works' standpoint. If an ant starts marching again once it no longer has another on top of it- once the supply of ants starts to dwindle (too many tied up in bridge structure), the bridge will start to deconstruct itself on its own. It's a status toggled by other elements having the same boolean column. If too many are 'busy' as bridges, that leaves too few to hold the toggle buttons of the remaining others. It's not a bridge, it's a self-sorting, self-clearing traffic jam. You drive when you can. Which as a civil engineer the traffic aspect is more interesting to me than the bridge aspect. But probably would get fewer clicks.
notfed 2y ago
"Wow, this article totally reminds me of Stephen Wolfram's work! It's amazing how these ants are following simple rules to create such complex behaviors. This is exactly what Wolfram has been talking about for years with his idea of cellular automata and complexity theory. I remember reading about how he pioneered the whole idea that simple rules can lead to complex systems. It feels like everything in nature is just proving him right, from the patterns in seashells to how these ants behave. I bet if we looked deeper, we'd find that Wolfram's principles are at the core of so many natural processes. It's incredible to see his influence everywhere!" I generated the above paragraph for satire reasons using ChatGPT.
all_these_years 2y ago
The second video seems to have some issues: https://www.youtube.com/watch?v=KJUk5VDCKbQ https://www.youtube.com/watch?v=KJUk5VDCKbQ
cen4 2y ago
Besides simple rules the other requirement is Altruism. Everyone in the colony takes care of the children of one queen. Can you imagine the chimp troupe doing any such thing? Siblings will kill each other and their kids rather than do that. People think its related to Ants being haplodiploid. Altruism in large social groups is something chimps haven't worked out yet. Which is why we have the MIL complex and war. So if you are an algo engineer check out Haplodiploidy.
- SeanAnderson 2y ago
  There are ant colonies with multiple queens :) But yes your general point still stands. https://en.wikipedia.org/wiki/Pleometrosis https://en.wikipedia.org/wiki/Pleometrosis
- Nursie 2y ago
  > Everyone in the colony takes care of the children of one queen. Depending on how you look at it, she is their prisoner and the only way they get to pass on a proportion of their genetic material is by looking after her kids that share some of it. Alternatively you can look at them as little automata largely driven by pheromones. Or you could see them as part of a larger pseudo-organism that doesn't really have individual wants and drives in the way we understand them. Fascinating either way though.
- daniel-s 2y ago
  This is really cynical. You live in a society where almost everyone goes to work to produce something that they will not directly consume. Humans figured out a way to organise everyone and get them to work cooperatively, it's the free market. In fact, it couldn't work any other way. Suppose you had 100% altruism and everyone wanted to promote the "greater good". You wouldn't ever know how to direct people towards that greater good. You need prices to drive production.
  - M95D 2y ago
    I doubt that free market drives people toward "greater good". What is "greater good" anyway? Can you give some examples of "greater good" and how the free market achieved it? BTW, no market is truly free, not unless it allows fakes, quack medicine, child labor, selling humans, or other horrors of the past.
xhevahir 2y ago
"Plant arithmetic" is another interesting phenomenon: https://en.wikipedia.org/wiki/Plant_arithmetic?wprov=sfla1 https://en.wikipedia.org/wiki/Plant_arithmetic?wprov=sfla1

120 points by silverkite 2y ago

Always found colony based insects but especially ants fascinating. Its as if the ant is not the organism but rather the colony is.

senthil_rajasek 2y ago
If you find Ants interesting you would probably like E.O. Wilson's works. That's all he did his entire life, study ants. https://en.m.wikipedia.org/wiki/Journey_to_the_Ants https://en.m.wikipedia.org/wiki/Journey_to_the_Ants
- SeanAnderson 2y ago
  E.O. Wilson's published work was read by Will Wright, creator of SimCity / The Sims / SimAnt, and the material is what inspired him to create SimAnt :) Will has also stated that the pheromone system designed for SimAnt was repurposed to be used in The Sims! Very interesting stuff if you ask me :)
- withants 2y ago
  Thank you. What a life!
jsenn 2y ago
Deborah Gordon (ant biologist) makes this argument in her book Ant Encounters. Because the ants in a colony are all sisters, the unit of reproduction is actually the colony rather than the individual ant. Great book!
- kevinventullo 2y ago
  The unit of reproduction, and thus the unit of Darwinian evolution!
- wonderwonder 2y ago
  Thanks for the rec, just ordered it. Its part of a series on complex systems, pretty cool
EdwardCoffin 2y ago
"A colony of ants is more than just an aggregate of insects that are living together. One ant is no ant." - E. O. Wilson
munificent 2y ago
An interesting follow-up thought is to what degree that is also true of humans and their tribes.
thowawatp302 2y ago
Have you extended that idea to the perspective of your cells and you, and thought up any major differences? (Genuine question, the biggest difference I can think of is most of our cells are in permanent physical contact.)
- wonderwonder 2y ago
  Actually yes, its an interesting thought. Not only are we a cooperative group of independent cells but those cells are fully replaced as we age so not even the original cells we start with. We are essentially walking cities of cells and somehow we gain consciousness. Pretty wild.
  - KineticLensman 2y ago
    > but those cells are fully replaced as we age Not nerve cells. Our post-childhood adult brains persist until we die, so the city of cells (nice image BTW) has a permanent and persistent ruling council, as it were, even if the citizens come and go.

digging 2y ago

It's a pretty cool finding, but a shallow article and arguably clickbait. > ... it’s very possible there’s more governing army ant behavior than two simple rules. Then don't say otherwise in the title!

EdwardCoffin 2y ago
I read the article. I think that though it is relatively short, it isn't shallow, and certainly isn't clickbait. They also point to the paper the article draws from if you want more. >> ... it’s very possible there’s more governing army ant behavior than two simple rules. >Then don't say otherwise in the title! The title constrains the rules to those dictating bridge building, whereas the quoted caveat above is more general.
beerandt 2y ago
Yea they don't really define stampede, let alone the subsidence/ threshold to break-off. That's not a simple rule. Is it traffic, weight, vibrations? An absolute threshold for decrease, or just a rate drop? Which made me think- maybe a second rule exists, but it probably isn't necessary from an 'algorithm works' standpoint. If an ant starts marching again once it no longer has another on top of it- once the supply of ants starts to dwindle (too many tied up in bridge structure), the bridge will start to deconstruct itself on its own. It's a status toggled by other elements having the same boolean column. If too many are 'busy' as bridges, that leaves too few to hold the toggle buttons of the remaining others. It's not a bridge, it's a self-sorting, self-clearing traffic jam. You drive when you can. Which as a civil engineer the traffic aspect is more interesting to me than the bridge aspect. But probably would get fewer clicks.

notfed 2y ago

"Wow, this article totally reminds me of Stephen Wolfram's work! It's amazing how these ants are following simple rules to create such complex behaviors. This is exactly what Wolfram has been talking about for years with his idea of cellular automata and complexity theory. I remember reading about how he pioneered the whole idea that simple rules can lead to complex systems. It feels like everything in nature is just proving him right, from the patterns in seashells to how these ants behave. I bet if we looked deeper, we'd find that Wolfram's principles are at the core of so many natural processes. It's incredible to see his influence everywhere!" I generated the above paragraph for satire reasons using ChatGPT.

all_these_years 2y ago

The second video seems to have some issues: https://www.youtube.com/watch?v=KJUk5VDCKbQ https://www.youtube.com/watch?v=KJUk5VDCKbQ

cen4 2y ago

Besides simple rules the other requirement is Altruism. Everyone in the colony takes care of the children of one queen. Can you imagine the chimp troupe doing any such thing? Siblings will kill each other and their kids rather than do that. People think its related to Ants being haplodiploid. Altruism in large social groups is something chimps haven't worked out yet. Which is why we have the MIL complex and war. So if you are an algo engineer check out Haplodiploidy.

SeanAnderson 2y ago
There are ant colonies with multiple queens :) But yes your general point still stands. https://en.wikipedia.org/wiki/Pleometrosis https://en.wikipedia.org/wiki/Pleometrosis
Nursie 2y ago
> Everyone in the colony takes care of the children of one queen. Depending on how you look at it, she is their prisoner and the only way they get to pass on a proportion of their genetic material is by looking after her kids that share some of it. Alternatively you can look at them as little automata largely driven by pheromones. Or you could see them as part of a larger pseudo-organism that doesn't really have individual wants and drives in the way we understand them. Fascinating either way though.
daniel-s 2y ago
This is really cynical. You live in a society where almost everyone goes to work to produce something that they will not directly consume. Humans figured out a way to organise everyone and get them to work cooperatively, it's the free market. In fact, it couldn't work any other way. Suppose you had 100% altruism and everyone wanted to promote the "greater good". You wouldn't ever know how to direct people towards that greater good. You need prices to drive production.
- M95D 2y ago
  I doubt that free market drives people toward "greater good". What is "greater good" anyway? Can you give some examples of "greater good" and how the free market achieved it? BTW, no market is truly free, not unless it allows fakes, quack medicine, child labor, selling humans, or other horrors of the past.

xhevahir 2y ago

"Plant arithmetic" is another interesting phenomenon: https://en.wikipedia.org/wiki/Plant_arithmetic?wprov=sfla1 https://en.wikipedia.org/wiki/Plant_arithmetic?wprov=sfla1