The "ant algorithm" has been used for a long time to solve "shortest path" type problems, it's not really intelligence just evolved behavior. Granted there is probably a fuzzy line between the two at some point.
I once mentioned using ACO to a former (mostly dumb, non-technical) manager and was nearly laughed out of the room. Her misapprehension was that to solve the particular problem at hand, which I think was some routing issue for support crews in a utility company, that we'd use actual ants!
("ants as fluids" linked at the top of that page grabbed my attention a few years ago; myrmycology meets physics, rather than "just" the robotics angle - see again a few paragraphs further below)
An article about Nathan Mlot's work, vintage 2011:
Their constructions are individually incredible, and there are so many of them. ("... in northeast Brazil ... about 200 million termite mounts spread over an area the size of Great Britain. Some of the mounds are 3m (10 ft) tall and 10 m (33 ft) wide and they are spaced about 20m (66 ft) apart. Underneath the mounds are networks of tunnels ... radioactive dating on 11 mounds ... youngest mound was 690 years old. The oldest was at least 3,820 years and possibly more than twice that").
If we could sensibly communicate, I wonder what they would make of human skyscrapers, and whether they think of themselves as living in arcologies.
Sadly, they have an enemy that makes Godzilla seem tame. Thirty-five thousand "urban" victims a day per giant anteater: https://www.bbc.co.uk/programmes/p00v0vc4 (this is a great video for seeing a number of ant mounds that are "only" about 150cm tall).
The horrifying things about fire ant rafts is that in floods, like after hurricanes and tropical storms on the US gulf coast, whole colonies will raft in the floodwaters. Now think of all the pictures you've seen of people wading through post-storm floods.
I really want to figure out how to have ants form bridges/towers to navigate gaps. It's on the roadmap for long-term goals, but there's a lot more to go in the short-term :) Above ground view where the ants leave the nest and collect food using pheromone trails is the next major feature coming up!
If you want to talk shop about ants, or help me tinker with code, discord link is in my profile :)
>I'm also trying to program a digital ant farm: https://ant.care/
Totally off topic, but your domain reminded me of: https://antville.org JS on the server-side (On JVM since the late 90's)
My favorite ant behavior is that many tree-dwelling species will actively glide back towards the trunk if they're dislodged so they're less likely to depart the tree entirely. There's a few papers on the topic, this article has some videos :) https://insider.si.edu/2011/06/lofty-experiments-with-glidin...
I didn't see it when I searched, so apologies if I overlooked it.
Why don't autonomous vehicles use swarm intelligence? Why are the companies trying to make a sufficiently advanced AI for each car, isolated from others, instead of https://en.wikipedia.org/wiki/Swarm_robotics ?
Also bicycles, motorbikes, e-scooters, trams, trains, runaway objects...
I always thought that the city itself could be part of the swarm system. Roadside sensors could broadcast information to all relevant participants - such as "emergency vehicle incoming, move to the right lane, vehicles ID 123, 456 and 789 stop immediately" or "danger - static object at 45th km", or even "slow down/speed up to 60 km/h to optimize traffic flow".
In the science fiction of my youth, where roads or similar systems existed, there were cases where the infrastructure was part of the autonomous vehicle system. It was often car-exclusive, never mentioning pedestrians, mobility devices, or anything that wasn't motorized, but it was an interesting bit of speculation.
I think to some degree it will play a part, esp. once most of the vehicles have autonomous functionality. But without that homogeny and having to deal with many outside variables it's not enough.
There's a huge boon to be had esp. with traffic shaping.
This kind of behavior has resulted in an amazing class of “Swarm Intelligence” algorithms that somehow never gained widespread adoption. The classic SFI reference that I invested some time working through a few years ago is: https://amzn.to/3GdKJ9o
So a brain is "just" a bunch of neurons communicating, and while no single neuron is intelligent, the whole is intelligent. And ants exhibit somewhat intelligent behaviour as a whole, while none of them are themselves something we consider intelligent.
I think this is super interesting and we can go completely off the rails taking this to its logical extremes. Like, the entire planet earth is an intelligence that arises from all interactions between all living things on the planet. In this sense, the speed of thought must be considered extremely slow, with even the most minute decision taking decades if not centuries.
There is a theory of consciences (using theory very lightly) that says that conscience is an inherent property of the universe (like gravity) and that any and all systems has conscience. Most of these are not recognizable by conscience humans (although we all seem to agree that we have a sub-conscience, so we are open to the concept of different types of consciences).
"Waking up: a guide to spirituality without religion" has a good chapter on this, where he starts with how people who've had their corpus callosum cut, and through experiments we can see that the two halves work independent, but the person is unaware of this. Their experience is still one conscience, even though their do things or know things that is not apparent to the conscience that outsiders generally would communicate to. He then goes on to argue how there can be multiple consciences that are unaware that they are "trapped" and not in complete control.
The author is not the only one to bring out this argument, but I thought it was well put in that book.
I really like this theory, while at the same time considering pretty far out.
And going the other direction: I don't expect my individual brain cells to feel conscious, but if they do, they're probably not aware that they also together form consciousness.
The fact that the universe exists is pretty weird; that it is (at least partially) self-aware is either far more strange, or completely... natural. Perhaps even inevitable for sufficiently complex closed systems with stabilizing forces.
If you're talking about collective intelligence, particularly about intelligence emerging from the colaboration of simple entities working together to form larger structures that feed back on the simple entities, you're awfully close to Hofstadter's Strange Loops.
38 comments
[ 4.9 ms ] story [ 87.8 ms ] threadhttps://en.wikipedia.org/wiki/Ant_colony_optimization_algori...
If we find a sharp definition of such a line we will know there is truly a difference. Personally though my bet is there is no such line.
https://antlab.gatech.edu/antlab/The_Ant_Raft.html
("ants as fluids" linked at the top of that page grabbed my attention a few years ago; myrmycology meets physics, rather than "just" the robotics angle - see again a few paragraphs further below)
An article about Nathan Mlot's work, vintage 2011:
https://www.nationalgeographic.com/science/article/fire-ants...
(Outcomes for the ants are sadly not always great in Hu's lab at GATech, e.g. the ones who died in a CT scanner: <https://www.nature.com/articles/nature.2014.15400>).
Some 2021 CU Boulder work:
https://www.colorado.edu/today/2022/03/02/fire-ant-rafts
(and their "Emergent Behavior in Fire Ants", <https://www.youtube.com/watch?v=IrLc-uDv7GU>)
Another recent study of ant raft dynamics from a physics perspective:
https://royalsocietypublishing.org/doi/10.1098/rsif.2021.021... (Also, ants are small: "To conduct experiments, we collected 3–10 g of worker ants (or approx. 3000–10 000 ants)")
Another interesting but somewhat sadder behaviour in ants is the
https://en.wikipedia.org/wiki/Ant_mill
Also really cool are the structures made by burrowing ants <https://en.wikipedia.org/wiki/Ant_colony#/media/File:Walter_...>, but the eusocial insects who are the builders most interesting (in a casual sense) to me are https://en.wikipedia.org/wiki/Mound-building_termites
Their constructions are individually incredible, and there are so many of them. ("... in northeast Brazil ... about 200 million termite mounts spread over an area the size of Great Britain. Some of the mounds are 3m (10 ft) tall and 10 m (33 ft) wide and they are spaced about 20m (66 ft) apart. Underneath the mounds are networks of tunnels ... radioactive dating on 11 mounds ... youngest mound was 690 years old. The oldest was at least 3,820 years and possibly more than twice that").
If we could sensibly communicate, I wonder what they would make of human skyscrapers, and whether they think of themselves as living in arcologies.
Sadly, they have an enemy that makes Godzilla seem tame. Thirty-five thousand "urban" victims a day per giant anteater: https://www.bbc.co.uk/programmes/p00v0vc4 (this is a great video for seeing a number of ant mounds that are "only" about 150cm tall).
I'm also trying to program a digital ant farm: https://ant.care/ (code: https://github.com/MeoMix/symbiants/)
I really want to figure out how to have ants form bridges/towers to navigate gaps. It's on the roadmap for long-term goals, but there's a lot more to go in the short-term :) Above ground view where the ants leave the nest and collect food using pheromone trails is the next major feature coming up!
If you want to talk shop about ants, or help me tinker with code, discord link is in my profile :)
this is really cool!
I didn't see it when I searched, so apologies if I overlooked it.
https://en.wikipedia.org/wiki/Swarm_intelligence
I always thought that the city itself could be part of the swarm system. Roadside sensors could broadcast information to all relevant participants - such as "emergency vehicle incoming, move to the right lane, vehicles ID 123, 456 and 789 stop immediately" or "danger - static object at 45th km", or even "slow down/speed up to 60 km/h to optimize traffic flow".
In what way does introducing swarm intelligence preclude coping?
There's a huge boon to be had esp. with traffic shaping.
I think this is super interesting and we can go completely off the rails taking this to its logical extremes. Like, the entire planet earth is an intelligence that arises from all interactions between all living things on the planet. In this sense, the speed of thought must be considered extremely slow, with even the most minute decision taking decades if not centuries.
There is a theory of consciences (using theory very lightly) that says that conscience is an inherent property of the universe (like gravity) and that any and all systems has conscience. Most of these are not recognizable by conscience humans (although we all seem to agree that we have a sub-conscience, so we are open to the concept of different types of consciences).
"Waking up: a guide to spirituality without religion" has a good chapter on this, where he starts with how people who've had their corpus callosum cut, and through experiments we can see that the two halves work independent, but the person is unaware of this. Their experience is still one conscience, even though their do things or know things that is not apparent to the conscience that outsiders generally would communicate to. He then goes on to argue how there can be multiple consciences that are unaware that they are "trapped" and not in complete control.
The author is not the only one to bring out this argument, but I thought it was well put in that book.
I really like this theory, while at the same time considering pretty far out.
I like the thought experiment though!
The fact that the universe exists is pretty weird; that it is (at least partially) self-aware is either far more strange, or completely... natural. Perhaps even inevitable for sufficiently complex closed systems with stabilizing forces.
Thanks for the book recommendation.
I'm disappointed.