Hacker News new | past | comments | ask | show | jobs | submit login
Ants swarm like brains think (nautil.us)
144 points by dang on Apr 28, 2014 | hide | past | web | favorite | 39 comments

There's an appealing idea that the core of human intelligence is one general purpose algorithm. I like thinking that there is something fundamental, simple and elegant waiting to be discovered like evolution by natural selection or Newtons laws (of mechanics).

A beautiful thing about Evolution by Natural Selection is that the principle itself is metaphysical, not physical. It exists in the same realm as math or logic. The principle is embodied in biology just like the concept of a square can be embodied as a planet. Even if we didn't have any physical squares our metaphysical ideas about metaphysical squares would still valid, hence n-dimensional geometry.

Another cool thing about Evolution by Natural Selection is that it is nicely self contained. It's the bottom of a finite rabbit hole. Understanding what evolution is made of will help us understand a particular instance of evolution, but not the principle. So, knowing about DNA and genetic mutations helps us understand evolution in biology but it doesn't tell us more about evolution in abstract. IE, it doesn't get us to a more fundamental understanding. Even if Darwin had been wrong about the origin of species he would still have discovered something awesome.

Newtons principles have a different kind of elegance. They are extremely fundamental and a pleasingly small number of them are needed. The difference is that they are very much physical laws. They are fundamental enough to be treated as fundamental rules of the universe. Almost. Newtonic physics still leaves open the idea of more fundamental questions and answers. Learning what gravity is made of does get us closer to the fundamental.

Darwin's evolution is fundamental in the way that math is. Newton's mechanics is fundamental in the way that atoms are. The way I have these defined in my own mind is Philosophical and Scientific. It's a sort of fuzzy distinction and it breaks down a lot. But it seems useful to me sometimes. "Anthropic," for example, is something you need to get your philosophical mind familiar with.

Back to AI. The single algorithm idea hints, I think a "philosophical" explanation. We're looking for some metaphysical property of the universe that intelligence is an instance of.

Upvote for your beautiful formulation and analogy.

What you are referring to was coined by Kant as “transcendental”, as opposed to “transcendent” — there’s theological/mythological/naive metaphysics and there’s abstract/logical metaphysics. The respective practitioners of those are a very different kind of believers. Unfamiliar with that essential difference, yet another party (mainstream scientocracy) is strongly averse to any kind of metaphysics and denies everything that is not touchable matter. They look at cloning the brain and will have difficulty to find the underlying structure of the mind.


I didn't remember the Kant connection (I can't seem to get to the end of a paragraph without forgetting the beginning). My definition is a kind of blurry augmentation of Physics & Metaphysics. Scientific or Philosophical aren't perfect synonyms for these but I think they give us a little more of a hint as to what kind of process to employ when thinking about them.

I think modern usage of the term dates to Aristotle's books: Physics & Metaphysics. 'Physics' is about nature and feels like a direct ancestor to modern science. 'Metaphysics' is pretty close to what we still call metaphysics. Most people think of these things (logic, ethics, epistemology as science).

Physical and metaphysical are the better terms. More accurate. But, I just don't like the associations most people have with them. Physics is impenetrable domain of extreme experts. Metaphysical doesn't have any expectation of rigor and every answer is equally valid. There are other problems like thinking of Metaphysics as completely separate and not impacting on physics like the mind/brain issue you discuss. I think to a modern, rigorous thinker the idea that the mind being a metaphysical concept which is embodies in a brain makes sense. It doesn't imply the kind of dualistic worldview usually associated with spirit and matter.

> I think modern usage of the term dates to Aristotle's books: Physics & Metaphysics. 'Physics' is about nature and feels like a direct ancestor to modern science. 'Metaphysics' is pretty close to what we still call metaphysics. Most people think of these things (logic, ethics, epistemology as science).

Tangential to all this, but a fun apropos: it's interesting that the original title of Aristotle's (multiple 'books' collectively known as) 'Metaphysics' is (widely believed to be) a matter of noting down the arrangement/order of his treatises: τὰ μετὰ τὰ φυσικά (ta meta ta fysika) - '[writings] after the Physics' (Physics being another arrangement of prior treatises of his.)

'What comes after Physics' can of course be evaluated conceptually, but many scholars think that 'τὰ μετὰ τὰ φυσικά' here means 'writings that come after the Physics'.

Which is a somewhat humbling (if that's the word) idea, and which might be able to reduce the aura of drama and mystique oft associated with the word of 'metaphysics'. :)

What I find especially interesting is why we think that these metaphysical concepts make sense. We do not only seem to think 'it makes sense', there seems to be an actual _emotion_ connected to it. We like it!

I mean, we look at evolution and at some point it 'just clicks', we wrap our head around it and see the abstraction clearly everywhere. And we like that we understand.

But we do we like it? Why do we like the explanation?

It also works the other way around: For example, look at the concept of the infinite regress. When one comes up, we actually seem to dislike the ideas that lead to it. Emotionally. There is no logical/rational reason for the dislike. Why so?

We have come to a lot of the same conclusions, particularly w.r.t. evolution being a metaphysical phenomenon and its being deeply interwoven with (or perhaps analogous to) intelligence.

Do you have any favorite authors or texts in this area?

Daniel Dennett expresses almost that exact idea (about evolution being abstract principle) in his book Darwin's Dangerous Idea:


This book isn't directly related to all of OP's comment, but it addresses emergent, meta-properties of systems---which seems to me to be an important component of OP's (outstanding) comment. The book is an exploration of complex behaviours and properties that arise from seemingly simple actors, actions, and individual characteristics taken in aggregate; essentially, the dynamics and structures of meta-properties/behaviours in systems.

More directly, the book also addresses examples of swarm dynamics and intelligence.

"Self Organization in Biological Systems"


Not really. Richard Dawkins turned me on to the idea.

Interestingly, the whole "intelligent design" debate is a good proxy for 'evolution as intelligence' arguments. One side offers example of how biology/nature is 'designed' and the other finds explanations for how it is evolved. Evolution does seem intelligently designed. It exhibits problem solving.

The similarity between the way an intelligent human thinks and the way nature "thinks" is striking. It's striking to the point that simply defining nature as an intelligence (or even going further as defining it as an intelligent creator) doesn't seem out of the question. We know a lot about the mechanics of evolution. Less about the mechanics of human intelligence. In the absence of a better understanding of the latter, it's more or less just a semantic choice.

But, there could very well be a stronger connection. Maybe something similar to evolutionary intelligence powers human intelligence. Maybe both are instances of a more fundamental thing. I'm using 'emergence' as the placeholder in my brain for that fundamental principle/algorithm or collection of these. 'Emergence' also covers concepts from economics, culture and ecologies and finds logical solutions for things without the intervention of an intelligent human.

But no, I don't have good reading recommendations. If anyone does, I would love to hear them. Fiction, non-fiction. I would love to hear about these ideas from different perspectives.

Well, you already mentioned Dawkins, but to anyone else reading this, I can't recommend The Selfish Gene enough. It dramatically changed the way I understand nature, and with the chapter on memes, the way I understand culture. I hear The Extended Phenotype, again by Dawkins, is also excellent, but regrettably, I haven't read it yet.

I also highly recommend Dan Dennett. He covers these topics more directly. Intuition Pumps and Other Tools for Thinking covers his thoughts on consciousness and evolution in a very direct and superbly clear manner.

Dennett also co-edited an anthology of fiction and non-fiction called The Mind's I with Douglas Hofstadter (of Godel, Escher, Bach fame). It was one of the most interesting and enjoyable books I have ever read. It covers a very wide variety of topics surrounding the editors' theory of the mind, including the classic dualist/materialist debate (the editors are distinctly materialists), evolution, the mechanics at the core of consciousness, and so much more. It's another one that I can't recommend enough.

I've been thinking about this a bit lately. I came to the conclusion that brain could basically be a process to simulate evolution (not computational, but analogous mimicry). If something seemingly 'intelligent' can exist because of evolutive process ('emergence'), brain should mimic the process to arrive at the same solution, faster. For example, when I'm thinking about this comment I'm unknowingly eliminating all the pathways that lead to bad comments and this is the only one that survives ('emerges').

Why 'mimic?' IE, if this is correct then biological evolutions mimics intelligence just as much as intelligence mimics biology.

>A beautiful thing about Evolution by Natural Selection is that the principle itself is metaphysical, not physical. It exists in the same realm as math or logic.

Evidence of this is genetic algorithms. They would work even if darwin was completely wrong. Physical evolution just happens to follow "metaphysical" evolution.


Though I think people if you are not thinking of evolution as an abstract thing already, you will categorize genetic algorithms as something that mimics evolutions rather than expressing or embodying it.

Richard Dawkins made this click for me with his concept of memes.

There is also an idea for universes evolution. It builds on (and flavours) the ideas for an anthropic universe. A Universe spawns child universes (in a black hole?) and these inherit most but not all (mutation) of the fundamental aspects of the parent universe.

If you want to get very wacky and go into 'simulated universe' speculations (I do!), evolution can play a role here too. If we were to create our own simulated universe, it would probably be some variation of our own.

The mechanisms of evolution have also been attributed to the development of different religions. One guess to the reason why there are so few religions with no obvious reward system is that it's more difficult to convert people to these, compared to major religions like Christianity and Islam where there are great, clearly defined rewards and punishments.

I guess, at one point in time, there were a lot more religions with neither reward or punishment system in place or just either one of these, and adopting more feedback systems to a religion at the time greatly boosted its breadth and penetration across populations.

I don't remember where I heard this, but I wouldn't be surprised if it was true.

It's very profound to see the math behind both events. I view this in line with finding Fibonacci's sequence all over the place - a core idea operates in multiple places.

For some reason I find ants fascinating - perhaps because I've lived in a few houses that were infested? Now what would truly be exciting is if we found symptoms or clues of the greater hive achieving consciousness-like behavior. I suspect the speed of ant reactions is too slow, but we're allowed to dream, right?

What are your thoughts on the evolution of evolution itself?

(the DNA/RNA/Ribosome complex has too many parts to have sprung into being just like that).

I highly recommend reading Chapter 2 of "The Selfish Gene", titled "The Replicators". It is possibly my favorite book chapter of all time. I'm sorry that I can't link you a copy of the entire chapter due to copyright. Here's a tiny excerpt:

“At some point a particularly remarkable molecule was formed by accident. We will call it the Replicator. It may not necessarily have been the biggest or the most complex molecule around, but it had the extraordinary property of being able to create copies of itself. This may seem a very unlikely sort of accident to happen. So it was. It was exceedingly improbable. In the lifetime of a man, things that are that improbable can be treated for practical purposes as impossible. That is why you will never win a big prize on the football pools. But in our human estimates of what is probable and what is not, we are not used to dealing in hundreds of millions of years. If you filled in pools coupons every week for a hundred million years you would very likely win several jackpots.”

“Actually a molecule that makes copies of itself is not as difficult to imagine as it seems at first, and it only had to arise once. Think of the replicator as a mould or template. Imagine it as a large molecule consisting of a complex chain of various sorts of building block molecules. The small building blocks were abundantly available in the soup surrounding the replicator. Now suppose that each building block has an affinity for its own kind. Then whenever a building block from out in the soup lands up next to a part of the replicator for which it has an affinity, it will tend to stick there. The building blocks that attach themselves in this way will automatically be arranged in a sequence that mimics that of the replicator itself. It is easy then to think of them joining up to form a stable chain just as in the formation of the original replicator. This process could continue as a progressive stacking up, layer upon layer. This is how crystals are formed. On the other hand, the two chains might split apart, in which case we have two replicators, each of which can go on to make further copies.

A more complex possibility is that each building block has affinity not for its own kind, but reciprocally for one particular other kind. Then the replicator would act as a template not for an identical copy, but for a kind of ‘negative’, which would in its turn remake an exact copy of the original positive. For our purposes it does not matter whether the original replication process was positive-negative or positive-positive, though it is worth remarking that the modern equivalents of the first replicator, the DNA molecules, use positive-negative replication. What does matter is that suddenly a new kind of ‘stability’ came into the world.”

Just the other day I read something about Cambridge researchers saying metabolic processes existed before life as we know it... here it is:


I don't have anything particularly insightful to share except that it has occurred to me and that thinking about it seems like a deep rabbit hole. It's probably a key to figuring out how it all got started in the first place.

DNA does seem too complex to have sprung into being fully formed. Also, DNA & RNA probably share a lineage which suggests they the are evolved.

The basic rules of the evolution game in the plant and animal kingdoms seem to be genes encoded in DNA passed through sexual reproduction. These are the mechanisms we think of when we think of evolution. Darwin didn't know about genes or DNA so he used "organisms" as the basic unit and that seems to work just as well. This is a pretty stable long term system. It seems to be the main way organisms evolve.

But, these aren't the only rules/mechanisms of biological evolution. For a start, it doesn't explain how these rules got written. We know that there are other mechanisms for inheritance of genetic traits, "epigenetic" seems to be the general word for them. Richard Dawkins describes "memes" as units of culture which evolve in societies. Humans are organisms.

The rules and mechanisms that govern evolution are certainly subject to change and there are various mechanisms, perhaps competing. For example, right now there is a transmittable cancer tearing through the Tasmanian Devil population. The cancerous cells are transmitted when devils bite each other. The cancer itself can be considered an organism, like a virus or a bacteria. It has cells and reproduction and genes. It behaves similarly to a bacteria or virus but it's a descendant of a tasmanian devil (or a devil's cell).

So we have an example of an leap (or fall). The basic rules got broken in a way that. The evolutionary unit (tasmanian devil) was ignored in favor of a new unit (cell). Most the rules that apply to the previous unit (eg sexual reproduction) no longer apply. It's a reminder that the genes themselves that are "trying" to reproduce. The cell is a vehicle for the gene. The organism is a vehicle for the cell.

Did "Devil Facial Tumour Disease" evolve? I'm not sure we can call it evolution. But, it wasn't the first cancer that had a chance to become its own organism. Every cancerous mutation is an opportunity for a transmittable cancer to pop into existence. Given enough time, maybe it evolve into Maybe it will develop organs and limbs and fur and then maybe it will develop its own parasitic cancer: generation 2.

Sexual reproduction is itself a mutation that probably occurred several times and it changes the rules. Epigenetics seems to be in the zetgeist at the moment, I hear the term mentioned a lot.

I don't know how to tie these thoughts together.

Douglas Adams is my go to for these kinds of issues:

“Anything that happens, happens. Anything that, in happening, causes something else to happen, causes something else to happen. Anything that, in happening, causes itself to happen again, happens again. It doesn’t necessarily do it in chronological order, though.”

Like I started earlier on, I my instinct is that evolution the principle is best explored outside of biology. Biology is a magnificent example of evolution, but it's an embodiment of evolution.

"Anything that, in happening, causes itself to happen again, happens again" seems like a good starting point.

Thank you! Makes you wonder what the minimum viable product version of DNA supporting 'life as we recognize it' was like... Maybe CG or TA only?

Douglas Hofstadter used this analogy in his 1979 book, Gödel, Escher, Bach. GEB presents an analogy about how the individual neurons of the brain coordinate to create a unified sense of a coherent mind by comparing it to the social organization displayed in a colony of ants. It's a fascinating book, and well worth the read for the inquiring mind.

This reminds me of how some biologists consider a colony to actually be a single organism, and each individual ant an organ system. The queen could be thought of as the reproductive organs, workers the limbs, scouts sensory organs. It's especially interesting considering that all of the workers are sterile, so fitness is shared, and dependent on the queen.

> It's especially interesting considering that all of the workers are sterile, so fitness is shared, and dependent on the queen.

That's why it makes sense to think of the colony as a single organism. In the same way, your hands can only reproduce when the rest of you does, in an orderly manner, and we think of you as a unified organism. If hands could bud off independent mini-hands, you might see less cooperation from them over the course of your life.


On a similar, but different tangent is Marvin Minksy's Society of Mind.

"You cannot train a beaver to build a termites nest or teach termites to build beaver dams."

GEB is an amazing book, recommended to anyone into CS and logic.

The 2004 ICFP contest was based around this theme. The contestants had to program an ant state machine in a very simple and resource-restricted language. Still it allowed for pretty complex strategies like ambushing opponent's ants trying to steal your food.

Here's the link to the results which also contains links to some summaries by contestants: https://alliance.seas.upenn.edu/~plclub/cgi-bin/contest/resu...

The foraging response to food is an example of a positive feedback loop, and familiar to anyone who has had a picnic ruined by a line of ants marching in single file toward their meal.

Seeing how hundreds of ants in a long line harvest a package of cereal high upon the shelf is an impressive sight, despite the food casualty . . . :-)

Reading the article makes me want to heed the thousands of years old advise:

Go to the ant, you sluggard; consider its ways and be wise! It has no commander, no overseer or ruler, yet it stores its provisions in summer and gathers its food at harvest. (Proverbs 6:6-8, NIV)

IANAAE (I Am Not An Ants Expert), but even though colonies have queens, there is no central control from what I understand:

There are about 10,000 species of ants. They all live in colonies consisting of one or a few queens, and then all the ants you see walking around are sterile female workers. And all ant colonies have in common that there's no central control. Nobody tells anybody what to do. The queen just lays the eggs. There's no management. No ant directs the behavior of any other ant. And I try to figure out how that works. And I've been working for the past 20 years on a population of seed-eating ants in southeastern Arizona. http://www.ted.com/talks/deborah_gordon_digs_ants/transcript...

> even though colonies have queens, there is no central control

This is important. A lot of times I describe to people how ants are a perfect example of a cooperative anarchical society, and they say "Well what about the queen!!"

"Queen" is a misnomer. She is really just another ant whos job it is to lay eggs. This job tends to be less expendable than others, so the queen is more protected (from what I know), but she's still "equal" in the sense that nobody gives orders, everything is self-organizing.

All sorts of things can be modeled on ants, like distributed message passing and congestion control algorithms, as well as routing in general. There's a list of ant colony-inspired algorithms here https://en.wikipedia.org/wiki/Ant_colony_optimization_algori... And Google Scholar has a wealth of papers found here http://scholar.google.com/scholar?hl=en&q=ant+colony

Even rivers are intelligent, they find a way from the lake to the sea and carve a comfortable path for themselves. :)

Evolution and gravity are consequences of the natural computation of the universe.

"The Invincible" by Stanisław Lem comes to mind.

Spoilers inside! -> http://en.wikipedia.org/wiki/The_Invincible

They actually are. Depends on what you consider the river. If it is just the molecules of water, probably not. But if you look at the whole ecosystem of a river, you will be amazed. And in this situation with beavers building dams and changing the course, and everything doing its bit, yes, the river will do amazing things and will show to have a life of its own, just like the ant colony.

I just finished reading the paper about a model of ant foraging rates.

Observing that ant nest size should be a stochastic exponential process, depending on how many ants bring back food and how many die trying, is this model detailed enough to show that ants are performing portfolio optimization?

I want to know how ants got into a sealed bag of brown sugar when I can't.

These little investment bankers sure are sneaky!

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact