Slashdot Mirror
Incorporating Swarm Intelligence Into Computer AI
An anonymous reader writes "From optimizing truck delivery routes to inspiring nerve-cell-based cognition models, ant intelligence has arrived. From the Economist: 'In 1992 Dr. Dorigo and his group began developing Ant Colony Optimisation (ACO), an algorithm that looks for solutions to a problem by simulating a group of ants wandering over an area and laying down pheromones. ACO proved good at solving travelling-salesman-type problems. Since then it has grown into a whole family of algorithms, which have been applied to many practical questions. ... Ant-like algorithms have also been applied to the problem of routing information through communication networks. Dr. Dorigo and Gianni Di Caro, another researcher at IDSIA, have developed AntNet, a routing protocol in which packets of information hop from node to node, leaving a trace that signals the "quality" of their trip as they do so. Other packets sniff the trails thus created and choose accordingly. In computer simulations and tests on small-scale networks, AntNet has been shown to outperform existing routing protocols."
does it scale?
The Government can do it better.
Some ants got into my kitchen the other day walked around in a circle and eventually died (the ones I didnt manage to kill) hours later
sounds like a excellent plan
Terry Pratchett got there first.
Anything you can do, HEX can do better!
!! -Imagine a beowulf cluster of thos--*gunshot*
It just sounds like the classic hill climbing algorithm to me.
"I bless every day that I continue to live, for every day is pure profit."
That would be the routing table from hell
I split in their general direction.
Who/what the fuck are they, Mr. Summary Writer?
Until it passed by a group of people having a picnic.
Swarm logic would be wonderful as a routing protocol, though would prevent protocols like UDP from ever getting packets through in any sort of decent order. though the system would be wonderful for many protocols, anything with ordered sequential data streams would see little to no benefit.
Behavioural/colonial/pheromonal definitions are the key to performance in this type of approaches.
Fine-tuning these definitions can be quite a manual task.
Yes, the rest is done by "ants" but still you have to do the real job in the beginning.
e.g. Without "turn at cross-roads with degree of 45" definition, it doesn't work well for even food gathering.
Dorigo et al. made their groundbreaking paper in 1996 (based on observations by Deneubourg in 89), and then nothing. Nothing new on the theoretical part, no new application. Even routing using ACO like algorithm has been published as early as 1994. The newest extension based on these algorithms is an interactive distributed image retrieval system by Picard et al. back in about 2006 (as far as I know). So nothing new nor groundbreaking here.
And with that, the seeds to build SkyNet are sown. :P
How is "AntNet, a routing protocol in which packets of information hop from node to node, leaving a trace that signals the "quality" of their trip as they do so..." any different from bog standard hop count updates on existing routers, and routing on the basis of the shortest path?
I think the authors are playing semantic games here, not doing research.
" I welcome our insect swarm AI overlords"
May be quite appropriate here
I just covered Ant-based load balancing on communications networks in a distributed systems class. Here's the paper we read. It's an easy read, and quite interesting.
http://www.hpl.hp.com/techreports/96/HPL-96-76.pdf
Uhuh, would you really want "Swarm A.I." to control your fleet of trucks? Imagine if you're the clerk at a warehouse just cooling your heels the whole day. You look at the clock and it is 4:55 pm when suddenly 10 delivery trucks Zerg rushed into the parking lot.
Kinda sounds like an adaptive traceroute. Perhaps traceroute was antlike before it became in vogue :)
Uh, Linux geek since 1999.
How do you incorporate Swarm intelligence into computer AI? Simple - SPAWN MORE OVERLORDS!
"The tree of liberty must be refreshed from time to time with the blood of patriots and tyrants." ~Thomas Jefferson
You mean to say that this thing is measured as outperforming non-statistical recursive routing methods?
"good at solving" - define "good" here, and how does it compare to a directed algorithm
"good at thinking up solutions" - wait, this is different. now they're not solving, they're thinking-up solutions. instead of being dumb actors with rule-based behaviors and reacting to external stimuli, they're now using their own internal models to plan how they will route themselves. (no, it isn't, i'm being sarcastic, the author just used a bad metaphor that inadvertently inverted the whole premise of the experiment.)
"working on something that can act as well as think" - oops. and the bad metaphor does a double gainer with a full twist. they weren't thinking, they were acting and solutions fell out of their actions. so you need something that can think as well as act, not just act and hand you its results which you misinterpret as thinking. and you need a better way to say you need something physical instead of just a simulation.
"there are those who think that, far from being an illusion of intelligence, what Dr Dorigo and his fellows have stumbled across may be a good analogue of the process that underlies the real thing." - it's the "there are those that" part that allows you to suggest any fantastical nonsense in the space between it and the period. And then you run off to construct the fantasy:
"the way bees select nesting sites is strikingly like what happens in the brain...explore an area...return to the nest and perform a waggle dance...Substitute nerve cells for bees...electric activity for waggle dances...good description of what happens when a stimulus produces a response..." - i think my brain just asploded. there's a chance that you can stretch this metaphor until it covers training of a neuron, but it's the same chance you have of stretching your rubber underpants until it encompasses your cubicle. but the feedback-induced avalanche in bees is nothing at all like the one-way dendritic stimulus producing a cascade-avalanche response in the soma of a neuron.
"Those who speak of intellectual buzz, then, might be using a metaphor which is more apt" - you have bees in your bonnet, mr. anonymous roman TFA author. that's the most apt metaphor anywhere near your essay.
>format /dev/fd0
>ANT: We must save the queen!
>ANT: Which one of us is the queen?
>ANT: I'm the queen!
>ANT: No, I'm the queen!
(smashing of glass sound)
>ANT: Freedom! Horrible, horrible freedom!
But how about when AntNet isn't forwarding packets (workers and soldiers) to sugar.com, honey.com, catBarf.com, or enemyAntNest.com?
rosetta stone rosetta stone rosetta stone language rosetta stone language rosetta stone spanish rosetta stone spanish abercrombie and fitch abercrombie and fitch Abercrombie Fitch Abercrombie Fitch Abercrombie Clothing Abercrombie Clothing pandora pandora pandora schmuck pandora schmuck pandora armband pandora armband tiffany tiffany tiffany jewellery tiffany jewellery tiffany rings tiffany rings
I wrote a program two years ago that used ACO to solve the TSP. That crap was the fastest thing in the west.
That may sound moderately well and good on the surface to most people, it spells a nightmare for the network architects and engineers. How do you troubleshoot traffic flow problems? I don't think I'll ever hear a network engineer try to track down a problem by asking "well did you follow the pheromone packet trail?!" What it boils down to are the people who build a network want nice, predictable traffic flows; whether OSPF/EIGRP/BGP dynamically picks them or not, there's still a layer of predictability once the initial traffic paths are determined.
Why are we reposting an economist article? I would think SlashDot could come up with an article that goes somewhat deeper than this.
Can I join in with your sentiment, and just say, WTF?
It's like Slashdot has just discovered early 1990s AI research and confused it with modern news. ACO has been around for so long now that it's a standard tool in the toolbox of the developer who has a basic grounding in AI along with genetic algorithms and particle swarm optimisation. As the AC says, it's already been used in routing protocols.
Why is Slashdot reporting this? What is the story meant to be here exactly? Even TFA just seems to be a very very brief history of this technique following a mention of it.
Still, if anyone finds it interesting then may I suggest you also look at the above mentioned techniques too- genetic algorithms, particle swarm optimisation, neural networks and so forth? None of this is new by a long shot, but it's all part of the larger field of AI, and specifically they all focus on the core principle of achieving order through emergence and self-organisation. In fact, if you really want to delve into it then learning all about emergence and emergent processes is in itself worthwhile.
I should note that ants are one of many species that exhibit this kind of behaviour- bee colonies work in much the same way for example in producing hives and such.
But again, I know Slashdot is slow when it comes to posting news, but being about 15 years give or take a few late- that's a new record even for Slashdot right?