Agreed. Most of his short stories are great. The Screamers movie is typical of other PKD adaptations, though.
If you can put that aside it's actually decently entertaining early-90's SF schlock.
If you use a Mac and are in a Latex-centric field, I find Bibdesk (http://bibdesk.sourceforge.net/) really great for managing reference pdfs and use cvs or svn if I really want to manage a document I'm working on. There's no annotation in Bibdesk but you can record notes and it generates bibtex for you.
Accuracy, to an experimentalist, is a quantitative measure of closeness to the "correct" value (i.e. accuracy vs. precision).
Anyways, as the pp said, this is really just an example of poor reporting. The actual article doesn't really make this kind of claim, in fact the word "accuracy" doesn't even appear in TFA. They'd never get published in Nature (it's actually Nature Nanotechnology) saying something was "100% accurate" because it doesn't make any sense formally.
"100% Accuracy" implies a positional error of zero meters (to infinite decimal places), which is obviously not what they're talking about.
I caught that, too. But really "percent" doesn't even make sense as a unit of accuracy, does it? Unless it's fractional, in which case I'd take it to mean that if you want to make a relative move of x, you'll get something in the range (0,2x) or maybe (0.5x, 1.5x)? I mean, on the nano scale that's still kind of remarkable, but as you've pointed out it's just not what they mean./pedantic
IT is part of a business. Making IT's job harder in that business costs money. The article is making the point that there are some pretty serious cons about using laptops, and these need to be considered as part of their cost.
Having pissed-off employees who feel chained to their workstation (and consequently horribly unmotivated) can also be a pretty big cost.
They are similar rules to boids. The aim is a little different though.
Nice game. Check out glSwarm. I always thought it would be fun to make a game based on these ideas.
Re:The Rules of the Swarm... on slashdot.
on
The Rules of the Swarm
·
· Score: 2, Informative
Then there are things like, we'd like to make robot swarms and we want them to act like birds except do X instead of Y. Even if the natural behaviors are modeled it is very difficult to figure out what rules to add/change/delete to get the desired change. Or, we want the robots to do natural thing X, but we need some guarantee that they will do it with some level of accuracy and we need a kill switch in case they become self-aware. This field is wide open and is extremely interesting, because even if we're not explicitly mimicking natural phenomenon we are anyway when we have groups of things (computerized or otherwise) that have many individual components with complex and somewhat autonomous interactions.
We apply tools like nonlinear control theory and graph theory to study these kinds of "rules" with rigor, with the aim of a) designing robotic (specifically mobile sensor) networks that are bio-inspired in the way you mentioned and b) help the biologists by providing insight from our perspective.
My favorite thing about Dr. Couzin is his willingness to work with people in other disciplines - particularly the "harder" sciences. It's mentioned towards the end of the article. My advisor, Naomi Leonard, and her students have published several papers with Iain as a co-author - see http://www.princeton.edu/~naomi/ and search for "Couzin" and "Levin" on the page for a few references. Dr. Grunbaum, who is also mentioned in the article, is great with this as well (also on the page). They are both fantastic guys to work with.
Certainly no new work in artificial life simulation has been done in the last 21 years, and it's not possible that this has been applied to something new. Certainly not publishable, especially not in respected publications like Science or Nature or Nature.
I didn't RTFA, but I wonder if the guy's work is related to an old artificial life simulation called Boids. The developer of Boids was able to model the flocking behavior of birds by formulating some relatively simple rules. When I first learned about the program, it sounded really neat, and according to Wikipedia it's still used in computer graphics to model flocking behavior.
The rules are quite similar, but the treatment is different. Boids was an artificial life simulation experiment. The stuff TFA refers to is intended as a simplified model of a set of mechanistic rules to describe animal behavior. The parameter space is explored with the goal of discovering behaviors that may be argued to have analogs in actual animal groups.
Re:The Rules of the Swarm... on slashdot.
on
The Rules of the Swarm
·
· Score: 2, Interesting
The article is a popular science article, but addresses this, more interesting, question much more than the summary. They discuss some of the rules involved in specific situations (ants), and even look at "human swarms" (although that bit is a little cheesy). There is no general theory posited about how to make these rule sets though, apart from trial and error (in simulation if you can). They say that the researchers are starting to see patterns, but don't talk about what those patterns are - pity really, as that would have been very interesting.
One common set of rules is (related to boids) 1. Move away from those closest to you. 2. Move towards those farthest from you. 3. Align with those in the middle.
On the other hand, it would be erroneous to posit a common set of rules for every species. Different evolutionary pressures produce different behaviors in different situations. An interesting facet of this type of research is to see how different sets of parameters for the same set of rules can produce different behaviors, or different sequences of behaviors.
How about you don't live in an apartment, or if you do, suffer the consequences instead of being a whiny bitch?
Could you make a more plutocratic, ignorant statement? There's no way in hell my wife and I could afford a house (even a cheap one, let alone a decent one) right now.. And we lead a relatively comfortable, priviledged life in a relatively inexpensive housing market. I can't even imagine how much worse the situation is for someone who is just barely making ends meet by living in a dirt-cheep tenement, and don't tell me that they could cancel their $40/mo cable and magically be able to afford a mortgage payment, let alone save up for a down payment. Believe it or not people who don't make enough money to afford a house do exist, do have rights, and do have the right to make a ruckus when they're being trampled on.
My daughter's teachers sit in front of an active whiteboard, hooked up to a laptop, so that, for example, she can see the effect of changing parameters in an equation graphed immediately. She then puts headphones on in the language lab, then uses Google Docs to do he creative writing (so she can easily finish it off at home).
That's fantastic. Though probably not at all the norm.
Where is it that teaching isn't using technology?
Outside of your affluent community. Inner cities. Rural areas. You know, where 75% of the people and 10% of the wealth are (numbers admittedly made up, but the sentiment is there). I'm not trying to say anything bad about your daughter's schooling - I think it's great - but don't think this is normal.
One of the seminal analytic papers in this area is
Tamás Vicsek, András Czirók, Eshel Ben-Jacob, and Inon Cohen ``Novel type of phase transition in a system of self-driven particles'' Phys. Rev. Lett. 75 1226 (1995)
In the above, a phenomenon called "collective memory" was exhibited in a model similar to Reynolds'. Individual members of the group have no explicit memory, but the group as a whole exhibits behavior that differs depending on the previous state of the group - in effect a "group memory".
Also, a shameless plug for my own software/API designed for similar simulations: glSwarm. Admittedly in a very early state of development, but functional enough to play with.
I was there, saw the demonstration, and wondered why this was being done on physical robots rather than on virtual robots. I guess its more impressive to see the little cars roaming around with their blinky lights.
As someone who has done both (virtual and real groups), doing things with real robots tends to open up a whole can of worms in terms of practical issues. This is true in a great variety of engineering fields... start with a simulation to get the major kinks out, move to the real thing and realize you hadn't accounted for X,Y, and Z in your simulation.
There is also immense military interest. Research doesn't get done on a large scale without funding. Funding, generally speaking - at least in engineering, doesn't come without someone with some influence being convinced that there will be applications.
This is just for the sake of conversation. I have studied evolutionary biology somewhat (and work with a few people whose main field this is, particularly with respect to altruism/selfishness and collective behavior), so I find it interesting.
You're right, much of the theory assumes a constant environment or some consistent probabilistic model of one. It still has done quite well as a theory. Also, it's a common misconception to assume that evolution finds a globally optimal strategy - this is in fact not (necessarily) true. Evolution finds a strategy that cannot be invaded by a mutant contingent. You may think of this as a "local" maximum, though it's somewhat different. This was Maynard Smith's brilliance of bringing a game-theoretic approach into the theory; an approach that owes very much to John Nash. One must also be very careful when considering good/bad leadership situations - on the (typically) massive scale of evolutionary action it is average leader, or the statistical distribution of leaders, that counts in some sense.
That said, the question of acting altruistically (i.e. do whatever is best for the group) vs. acting selfishly is a somewhat open research question. There are also several open questions in the area of leadership. I'd be glad to share references if you are interested.
Ant colonies sound a lot like slashdot it seems...
Actually, this is literally one of the main points of the book Emergence by Stephen Johnson. Personally, I didn't like the book, but only based on execution.. it was a good premise.
The key to swarm intelligence is evolution and experience over a long time. Individuals who aren't responsible and don't follow the rules get wiped out by predators or accidents (or the society itself), and individuals have strong reasons for following the society's rules (personal survival). The corollary of this is that if we want to make humans behave intelligently within a group, we must make sure they know the rules and that failure to observe the rules would lead to personal and group death.
Actually, a key idea of evolution is the very fact that some individual *will* break the rules. If a population follows strategy X, and an individual can do better by following strategy X+epsilon, the theory says that the population will tend towards the new strategy. Check out Evolutionarily Stable Stragies, one tool that is often applied to studies of collective behavior.
The correct term is Dynamical Systems, and its common, consistent language is the branch of mathematics dealing with dynamical systems (complete with its own vocabulary -- strange attractors, manifolds, emergence, chaos, etc.)
Dynamical systems is *one* approach to studying collective motion/swarming/whatever you want to call it. In this framework we consider the "collective" as a group of coupled ordinary differential equations (representing the states of each "agent"), and see what we can say about the whole system (or vice versa). The applications of dynamical systems theory are VERY far reaching. Generally speaking, this is an example of a Lagrangian approach.
Another type of approach altogether is the Eulerian approach. Here we consider a point in space and talk about the volume or density of individuals over time, usually using partial differential equations.
Both of the above are only examples of mathematical frameworks. Others are used in this field as well, for example game theory and algorithm analysis. Even then you have to back up a step and consider that there are many many others who study these problems from non-mathematical approaches - there are lots of experimental biologists doing really cool work here.
The bottom line is that the OP hits on a key issue in this field: there are a LOT of different people working in it, and they all speak different languages and have different interests and goals. It can throw up barriers at times, but for the most part it makes it a really exciting field to work in.
ease traffic back-ups at lower cost, particularly in rural areas.
For my next trick, I will quickly and efficiently shave all the hairs off a dolphin!
I had the same thought (though not phrased as amusingly). Then I bravely clicked the link and found that nowhere in the story nor on Coifmans website does it EVER say the word "rural". In fact it talks about implementations in Ohio CITIES and near UC Berkeley.
Hey dude, no problem. No personal harm meant. I'm just trying to be a smartass. You got me.
Purely in the name of dialog, though, the number of news stories on a given topic is not necessarily indicative of its scope or importance. Compare: Number of stories and/or blog posts (at least in the US) on Paris Hilton's prison release vs. the weekend's terrorist activity in Britain.
For me, at the top of cnn.com is a "Hot topics" bar, which includes an "iPhone" link that brings up 75 stories (not all about said problems, just about the phone in general). It comes up if you search the page for "iphone". Maybe this is determined from cookies or something.
The presence of three identical stories sourced from AP that quote "almost all customers" activating within minutes... nope, doesn't prove your argument, sorry.
*shrugs*
I didn't even read the stories, and I'm not really trying to make an argument. I just thought I'd call you on making a sweeping statement that was untrue. It caught my eye because it was all over the television this morning, so I figured it must also be all over the news sites.
Personally, I think both the phone and the ensuing problems are over-hyped (don't get me wrong, it's a pretty cool device). How else does the media operate but by sensationalizing things, though? It's a direct consequence of their social function.
Slashdot Mirror
Agreed. Most of his short stories are great. The Screamers movie is typical of other PKD adaptations, though. If you can put that aside it's actually decently entertaining early-90's SF schlock.
Also missed Screamers, based loosely on the short story "Second Variety".
If you use a Mac and are in a Latex-centric field, I find Bibdesk (http://bibdesk.sourceforge.net/) really great for managing reference pdfs and use cvs or svn if I really want to manage a document I'm working on. There's no annotation in Bibdesk but you can record notes and it generates bibtex for you.
Accuracy, to an experimentalist, is a quantitative measure of closeness to the "correct" value (i.e. accuracy vs. precision).
Anyways, as the pp said, this is really just an example of poor reporting. The actual article doesn't really make this kind of claim, in fact the word "accuracy" doesn't even appear in TFA. They'd never get published in Nature (it's actually Nature Nanotechnology) saying something was "100% accurate" because it doesn't make any sense formally.
"100% Accuracy" implies a positional error of zero meters (to infinite decimal places), which is obviously not what they're talking about.
I caught that, too. But really "percent" doesn't even make sense as a unit of accuracy, does it? Unless it's fractional, in which case I'd take it to mean that if you want to make a relative move of x, you'll get something in the range (0,2x) or maybe (0.5x, 1.5x)? I mean, on the nano scale that's still kind of remarkable, but as you've pointed out it's just not what they mean. /pedantic
They are similar rules to boids. The aim is a little different though.
Nice game. Check out glSwarm. I always thought it would be fun to make a game based on these ideas.
We apply tools like nonlinear control theory and graph theory to study these kinds of "rules" with rigor, with the aim of a) designing robotic (specifically mobile sensor) networks that are bio-inspired in the way you mentioned and b) help the biologists by providing insight from our perspective.
My favorite thing about Dr. Couzin is his willingness to work with people in other disciplines - particularly the "harder" sciences. It's mentioned towards the end of the article. My advisor, Naomi Leonard, and her students have published several papers with Iain as a co-author - see http://www.princeton.edu/~naomi/ and search for "Couzin" and "Levin" on the page for a few references. Dr. Grunbaum, who is also mentioned in the article, is great with this as well (also on the page). They are both fantastic guys to work with.
And a shameless plug for my tiny contribution - http://www.princeton.edu/~dswain/publications/2007/DSCDC07.pdf
One common set of rules is (related to boids) 1. Move away from those closest to you. 2. Move towards those farthest from you. 3. Align with those in the middle.
On the other hand, it would be erroneous to posit a common set of rules for every species. Different evolutionary pressures produce different behaviors in different situations. An interesting facet of this type of research is to see how different sets of parameters for the same set of rules can produce different behaviors, or different sequences of behaviors.
One of the seminal analytic papers in this area is
Tamás Vicsek, András Czirók, Eshel Ben-Jacob, and Inon Cohen ``Novel type of phase transition in a system of self-driven particles'' Phys. Rev. Lett. 75 1226 (1995)
Another great paper:
Couzin, I.D., Krause, J., James, R., Ruxton, G.D. & Franks, N.R. (2002) Collective memory and spatial sorting in animal groups Journal of Theoretical Biology 218, 1-11.
In the above, a phenomenon called "collective memory" was exhibited in a model similar to Reynolds'. Individual members of the group have no explicit memory, but the group as a whole exhibits behavior that differs depending on the previous state of the group - in effect a "group memory".
Also, a shameless plug for my own software/API designed for similar simulations: glSwarm. Admittedly in a very early state of development, but functional enough to play with.
ASAP program, where a group of coordinated robots was used to perform oceanographic measurements.
As someone who has done both (virtual and real groups), doing things with real robots tends to open up a whole can of worms in terms of practical issues. This is true in a great variety of engineering fields... start with a simulation to get the major kinks out, move to the real thing and realize you hadn't accounted for X,Y, and Z in your simulation.
There is also immense military interest. Research doesn't get done on a large scale without funding. Funding, generally speaking - at least in engineering, doesn't come without someone with some influence being convinced that there will be applications.
This is just for the sake of conversation. I have studied evolutionary biology somewhat (and work with a few people whose main field this is, particularly with respect to altruism/selfishness and collective behavior), so I find it interesting.
You're right, much of the theory assumes a constant environment or some consistent probabilistic model of one. It still has done quite well as a theory. Also, it's a common misconception to assume that evolution finds a globally optimal strategy - this is in fact not (necessarily) true. Evolution finds a strategy that cannot be invaded by a mutant contingent. You may think of this as a "local" maximum, though it's somewhat different. This was Maynard Smith's brilliance of bringing a game-theoretic approach into the theory; an approach that owes very much to John Nash. One must also be very careful when considering good/bad leadership situations - on the (typically) massive scale of evolutionary action it is average leader, or the statistical distribution of leaders, that counts in some sense.
That said, the question of acting altruistically (i.e. do whatever is best for the group) vs. acting selfishly is a somewhat open research question. There are also several open questions in the area of leadership. I'd be glad to share references if you are interested.
Another type of approach altogether is the Eulerian approach. Here we consider a point in space and talk about the volume or density of individuals over time, usually using partial differential equations.
Both of the above are only examples of mathematical frameworks. Others are used in this field as well, for example game theory and algorithm analysis. Even then you have to back up a step and consider that there are many many others who study these problems from non-mathematical approaches - there are lots of experimental biologists doing really cool work here.
The bottom line is that the OP hits on a key issue in this field: there are a LOT of different people working in it, and they all speak different languages and have different interests and goals. It can throw up barriers at times, but for the most part it makes it a really exciting field to work in.
Excellent editing job, guys.
Hey dude, no problem. No personal harm meant. I'm just trying to be a smartass. You got me.
Purely in the name of dialog, though, the number of news stories on a given topic is not necessarily indicative of its scope or importance. Compare: Number of stories and/or blog posts (at least in the US) on Paris Hilton's prison release vs. the weekend's terrorist activity in Britain.
For me, at the top of cnn.com is a "Hot topics" bar, which includes an "iPhone" link that brings up 75 stories (not all about said problems, just about the phone in general). It comes up if you search the page for "iphone". Maybe this is determined from cookies or something.
I didn't even read the stories, and I'm not really trying to make an argument. I just thought I'd call you on making a sweeping statement that was untrue. It caught my eye because it was all over the television this morning, so I figured it must also be all over the news sites.
Personally, I think both the phone and the ensuing problems are over-hyped (don't get me wrong, it's a pretty cool device). How else does the media operate but by sensationalizing things, though? It's a direct consequence of their social function.