Yes I do understand that. In fact it was the whole point that I was trying to make! It's also extremely obvious that any function that can be performed by a living creature can be copied by a sufficiently advanced neural network, since that's exactly how the living creature does it. What's that called, a tautology?
Anyway if you've got a link to a study showing a neural network that can strike a flying insect out of the sky with it's tongue, I'd be exceedingly interested in reading it..?
Fine, except chimps made it to the third step in that chain (well more or less) so it doesn't really explain much. AFAIK there are a lot of reasons that come together in a "perfect storm" - roaming lifestyle, opposable thumb, vocal capability and so on.
It's a bit like us when we can simply immediately point to an intercept between two curves on a graph.
When we do that, there is some maths happening in our brains, it just isn't conscious. You're right, that is exactly what is happening in the spider's case. However to "just point" to an intercept seems like an incredibly simple thing to us, but to do it with the amount of brain cells a spider has is quite a trick. Bear in mind this all has to come from sensory data - it has to find branches, blades of grass or whatever and make a decision whether it is feasible to spin a web there, using very rough input from it's eyes. Try writing software for a robot to do that - if you manage it you might get a nobel prize. Even in a very simplified virtual world with perfect data, there would be a fair bit of maths, even if it's just basic trig.
I've just read the first 5 lines of the wikipedia article on Markov chains and am therefore the closest thing to an expert on this subject you're likely to find any time soon. Does the Markov property prohibit causal interdependence between events? That is to say, is it in keeping with a markov chain if a species develops a higher intelligence in order to evade an otherwise unrelated predator species? I would suggest not.
So the upshot here is that the intelligence of any given creature is not a function of it's size or age (in evolutionary terms) but is very tightly geared towards the problems it likely faces in it's natural environment.
For example, even a spider can do quite tricky maths in order to work out how to spin a web between arbitrary fixed points, yet is completely flummoxed by even the simplest general knowledge quiz.
So what I want to know is, what was it about human beings that caused us to develop the capacity to drive cars, build computers and walk on the moon?
Whenever I hear "Chicken Little" I think of the novel The Space Merchants by Pohl & Cornbluth... So I was wondering what a giant GM KFC had to do with the sky being slightly closer than expected..?
At least the windows registry keeps everything in one place. I use Linux a lot at work and at home and I find I spend quite a lot of time hunting around for config files. Don't get me wrong, I love Linux, but it's clearly a pro setup.
I agree that businesses that make bad decisions should be allowed to fail. So did the UK government, until they saw what happened after Northern Rock (a bank) failed and caused a run.
I'm no expert on this but my understanding is that they then realised it was just the first of many. It then looked like if there was no intervention, most of the banking industry would fall apart. Sure, Lloyds TSB might have been ok plus a few smaller ones but it would've taken out the savings and current accounts of millions of people, plus destroyed a lot of infrastructure, making it difficult for the survivors to take up the slack as would be ideal.
In the meantime, very many cashpoints would stop working and therefore many people would be unable to get cash or use their cards in order to buy food. BANG! Society falls apart.
If this is anything like the UK banking sector, you may be greatly underestimating the danger that the memory chip business is in!
What you're missing here, is that it's wrong to take what anyone earns to give it to someone else
No it isn't! I've never seen such a sweeping generalisation. So, in your view, we should get rid of all taxes, let all public services cease and prepare for anarchy?
What are you, crazy? You just have to write a completely generalised GA then evolve the parameters! Duh! I have such a thing on my laptop right here, honest. Not showing you though.
No, trust in the laws of nature! If you think of it like that you'd be paranoid of falling over all the time in case your legs evolved wrong. It is just random chance, after all! Try not to think about congenital birth defects.
Brilliant! That is the sort of thing they're good for - optimisations (like TSP). Have been trying to think of something I can use it for in the office - would probably get me a promotion just for cleverness - but I haven't thought of anything yet.
Also I forgot the mention the application in real-time scenarios like the city (I forget where) that had the water supply balance supply using computer-operated valves. Or the real-time missile guidance system that would rotate different tracking strategies and evolve them while it was in the air, so being able to adapt to more or less any evasive maneuver by the target.
Awesome will give those a read! Before I do, I did read a couple of papers (years ago) on self-programming and had a play with the application of GA to it. Basically you'd start with a fixed input and model output then evolve a source file to produce the bit in the middle and try to compile it. You'd then match the output to your model as a fitness score which is trivial if you're just manipulating text. I could never think of a way to improve the viability of a semi-random source though- in other words, an evolved and mutated source file has a very low chance of actually compiling in any language you'd want to use!
Yeah I wrote a GA at university. Unfortunately the professor didn't like it and gave it a crappy mark. Anyway, it's not too tricky to make something look impressive but the problem with it is finding some real-world application for it...
The main thing you learn from GAs is that we know a lot already and to get useful information from a simulation it has to be VERY complex. Finding a sensible fitness measure is often a big problem too. It can be a good idea to set up two or more different populations and have them compete against each other- then you don;t need a fitness measure.
Eventually the complexities of the evolved individuals plateau and you don't see any further evolution even if you make the "DNA" longer.
One other thing you can do with it is to find optimal solutions like in volume/area problems or travelling salesman-y stuff. But again, it can be tough to find a useful application! The best thing I did with it was coming up with an optimal maze solution and unit placement for Desktop Tower Defense! Still couldn't crack level 100 though...:(
I'd go even further and ask the question: how much do we really care about GDP? OK it's a broadly representative number which affects employment and so on but it can't be a particularly accurate figure at any point. Who knows exactly how much stuff the whole country made in a year?
Moreover, we should probably wrench our sight away from pure cash wealth and think about things in the round a bit more.
There are so many mainstream sources around UNSCOM and the IAEA that have come forward since the Iraq war that the truth is no longer in question. It goes like this:
Firstly, the Iraqi military and economy had been smashed by the first gulf war and subsequent sanctions.
Secondly, Hussein and the rest weren't stupid and clearly knew the US government, public and media were all baying for war.
Thirdly, the UNSCOM inspections were very thorough and even well funded and equipped (largely by the US taxpayer) and had a great deal of success in pressurizing the Iraqi regime into getting rid of what it had left, which in any case wasn't much because it was all so old. Dozens of Iraqi army officers defected through Syria or Jordan and confirmed the story.
Because of all this, successive administrations tried and failed to find a pretext to war. Parent is entirely correct - it's the defense industry wanting cash and the government finding any excuse to pump tax dollars to their well-to-do pals. It's good for the economy, according to the politicians.
What helped most of all to tip the balance in favour of war was when someone or other (probably the CIA) forged a now infamous document purporting to show the sale of yellow-cake uranium by Niger to Iraq. It was by all accounts a hilariously bad forgery and contained many, many obvious errors that clearly showed it could not be genuine. However, the White House released it to the media as genuine, who immediately, without checking it, presented it as causus belli to the trusting public. By the time the IAEA's Mohammed El Baradei announced a couple of days later that it was utterly false, it was too late. Not that the same, supposedly liberal media made a big deal of that.
That, my friend, is how rich, powerful people can manipulate the public into doing whatever they see as necessary, even when it calls for the deaths of hundreds of thousands of normal, working class people on all sides.
a 3 1/2 year old isn't going to choke to death on a keyboard key for two reasons.
First, if they did anything they'd swallow it like food. Just because it isn't edible doesn't mean they'll automatically breathe it in somehow.
Secondly, if you're any kind of parent they should know by now what's FOOD and what's TOYS.
I mean, you sound like one of these typical overwrought parents who get worked up about things and won't let their kids do this or that and end up with useless, beaten down kids with no personality. Lighten up a bit.
What's the problem? Assuming this person isn't a horrible, horrible person then anything that can keep a child occupied for half an hour is a real boon for any parent.
My boy's 3 and a half and the only time I get to do anything useful is when he's sat in front of the TV. Like it or not, that's what parenting toddlers is about - spending time with them yes, but also managing to clean the house, cook dinner, wash clothes etc.
I wouldn't buy one new because it may well get trashed (especially if they think of it as "their toy") but if you've an old box hanging around like I have, then why not? The BBC Cbeebies site is excellent and can keep a child busy for a little while. Kids pick up how to use a mouse within days and point-and-click websites at least make them think and concentrate for a while, as opposed to the complete mental switch-off that happens when watching a Barney DVD or somesuch.
You may be right but I don't think it's over yet - it wouldn't be too tough to routinely encrypt most or all traffic. Anyhow, even if the media companies do manage to maim the internet by throwing lawyers at ISPs, I doubt this'll save the record companies, because the internet is only part of their problem. If I want to I can get 100+ free albums just by asking people who sit near me at work, without having to be online at all. Virtually everyone's got a cheap USB hard disk plugged into their laptops, each one of which can store enormous amounts of music.
Slashdot Mirror
Yes I do understand that. In fact it was the whole point that I was trying to make! It's also extremely obvious that any function that can be performed by a living creature can be copied by a sufficiently advanced neural network, since that's exactly how the living creature does it. What's that called, a tautology?
Anyway if you've got a link to a study showing a neural network that can strike a flying insect out of the sky with it's tongue, I'd be exceedingly interested in reading it..?
OK I should have said "calulation" and not "maths". Maths, as you said, is a representational system and a spider is indeed not capable of any maths.
Your second point is mere nitpicking.
Fine, except chimps made it to the third step in that chain (well more or less) so it doesn't really explain much. AFAIK there are a lot of reasons that come together in a "perfect storm" - roaming lifestyle, opposable thumb, vocal capability and so on.
It's a bit like us when we can simply immediately point to an intercept between two curves on a graph.
When we do that, there is some maths happening in our brains, it just isn't conscious. You're right, that is exactly what is happening in the spider's case. However to "just point" to an intercept seems like an incredibly simple thing to us, but to do it with the amount of brain cells a spider has is quite a trick. Bear in mind this all has to come from sensory data - it has to find branches, blades of grass or whatever and make a decision whether it is feasible to spin a web there, using very rough input from it's eyes. Try writing software for a robot to do that - if you manage it you might get a nobel prize. Even in a very simplified virtual world with perfect data, there would be a fair bit of maths, even if it's just basic trig.
I've just read the first 5 lines of the wikipedia article on Markov chains and am therefore the closest thing to an expert on this subject you're likely to find any time soon. Does the Markov property prohibit causal interdependence between events? That is to say, is it in keeping with a markov chain if a species develops a higher intelligence in order to evade an otherwise unrelated predator species? I would suggest not.
Speak for yourself, titch!
So the upshot here is that the intelligence of any given creature is not a function of it's size or age (in evolutionary terms) but is very tightly geared towards the problems it likely faces in it's natural environment.
For example, even a spider can do quite tricky maths in order to work out how to spin a web between arbitrary fixed points, yet is completely flummoxed by even the simplest general knowledge quiz.
So what I want to know is, what was it about human beings that caused us to develop the capacity to drive cars, build computers and walk on the moon?
Come on slashdot - someone back me up on this classic bit of golden age SF!
Whenever I hear "Chicken Little" I think of the novel The Space Merchants by Pohl & Cornbluth... So I was wondering what a giant GM KFC had to do with the sky being slightly closer than expected..?
At least the windows registry keeps everything in one place. I use Linux a lot at work and at home and I find I spend quite a lot of time hunting around for config files. Don't get me wrong, I love Linux, but it's clearly a pro setup.
I agree that businesses that make bad decisions should be allowed to fail. So did the UK government, until they saw what happened after Northern Rock (a bank) failed and caused a run.
I'm no expert on this but my understanding is that they then realised it was just the first of many. It then looked like if there was no intervention, most of the banking industry would fall apart. Sure, Lloyds TSB might have been ok plus a few smaller ones but it would've taken out the savings and current accounts of millions of people, plus destroyed a lot of infrastructure, making it difficult for the survivors to take up the slack as would be ideal.
In the meantime, very many cashpoints would stop working and therefore many people would be unable to get cash or use their cards in order to buy food. BANG! Society falls apart.
If this is anything like the UK banking sector, you may be greatly underestimating the danger that the memory chip business is in!
What you're missing here, is that it's wrong to take what anyone earns to give it to someone else
No it isn't! I've never seen such a sweeping generalisation. So, in your view, we should get rid of all taxes, let all public services cease and prepare for anarchy?
That doesn't mean it isn't virtualised! If they can make it work on ARM then it can't just be a straight plug through to native code.
What are you, crazy? You just have to write a completely generalised GA then evolve the parameters! Duh! I have such a thing on my laptop right here, honest. Not showing you though.
No, trust in the laws of nature! If you think of it like that you'd be paranoid of falling over all the time in case your legs evolved wrong. It is just random chance, after all! Try not to think about congenital birth defects.
Brilliant! That is the sort of thing they're good for - optimisations (like TSP). Have been trying to think of something I can use it for in the office - would probably get me a promotion just for cleverness - but I haven't thought of anything yet.
Also I forgot the mention the application in real-time scenarios like the city (I forget where) that had the water supply balance supply using computer-operated valves. Or the real-time missile guidance system that would rotate different tracking strategies and evolve them while it was in the air, so being able to adapt to more or less any evasive maneuver by the target.
Awesome will give those a read! Before I do, I did read a couple of papers (years ago) on self-programming and had a play with the application of GA to it. Basically you'd start with a fixed input and model output then evolve a source file to produce the bit in the middle and try to compile it. You'd then match the output to your model as a fitness score which is trivial if you're just manipulating text. I could never think of a way to improve the viability of a semi-random source though- in other words, an evolved and mutated source file has a very low chance of actually compiling in any language you'd want to use!
Yeah I wrote a GA at university. Unfortunately the professor didn't like it and gave it a crappy mark. Anyway, it's not too tricky to make something look impressive but the problem with it is finding some real-world application for it...
:(
The main thing you learn from GAs is that we know a lot already and to get useful information from a simulation it has to be VERY complex. Finding a sensible fitness measure is often a big problem too. It can be a good idea to set up two or more different populations and have them compete against each other- then you don;t need a fitness measure.
Eventually the complexities of the evolved individuals plateau and you don't see any further evolution even if you make the "DNA" longer.
One other thing you can do with it is to find optimal solutions like in volume/area problems or travelling salesman-y stuff. But again, it can be tough to find a useful application! The best thing I did with it was coming up with an optimal maze solution and unit placement for Desktop Tower Defense! Still couldn't crack level 100 though...
I'd go even further and ask the question: how much do we really care about GDP? OK it's a broadly representative number which affects employment and so on but it can't be a particularly accurate figure at any point. Who knows exactly how much stuff the whole country made in a year?
Moreover, we should probably wrench our sight away from pure cash wealth and think about things in the round a bit more.
You must live in a parallel universe!
There are so many mainstream sources around UNSCOM and the IAEA that have come forward since the Iraq war that the truth is no longer in question. It goes like this:
Firstly, the Iraqi military and economy had been smashed by the first gulf war and subsequent sanctions.
Secondly, Hussein and the rest weren't stupid and clearly knew the US government, public and media were all baying for war.
Thirdly, the UNSCOM inspections were very thorough and even well funded and equipped (largely by the US taxpayer) and had a great deal of success in pressurizing the Iraqi regime into getting rid of what it had left, which in any case wasn't much because it was all so old. Dozens of Iraqi army officers defected through Syria or Jordan and confirmed the story.
Because of all this, successive administrations tried and failed to find a pretext to war. Parent is entirely correct - it's the defense industry wanting cash and the government finding any excuse to pump tax dollars to their well-to-do pals. It's good for the economy, according to the politicians.
What helped most of all to tip the balance in favour of war was when someone or other (probably the CIA) forged a now infamous document purporting to show the sale of yellow-cake uranium by Niger to Iraq. It was by all accounts a hilariously bad forgery and contained many, many obvious errors that clearly showed it could not be genuine. However, the White House released it to the media as genuine, who immediately, without checking it, presented it as causus belli to the trusting public. By the time the IAEA's Mohammed El Baradei announced a couple of days later that it was utterly false, it was too late. Not that the same, supposedly liberal media made a big deal of that.
That, my friend, is how rich, powerful people can manipulate the public into doing whatever they see as necessary, even when it calls for the deaths of hundreds of thousands of normal, working class people on all sides.
An aircraft navigation system that runs on AA batteries? Can't see any problems there then.
a 3 1/2 year old isn't going to choke to death on a keyboard key for two reasons.
First, if they did anything they'd swallow it like food. Just because it isn't edible doesn't mean they'll automatically breathe it in somehow.
Secondly, if you're any kind of parent they should know by now what's FOOD and what's TOYS.
I mean, you sound like one of these typical overwrought parents who get worked up about things and won't let their kids do this or that and end up with useless, beaten down kids with no personality. Lighten up a bit.
Careful, you're playing with fire: http://www.damninteresting.com/?p=715
What's the problem? Assuming this person isn't a horrible, horrible person then anything that can keep a child occupied for half an hour is a real boon for any parent.
My boy's 3 and a half and the only time I get to do anything useful is when he's sat in front of the TV. Like it or not, that's what parenting toddlers is about - spending time with them yes, but also managing to clean the house, cook dinner, wash clothes etc.
I wouldn't buy one new because it may well get trashed (especially if they think of it as "their toy") but if you've an old box hanging around like I have, then why not? The BBC Cbeebies site is excellent and can keep a child busy for a little while. Kids pick up how to use a mouse within days and point-and-click websites at least make them think and concentrate for a while, as opposed to the complete mental switch-off that happens when watching a Barney DVD or somesuch.
You may be right but I don't think it's over yet - it wouldn't be too tough to routinely encrypt most or all traffic. Anyhow, even if the media companies do manage to maim the internet by throwing lawyers at ISPs, I doubt this'll save the record companies, because the internet is only part of their problem. If I want to I can get 100+ free albums just by asking people who sit near me at work, without having to be online at all. Virtually everyone's got a cheap USB hard disk plugged into their laptops, each one of which can store enormous amounts of music.