The 'seamless hyperconnected information space' that a single page formed from multiple machines might represent, is in reality a simple web page with six ads and an animated cursor, all firing off demographic information to various companies.
So are you advocating that the lowest common denominator be enforced on everybody? Sure there are plenty of crude pages with six and more banners -- but that is not a good reason to ban everything more complicated than single-server plain-vanilla HTML...
A prompt pops up. "Do you want to add 209.207.224.245 to slashdot.org's trust realm?"
You know what? For a while I ran my browser with the setting "ask me about cookies" turned to "on". On the one hand, it was good and useful -- I knew who wanted to place a cookie on my machine and could always tell them to fuck off. Unfortunately, I tired rather quickly of clicking "no", "no", "no", on all those pop-up dialog boxes that appeared with dreadful regularity. Solution? I installed junkbuster and now live in peace.
I think that asking every time about the trust realm is going to irritate most about everybody. Instead I propose a side-window or a panel that shows you whom does your browser connect to (optional, of course). Add to this a built-in ability to exclude certain machines and domains (a la junkbuster) and all this becomes reasonable. If you care, you'll look and exclude. If you don't you'll ignore it.
Another solution is to use ZeroKnowledge's Freedom -- I'm perfectly happy to let the markedroids collect noise (what they believe to be info) about my nyms...
Asking the average user to do this to ensure security is of course ludicrous.
Of course. But there are a lot of people out there (actually, mostly here, on Slashdot) that are quite competent to handle tcpdump and are quite interested in what exactly does their browser do. It only take a single person to sound the alarm.
A bad thing for whom? Doubleclick or the user?
The user.
What I propose is that I have direct control over the network activity that my machine is initiating.
You are welcome. Just telnet www.webserver.com 80 and type away. For most users that's waaay over their heads. They know that if you click a link, a page will appear. They are not concerned (and should not be) which machine is serving the page, whether it is one machine or several, whether the server is real or there is an akamai/sandpiper server cloud generating the content, etc. etc.
I take issue with the implication that my browser should be a black box that is free to perform arbitrary network connections on my behalf. THAT is a Bad Thing, IMHO.
Well, don't use the web, then. The web is not a system of point-to-point connections as you seem to believe -- it is a multiconnected network and I see no good reason to impose this huge constraint on it: that a page cannot contain information from more than one machine. To repeat myself, the web is meant to be a seamless whole. You are arguing for the days of the UUCP connections where yes, you did know exactly when and to whom you were connected.
Instead of a simple 'explicit host only' solution, a 'trust realm' scheme may be effective
Well, "host only" solution would break most of the web immediately. I will not argue against including this choice in a browser (choice is always good) but switching this on will make the browser basically unusable.
A "trust realm" scheme suffers from the usual defect: who will define the trust? I, as a user, certainly don't want to be bothered with it and if a site will be able to define its own trust realm we're back to square one since that's mostly what you are trying to avoid.
Using junkbuster and selecting the proper domains to block will give you most of what you want anyway.
Detecting this kind of activity should be trivial. A nice feature for browsers (or any other net-aware software) would be to have an 'honor only explicit requests' mode, which would allow the browser to open connections only to the host that you've specified. No remote images, no remote IFRAME pages, no extraneous sockets opened up in applets or other embedded controls. I think this would neatly solve all sorts of network security problems. (As well as all sorts of advertizing 'problems'.)
Why should your browser be connecting to machines that you didn't ask it to?
Well, that's not trivial at all. Think about what the web is supposed to represent -- a seamless whole where the boundaries between machines that serve you information are irrelevant. What you propose would kill all web pages that pull content from more than one machine and that is a Bad Thing.
I'm afraid that this solution, to quote somebody's sig is simple, elegant and wrong. Looking at tcpdump output is still the preferred way to handle these issues.
For people interested in what TC calls "zenning" and what is usually called "the flow" check out the work of a guy with the improbable name of Csikszentmihalyi (search on Google) who is usually credited with first researching the concept. One place to start is www.flownetwork.com.
There is not logical notion that human kind has any implied fuzzy quasi-telepathic state wherin they gain "mystical" powers.
Nobody is saying anything about mystical powers. This is one of the so-called altered states of consciousness, specifically one in which you can achieve and maintain high concentration for a long period of time. It is often called 'flow'. The existence of such a state is widely recognized and documented. For example, being able to go into 'flow' is one of the characteristics of a world-class athlete in individual-competiton sports (martial arts, tennis, etc.)
I think it's outrageous that large companies would even think to do this
Welcome to the real world, pal.
I don't want to sound innocent, they're obviously doing this for profit
You mean this never crossed your mind before? Corporations exist to make a profit -- that's their purpose in life.
they must realize that it will backfire when everyone gets disgusted by their tactics
Not necessarily. Random J. Luser is quite happy to have all the information about him collected, stored, indexed and catalogued. He is told that this is for his own benefit so that he can have personalized service (i.e. most effective methods for separating him from his cash/bank balance). Poll after poll show that Americans (at least) are not all that much concerned about privacy erosion. I am afraid that people who take privacy seriously will soon get hit with a "paranoid nut" label. I've already seen it used.
The current bottleneck in all the mobile/wearable devices is battery power. Batteries are too heavy and last too short a time. The superslim notebooks -- ones that you can actually carry around -- survive for 2-3 hours at most without an electric outlet. A variety of PDAs can go through the day without recharging, but they all lack decent keyboards and tend to be quite limited in what they do.
Basically, for wearable computers to take off, batteries have to become much lighter.
For all you people drooling about the possibilities of controlling machines (computers) purely by the mind, stop and thing for a second. If your thoughts are going to be controlling anything in any reasonable fashion, you'd better be concentrating real hard! It's perfectly possible to interface with a computer (keyboard/mouse/monitor) and think about a couple of different things at the same time. Well, it's not going to work any more. Imagine a worker controlling, say, a factory robot by his thoughts. If any time the worker thinks about beer or [insert your favorite porn star here] that robot will jerk, or stop, or drill the wrong thing, then the factory better start recruiting tibetian monks for its workers.
The point is that thought control of any sophistication requires not only the neural interface. It also requires a lot of concentration and mind discipline. Operating stuff by mind control is going to be hard.
ever had an infected appendix? Would you think twice about having your stomach cut open and having it removed if your doctor told you it needed to be done? Do you know anyone who would? The only reason we don't all have it done (literally) is cost.
Bzzzzt. Sorry, bad argument. I would agree to have myself cut open and my appendix removed because the alternative is dying. If my doctor told me he wanted to remove my appendix just because he couldn't see any use for the damn thing, I would laugh long and hard and then change the doctor. It has nothing to do with cost, but is somewhat connected to the Murphy's law, which, loosely paraphrased, says "If anything can go wrong, it will".
Having it removed pre-emptively would be the safest thing to do, since there would then be no chance of infection.
Well, no. You are trading off a probability of your appendix becoming inflamed at the time when you are far away from medical facilities (as in cannot get to a hospital within 12 hours or so) against the probability of something going wrong during the operation (as in suddenly developing an intolerance of anesthesia). The choice is not at all clear-cut and depends on a lot of things, like how far from the hospital you live.
Besides, how do you know the appendix is useless? Contemporary medical science still does not understand a lot of things about human body.
In the US at least their is some sort of weird moral structure that says it is okay to "fix defects" but not to improve healthy people.
That's an old Judea-Christian hangup, nothing to do with the US specifically. Basically, the Big Three middle-eastern relgions (Judaism, Christianity and Islam) believe that since God created humans in his own image, trying to improve on his creation is a sin (of pride, mostly) and is to be forbidden. You will rarely find this spelled out explicitly, but the attitude is quite prevalent throughout the Western culture, even if most people cannot explain why they think so.
First, I should define an adaptive knowledge-based system. Basically, a KBS is a big database search program
Oh, OK. You mean what I would call an expert system.
In fact, it's been years since I've seen a backpropogation or Bayesian NN used inside these walls. We've been using realtime systems closer to CORE, which are not simple back-to-front-and-learn-backwards networks. They are latticed in all directions, a lot of them are at least partially hardware.
So? Backprop is admittedly a very inefficient search method, so barely anybody uses it any more. Besides, a "neural net" is a very fuzzy:) concept -- for example, the plain-vanilla feed-forward nets are very different from Kohonen nets which are different from, say, gated experts architecture. And whether something runs in real time, or is implemented in hardware is really irrelevant to the discussion.
You seem to have the misconception that an NN-based system outputs a number
More, I insist on it -- only not *a* number, but a set of numbers. Unless you are doing analogue neural net (which I'll admit I never heard about) there is no way you can avoid the fact that the output of the net is numbers.
Like a robot that demonstrates a fear of light for instance
And this is a big deal? I can build one out of my Lego Mindstorms set.
NN systems are not merely a statistical system, but an approximation of the brain
Bullshit. Historically, NNs were developed as an approximation to the animals' nervous system, but that's no longer relevant. Contemporary NNs are statistical models, typically with a very large number of parameters and sometimes with interesting search strategies in parameter space. If you are going to claim that a NN is an approximation of the brain, I'm going to claim the same for my projection pursuit regression (of which a three-layer feed-forward net is a special case). Unfortunately, it doesn't sound as cool.
we've seen some interesting things happen, and quite often, the NN jumps to a totally unexpected conclusion
To repeat myselc, your inability to predict the outcome does not say anything non-trivial about the sophistication of the system you are observing.
We've seen time and again hardware systems succeed where a software model of the same model fails. Why? Chaos. Computers are terrible at calculating chaos. Hardware, on the other hand, has lots of little transmission delays.
You are not making much sense. Hardware systems typically succeed because they are orders of magnitude faster. And what do you mean by chaos? There are some fairly precise definitions of chaotic systems, but I don't think you have them in mind. Are you talking about analytic solutions of the lack of them? And it's not like it's hard to implement delays in software -- again, the main difference is speed.
But I hope that clears up some of the common misconceptions
for instance not ANY Austrialian's computer can be tapped ie: (i) allowing the Minister to issue warrants which may authorise ASIO to:
- remotely access data that is relevant to security and is stored in a computer specified in the warrant;
hence, there has to be a warrant issued, so they/shouldn't/ ( key word ) shouldn't be snooping around Joe Smoe's computer for no good reason.
First, in the US the search warrants are asked for by the executive branch (cops) but are issued by the judicial branch (judges). This is supposed to keep the cops more or less in line (and somewhat does). However here it seems that the Minister (=executive branch) is going to be issuing the warrants. If so, he can perfectly well delegate his authority to lower levels until, say the section head at ASIO will be able to issue warrants and it'll be perfectly legal.
Second, ASIO may be snooping around Joe Shmoe's computer for a good reason -- to them. That does not necessarily make it a good reason to me, to Joe Schmoe, or say, to a "reasonable person". Maybe ASIO decides that running Linux strongly correlates with anti-government sentiments, so there is a "good reason" to periodically run checks on those weirdo Linux guys...
What good will encryption do when they can change the data to whatever they want?
Unless you break the encryption, you *cannot* change the encrypted data to whatever you want. Doesn't mean that you cannot add unencrypted stuff, of course...
On the other hand, that might look a bit strange to a judge: "Yes, Your Honor, the accused had his whole hard disk encrypted except for this single file where he admitted to planning anti-government activities".
YARTEE (Yet Another Reason To Encrypt Everything).
It wouldn't surprise me if they outlawed the use of non-government-sactioned encryption products next. I'm really surprised that no one in the US has tried that.
They thought about it. Unfortunately for them, in the US at least, doing so would involve major violence to the constitution and so far the consensus is that the Supreme Court will laugh long and hard and then throw it out (first amendment and all that).
Doesn't mean the situation will stay the same forever, though.
If the U.S. government did what the Australian government just did, I'd pull the plug on my internet connection permanently. A big middle finger to those idiots. I'd rather be without the internet than to have the government in my computer.
You're bluffing, I don't believe you. Besides,
(1) The government would be happy for you to disconnect from the net: another malcontent made powerless.
(2) You can get the same result by closing all ports, running no servers at all, reading e-mail in a pure ASCII text reader, and browsing the web with something that does NOT understand Java or Javascript.
I think that the intent of the "changing data" provision is backdoors. Once they got into your system, they want easy access from now on. However, just because this is the original intent, there is no guarantee that soon enough some "smart" guy figures out that altering data can make their life much more easy... And even if it doesn't stand up in court -- what a wonderful tool for pressure tactics. It's the same "drop a packet of coke/heroin/etc. during the search, find it, and lean on the guy real hard" technique.
So does this mean that if I am an Australian and I detect a network intrusion that comes from a government machine, I am supposed to spread my legs and bend over?
The one thing I can tell you right now is that you DO get robots with personality disorders. In one of my preliminary tests, I had a computer vision system go into "complete negative lockdown", which is the equivalent to suicidal depression.
When you work with neural nets, you do get to use nice metaphors. However, personality disorders they didn't get to yet. A neural net is a statistical model, a system to approximate the relationship between the inputs and the outputs. Sure, I can call a NN that fails to fit the data and example of clinical idiocy, or say that one which goes into oscillations is manic-depressive. But all these are just convenient labels for my bored mind and are not different at all from calling a car that doesn't start well in the morning cranky and grumpy.
But the point is, from the second I switch on a neural network, I hardly have the foggiest notion what it'll be like. The same matrix that went negative-lockdown stayed "sane" for about a full minute when I tried the same test with the room lights darker.
That has nothing to do with intelligence or sentience. Any chaotic system can do this easily. Throw a handful of sand in the air -- do you have the foggiest notion of where the sand particles end up? No? Does this make sand intelligent?
However, most of my decent tests of neural systems all point to the existance of true artificial intelligence in them.
Did you mean decent or recent? Anyway, define "true artificial intelligence" and then it'll be possible to talk about it. People seem to think that if a piece of software generates some behaviour not hardcoded into it by the programmer, it means that the software is intelligent. Unfortunately, no. The problems that AI's been having over the last 30 years is proof of that.
While other experimental techniques like genetic algorithms, fuzzy logic and adaptive knowledge-based systems do still have the capability to mess up in ways the designer did not intend (your favorite search engine is proof of that), neural networks (which most experts believe is the most promising currently) have the capability for a MUCH larger range of reactions.
Buzzword-o-rama! Genetic algorithms are just a global optimization mechanism. Fuzzy logic is just a way to talk about partial membership of sets. Adaptive knowledge-based systems -- hard to say what you mean. I'd probably call neural nets a subclass of adaptive learning systems.
And what do you mean by a range of reactions? A neural net outputs a bunch of numbers. Do you mean that a NN can output more numbers? or that you can interpret the NN's numbers in more imaginative ways?
Just the opinion of somebody who dabbles in AI professionally.
So, basing yourself on some Babelfish translation of a German article written by journalists about some fictional implications of some EU technical commitee, you already blame certain individuals of a certain nation?
You are a clear example of your own law. You are an ignorant and an idiot.
A bit uptight today, aren't you? I would hazard a guess that something long and straight has been shoved up your ass recently.
As to certain individuals of certain nations, it happens to be a demonstratable fact that the French tend to be the most anti-American of all Western European nations.
[Little bit of chauvinism: this does not surprise me since it comes from an American. Most Americans are semiliterate and don't even know where their country is on a map of the globe. No wonder their universities and research labs are staffed with foreigners.]
Didn't I hear you rise in indignation in response to a slur on "certain individuals of a certain nation"? My, my, how quickly does your attitude change...
But a more intersting question: why in the world do you think that I am an American? Slashdot, AFAIK, is a fairly international place. Kaa
An embargo on Pentum IIIs??? This got to be one of the most stupid things I've heard lately. Even putting aside the fact that all network cards have a factory built-in GUID, this idea is totally and completely unworkable. "You, sir, I see your computer has a PIII chip in it. I am afraid we cannot allow you to use this computer, sir." Not to mention the fact that, if implemented, this embargo will kill Europe's IT industry quicker than Windows gets to BSOD.
I suspect that some low-level French bureaucrat got drunk, forgot that Europe doesn't have a microprocessor industry and decided to stick it to those Yankee bastards. And privacy -- since when the EU gives a damn about privacy? Yes, I know about the laws that limit data gathering, but all these laws do nothing to limit the government's power -- and I'm still wary of the government much more than of corporations...
I just sent a Linux system to my sister, and she didn't even ask why it was running Linux! She wants to browse the web, and play some games - and for that she thinks "the computer" is cool.
Uh-oh. If playing game is going to be one of the major uses of the machines, Linux starts to suck very badly...
Slashdot Mirror
The 'seamless hyperconnected information space' that a single page formed from multiple machines might represent, is in reality a simple web page with six ads and an animated cursor, all firing off demographic information to various companies.
So are you advocating that the lowest common denominator be enforced on everybody? Sure there are plenty of crude pages with six and more banners -- but that is not a good reason to ban everything more complicated than single-server plain-vanilla HTML...
Kaa
A prompt pops up. "Do you want to add 209.207.224.245 to slashdot.org's trust realm?"
You know what? For a while I ran my browser with the setting "ask me about cookies" turned to "on". On the one hand, it was good and useful -- I knew who wanted to place a cookie on my machine and could always tell them to fuck off. Unfortunately, I tired rather quickly of clicking "no", "no", "no", on all those pop-up dialog boxes that appeared with dreadful regularity. Solution? I installed junkbuster and now live in peace.
I think that asking every time about the trust realm is going to irritate most about everybody. Instead I propose a side-window or a panel that shows you whom does your browser connect to (optional, of course). Add to this a built-in ability to exclude certain machines and domains (a la junkbuster) and all this becomes reasonable. If you care, you'll look and exclude. If you don't you'll ignore it.
Another solution is to use ZeroKnowledge's Freedom -- I'm perfectly happy to let the markedroids collect noise (what they believe to be info) about my nyms...
Kaa
Asking the average user to do this to ensure security is of course ludicrous.
Of course. But there are a lot of people out there (actually, mostly here, on Slashdot) that are quite competent to handle tcpdump and are quite interested in what exactly does their browser do. It only take a single person to sound the alarm.
A bad thing for whom? Doubleclick or the user?
The user.
What I propose is that I have direct control over the network activity that my machine is initiating.
You are welcome. Just telnet www.webserver.com 80 and type away. For most users that's waaay over their heads. They know that if you click a link, a page will appear. They are not concerned (and should not be) which machine is serving the page, whether it is one machine or several, whether the server is real or there is an akamai/sandpiper server cloud generating the content, etc. etc.
I take issue with the implication that my browser should be a black box that is free to perform arbitrary network connections on my behalf. THAT is a Bad Thing, IMHO.
Well, don't use the web, then. The web is not a system of point-to-point connections as you seem to believe -- it is a multiconnected network and I see no good reason to impose this huge constraint on it: that a page cannot contain information from more than one machine. To repeat myself, the web is meant to be a seamless whole. You are arguing for the days of the UUCP connections where yes, you did know exactly when and to whom you were connected.
Instead of a simple 'explicit host only' solution, a 'trust realm' scheme may be effective
Well, "host only" solution would break most of the web immediately. I will not argue against including this choice in a browser (choice is always good) but switching this on will make the browser basically unusable.
A "trust realm" scheme suffers from the usual defect: who will define the trust? I, as a user, certainly don't want to be bothered with it and if a site will be able to define its own trust realm we're back to square one since that's mostly what you are trying to avoid.
Using junkbuster and selecting the proper domains to block will give you most of what you want anyway.
Kaa
Detecting this kind of activity should be trivial. A nice feature for browsers (or any other net-aware software) would be to have an 'honor only explicit requests' mode, which would allow the browser to open connections only to the host that you've specified. No remote images, no remote IFRAME pages, no extraneous sockets opened up in applets or other embedded controls. I think this would neatly solve all sorts of network security problems. (As well as all sorts of advertizing 'problems'.)
Why should your browser be connecting to machines that you didn't ask it to?
Well, that's not trivial at all. Think about what the web is supposed to represent -- a seamless whole where the boundaries between machines that serve you information are irrelevant. What you propose would kill all web pages that pull content from more than one machine and that is a Bad Thing.
I'm afraid that this solution, to quote somebody's sig is simple, elegant and wrong. Looking at tcpdump output is still the preferred way to handle these issues.
Kaa
For people interested in what TC calls "zenning" and what is usually called "the flow" check out the work of a guy with the improbable name of Csikszentmihalyi (search on Google) who is usually credited with first researching the concept. One place to start is www.flownetwork.com.
Kaa
There is not logical notion that human kind has any implied fuzzy quasi-telepathic state wherin they gain "mystical" powers.
Nobody is saying anything about mystical powers. This is one of the so-called altered states of consciousness, specifically one in which you can achieve and maintain high concentration for a long period of time. It is often called 'flow'. The existence of such a state is widely recognized and documented. For example, being able to go into 'flow' is one of the characteristics of a world-class athlete in individual-competiton sports (martial arts, tennis, etc.)
Kaa
www.calcaria.net
Kaa
I think it's outrageous that large companies would even think to do this
Welcome to the real world, pal.
I don't want to sound innocent, they're obviously doing this for profit
You mean this never crossed your mind before? Corporations exist to make a profit -- that's their purpose in life.
they must realize that it will backfire when everyone gets disgusted by their tactics
Not necessarily. Random J. Luser is quite happy to have all the information about him collected, stored, indexed and catalogued. He is told that this is for his own benefit so that he can have personalized service (i.e. most effective methods for separating him from his cash/bank balance). Poll after poll show that Americans (at least) are not all that much concerned about privacy erosion. I am afraid that people who take privacy seriously will soon get hit with a "paranoid nut" label. I've already seen it used.
We are not numbers!
No, you are a string. Welcome, X65U776Z15!
Kaa
Gas based life forms don't exist
Anywhere in the universe? And how would you know that?
because of how hard it is to bind to most gases
Oh, yeah. Oxygen especially, right?
Carbon based life forms are abundant...
Well, here on Earth they are more than abundant -- they are the only game in town. Elsewhere in the universe? I have no clue, and neither do you.
Kaa
Just imagine a super slimy goo that responded to our radio signals with video streams of what Goo life was like
Just watch C-SPAN.
Kaa
The current bottleneck in all the mobile/wearable devices is battery power. Batteries are too heavy and last too short a time. The superslim notebooks -- ones that you can actually carry around -- survive for 2-3 hours at most without an electric outlet. A variety of PDAs can go through the day without recharging, but they all lack decent keyboards and tend to be quite limited in what they do.
Basically, for wearable computers to take off, batteries have to become much lighter.
Kaa
For all you people drooling about the possibilities of controlling machines (computers) purely by the mind, stop and thing for a second. If your thoughts are going to be controlling anything in any reasonable fashion, you'd better be concentrating real hard! It's perfectly possible to interface with a computer (keyboard/mouse/monitor) and think about a couple of different things at the same time. Well, it's not going to work any more. Imagine a worker controlling, say, a factory robot by his thoughts. If any time the worker thinks about beer or [insert your favorite porn star here] that robot will jerk, or stop, or drill the wrong thing, then the factory better start recruiting tibetian monks for its workers.
The point is that thought control of any sophistication requires not only the neural interface. It also requires a lot of concentration and mind discipline. Operating stuff by mind control is going to be hard.
Kaa
ever had an infected appendix? Would you think twice about having your stomach cut open and having it removed if your doctor told you it needed to be done? Do you know anyone who would? The only reason we don't all have it done (literally) is cost.
Bzzzzt. Sorry, bad argument. I would agree to have myself cut open and my appendix removed because the alternative is dying. If my doctor told me he wanted to remove my appendix just because he couldn't see any use for the damn thing, I would laugh long and hard and then change the doctor. It has nothing to do with cost, but is somewhat connected to the Murphy's law, which, loosely paraphrased, says "If anything can go wrong, it will".
Having it removed pre-emptively would be the safest thing to do, since there would then be no chance of infection.
Well, no. You are trading off a probability of your appendix becoming inflamed at the time when you are far away from medical facilities (as in cannot get to a hospital within 12 hours or so) against the probability of something going wrong during the operation (as in suddenly developing an intolerance of anesthesia). The choice is not at all clear-cut and depends on a lot of things, like how far from the hospital you live.
Besides, how do you know the appendix is useless? Contemporary medical science still does not understand a lot of things about human body.
Kaa
In the US at least their is some sort of weird moral structure that says it is okay to "fix defects" but not to improve healthy people.
That's an old Judea-Christian hangup, nothing to do with the US specifically. Basically, the Big Three middle-eastern relgions (Judaism, Christianity and Islam) believe that since God created humans in his own image, trying to improve on his creation is a sin (of pride, mostly) and is to be forbidden. You will rarely find this spelled out explicitly, but the attitude is quite prevalent throughout the Western culture, even if most people cannot explain why they think so.
Kaa
First, I should define an adaptive knowledge-based system. Basically, a KBS is a big database search program
:) concept -- for example, the plain-vanilla feed-forward nets are very different from Kohonen nets which are different from, say, gated experts architecture. And whether something runs in real time, or is implemented in hardware is really irrelevant to the discussion.
Oh, OK. You mean what I would call an expert system.
In fact, it's been years since I've seen a backpropogation or Bayesian NN used inside these walls. We've been using realtime systems closer to CORE, which are not simple back-to-front-and-learn-backwards networks. They are latticed in all directions, a lot of them are at least partially hardware.
So? Backprop is admittedly a very inefficient search method, so barely anybody uses it any more. Besides, a "neural net" is a very fuzzy
You seem to have the misconception that an NN-based system outputs a number
More, I insist on it -- only not *a* number, but a set of numbers. Unless you are doing analogue neural net (which I'll admit I never heard about) there is no way you can avoid the fact that the output of the net is numbers.
Like a robot that demonstrates a fear of light for instance
And this is a big deal? I can build one out of my Lego Mindstorms set.
NN systems are not merely a statistical system, but an approximation of the brain
Bullshit. Historically, NNs were developed as an approximation to the animals' nervous system, but that's no longer relevant. Contemporary NNs are statistical models, typically with a very large number of parameters and sometimes with interesting search strategies in parameter space. If you are going to claim that a NN is an approximation of the brain, I'm going to claim the same for my projection pursuit regression (of which a three-layer feed-forward net is a special case). Unfortunately, it doesn't sound as cool.
we've seen some interesting things happen, and quite often, the NN jumps to a totally unexpected conclusion
To repeat myselc, your inability to predict the outcome does not say anything non-trivial about the sophistication of the system you are observing.
We've seen time and again hardware systems succeed where a software model of the same model fails. Why? Chaos. Computers are terrible at calculating chaos. Hardware, on the other hand, has lots of little transmission delays.
You are not making much sense. Hardware systems typically succeed because they are orders of magnitude faster. And what do you mean by chaos? There are some fairly precise definitions of chaotic systems, but I don't think you have them in mind. Are you talking about analytic solutions of the lack of them? And it's not like it's hard to implement delays in software -- again, the main difference is speed.
But I hope that clears up some of the common misconceptions
Thank you for enlightening us, peons.
Kaa
for instance not ANY Austrialian's computer can be tapped ie: (i) allowing the Minister to issue warrants which may authorise ASIO to:
/shouldn't/ ( key word ) shouldn't be snooping around Joe Smoe's computer for no good reason.
- remotely access data that is relevant to security and is stored in a computer specified in the warrant;
hence, there has to be a warrant issued, so they
First, in the US the search warrants are asked for by the executive branch (cops) but are issued by the judicial branch (judges). This is supposed to keep the cops more or less in line (and somewhat does). However here it seems that the Minister (=executive branch) is going to be issuing the warrants. If so, he can perfectly well delegate his authority to lower levels until, say the section head at ASIO will be able to issue warrants and it'll be perfectly legal.
Second, ASIO may be snooping around Joe Shmoe's computer for a good reason -- to them. That does not necessarily make it a good reason to me, to Joe Schmoe, or say, to a "reasonable person". Maybe ASIO decides that running Linux strongly correlates with anti-government sentiments, so there is a "good reason" to periodically run checks on those weirdo Linux guys...
Kaa
What good will encryption do when they can change the data to whatever they want?
Unless you break the encryption, you *cannot* change the encrypted data to whatever you want. Doesn't mean that you cannot add unencrypted stuff, of course...
On the other hand, that might look a bit strange to a judge: "Yes, Your Honor, the accused had his whole hard disk encrypted except for this single file where he admitted to planning anti-government activities".
YARTEE (Yet Another Reason To Encrypt Everything).
Kaa
It wouldn't surprise me if they outlawed the use of non-government-sactioned encryption products next. I'm really surprised that no one in the US has tried that.
They thought about it. Unfortunately for them, in the US at least, doing so would involve major violence to the constitution and so far the consensus is that the Supreme Court will laugh long and hard and then throw it out (first amendment and all that).
Doesn't mean the situation will stay the same forever, though.
Kaa
If the U.S. government did what the Australian government just did, I'd pull the plug on my internet connection permanently. A big middle finger to those idiots. I'd rather be without the internet than to have the government in my computer.
You're bluffing, I don't believe you. Besides,
(1) The government would be happy for you to disconnect from the net: another malcontent made powerless.
(2) You can get the same result by closing all ports, running no servers at all, reading e-mail in a pure ASCII text reader, and browsing the web with something that does NOT understand Java or Javascript.
Kaa
I think that the intent of the "changing data" provision is backdoors. Once they got into your system, they want easy access from now on. However, just because this is the original intent, there is no guarantee that soon enough some "smart" guy figures out that altering data can make their life much more easy... And even if it doesn't stand up in court -- what a wonderful tool for pressure tactics. It's the same "drop a packet of coke/heroin/etc. during the search, find it, and lean on the guy real hard" technique.
Kaa
So does this mean that if I am an Australian and I detect a network intrusion that comes from a government machine, I am supposed to spread my legs and bend over?
Kaa
The one thing I can tell you right now is that you DO get robots with personality disorders. In one of my preliminary tests, I had a computer vision system go into "complete negative lockdown", which is the equivalent to suicidal depression.
When you work with neural nets, you do get to use nice metaphors. However, personality disorders they didn't get to yet. A neural net is a statistical model, a system to approximate the relationship between the inputs and the outputs. Sure, I can call a NN that fails to fit the data and example of clinical idiocy, or say that one which goes into oscillations is manic-depressive. But all these are just convenient labels for my bored mind and are not different at all from calling a car that doesn't start well in the morning cranky and grumpy.
But the point is, from the second I switch on a neural network, I hardly have the foggiest notion what it'll be like. The same matrix that went negative-lockdown stayed "sane" for about a full minute when I tried the same test with the room lights darker.
That has nothing to do with intelligence or sentience. Any chaotic system can do this easily. Throw a handful of sand in the air -- do you have the foggiest notion of where the sand particles end up? No? Does this make sand intelligent?
However, most of my decent tests of neural systems all point to the existance of true artificial intelligence in them.
Did you mean decent or recent? Anyway, define "true artificial intelligence" and then it'll be possible to talk about it. People seem to think that if a piece of software generates some behaviour not hardcoded into it by the programmer, it means that the software is intelligent. Unfortunately, no. The problems that AI's been having over the last 30 years is proof of that.
While other experimental techniques like genetic algorithms, fuzzy logic and adaptive knowledge-based systems do still have the capability to mess up in ways the designer did not intend (your favorite search engine is proof of that), neural networks (which most experts believe is the most promising currently) have the capability for a MUCH larger range of reactions.
Buzzword-o-rama! Genetic algorithms are just a global optimization mechanism. Fuzzy logic is just a way to talk about partial membership of sets. Adaptive knowledge-based systems -- hard to say what you mean. I'd probably call neural nets a subclass of adaptive learning systems.
And what do you mean by a range of reactions? A neural net outputs a bunch of numbers. Do you mean that a NN can output more numbers? or that you can interpret the NN's numbers in more imaginative ways?
Just the opinion of somebody who dabbles in AI professionally.
Ditto.
Kaa
So, basing yourself on some Babelfish translation of a German article written by journalists about some fictional implications of some EU technical commitee, you already blame certain individuals of a certain nation?
You are a clear example of your own law. You are an ignorant and an idiot.
A bit uptight today, aren't you? I would hazard a guess that something long and straight has been shoved up your ass recently.
As to certain individuals of certain nations, it happens to be a demonstratable fact that the French tend to be the most anti-American of all Western European nations.
[Little bit of chauvinism: this does not surprise me since it comes from an American. Most Americans are semiliterate and don't even know where their country is on a map of the globe. No wonder their universities and research labs are staffed with foreigners.]
Didn't I hear you rise in indignation in response to a slur on "certain individuals of a certain nation"? My, my, how quickly does your attitude change...
But a more intersting question: why in the world do you think that I am an American? Slashdot, AFAIK, is a fairly international place.
Kaa
An embargo on Pentum IIIs??? This got to be one of the most stupid things I've heard lately. Even putting aside the fact that all network cards have a factory built-in GUID, this idea is totally and completely unworkable. "You, sir, I see your computer has a PIII chip in it. I am afraid we cannot allow you to use this computer, sir." Not to mention the fact that, if implemented, this embargo will kill Europe's IT industry quicker than Windows gets to BSOD.
I suspect that some low-level French bureaucrat got drunk, forgot that Europe doesn't have a microprocessor industry and decided to stick it to those Yankee bastards. And privacy -- since when the EU gives a damn about privacy? Yes, I know about the laws that limit data gathering, but all these laws do nothing to limit the government's power -- and I'm still wary of the government much more than of corporations...
Kaa
I just sent a Linux system to my sister, and she didn't even ask why it was running Linux! She wants to browse the web, and play some games - and for that she thinks "the computer" is cool.
Uh-oh. If playing game is going to be one of the major uses of the machines, Linux starts to suck very badly...
Kaa