The ION-ETM system adds the IONTM Eye Control Software to the basic software for head control, and allows full control of a computer with your eyes. Two tiny cameras in the headset observe both your eye and the beacon on your monitor, allowing the computer and the IONTM Eye Control Software to determine where you are looking on the screen. As with ION-HTM, the ION-ETM system also senses intentional blinking, and uses that for clicking and dragging.
My main worry about IVHS is always this: will these systems see me on my bicycle? Stoplight sensors -- the only automatic vehicle detection system which has been widely deployed -- have always had trouble seeing bikes. That's actually getting better because the electronics are getting more sensitive. But I worry about the future systems, especially those that will eventually be controling vehicles. I'm never going to carry much of a transmitter because of power and weight limitations, and I don't present a very large radar or sonar cross-section.
The real root of our car problem is cheap fuel and sprawl development. In the long run, IVHS won't reduce accident rates or congestion anywhere near as effectively as a $10/gallon hike to the gas tax.
The previous article was about how the rumor was initiated by a microsoft employee who was (very naively) pretending to be a third party. He lied about who he was, but he what he said about the code is confirmed by the current piece. Here's the relevant sentence from the previous article:
In his message, he asserted that America Online is using a programming error that has created a security flaw -- one not found in Microsoft's clone program -- to detect the Microsoft Messenger program.
If they want to take a look at my computer, I'll gladly cooperate. All they have to do is show up at my door and with a warrant that they will show me, and I'll gladly let them look at whatever they want.
If they show up at your door and want to search your computer, you'll be lucky if they let you watch from across the room. More likely, the officers at your door will not be computer-competent, and they'll take your hardware with them so they can have someone who knows more look at it -- no chance to make a backup, never mind if you need it to work, and your "receipt" documenting what they took might just be a picture of the box.
Personally, having them get a warrant try to crack in bothers me a lot less than a physical search. At least this way you keep physical control of your box and you have a chance to defend it or detect the intrusion. And you can be sure that the officers doing the search are at least as competent at a keyboard as your average script kiddie.
I don't like the idea of having my privacy invaded this way; I don't imagine anyone does. It would be better if you were notified of the warrant after the fact -- permanently sealed records keep voters from knowing about patterns of abuse. But of all the proposals I've heard for helping law enforcement cope with computers, I'd say this is most benign.
Natural selection could still be considered "theory." There are several variations on the theme. There is not, as yet, a single theory of natural selection that has achieved massive concensus. This is a topic of much ongoing research.
I don't understand how anyone could say that natural selection is in any doubt. Yes, there are varying ideas about the patterns of change in evolution caused by natural selection, notably punctuated equilibrium vs. uniformitarianism. But natural selection itself is simple. The necessary assumptions can be observed in everyday experience; that it can produce signifigant change can be seen in a human lifetime and demonstrated in a laboratory to produce appreciable changes (e.g. DDT resistant mosquitos and antibiotic resistant bacteria) and demonstrated in a laboratory (varieties of broccoli which produce mature seeds in 18 days). It is awfully hard to deny. In fact, the creationists who make the decision in Kansas specifically included "micro-evolution", which is just natural selection.
Here's what you need to have natural selection
individuals aren't all the same
some of these differences are hereditary
some of the hereditary differences affect fitness (defined in terms of offspring produced)
Can anyone deny that real populations have these properties? Given these properties, an anyone deny that natural selection doesn't happen?
Most creationists have had to admit that natural selection happens. What they won't accept is that they, and all the species they see around them, descended from a common ancestor.
Not everything on your redhat box is editable from the command line. Try managing your RPM database by hand. Whups, it's a berkeley db database, ain't it?
While it might be rather difficult to manage the rpm database with a text editor, it isn't hard at all to manage it from the command line. I can't stand RedHat's GUI front ends to rpm, but I get along with rpm itself just fine. Try something line "rpm -qfi/bin/ls".
As I read this, most of what they're talking about is R&D for tools to detect intrusions and attempted intrusions at the host and LAN level in -- things decent sysadmins already use, like tripwire and portscan detectors. They're also talking about tools for correlating the alerts these detectors send out. Correlating them at the LAN or organization level seems to me to be a reasonable thing for a sysadmin to do. Correlating them nationwide is only a problem if they are gathering more information than they're supposed to -- if we insist on an open protocol and open-source, it's not a privacy threat.
Here's the kind of thing they're worried about: imaging a virus or worm whose payload is a packet sniffer. Mostly it spreads as quietly as it can, but when a copy finds itself on a host in the target domain, it starts sniffing for l/p pairs and other critical information. When it has them, it sends out an http request. I'll leave the contents of the response to your imagination.
Is this so implausible? Could it do serious damage to your organization? Do you know how to prevent it?
AI went through a spell, sometimes called the "AI winter", when it was hard to get support to do R&D (unless you hid the fact that you considered it to be AI -- lots of products, from games to elevator controllers, use techniques derived from AI to solve fairly simple problems). That's since eased somewhat, though as far as I know nobody's aiming at the Turing test right now. Nevermind having the intelligence of a dull chimp -- we're at the cockroach level, and accurately imitating mouse's brain would be a breakthrough
Symbolic (traditional) AI got into trouble for two substantial reasons. First, the real world is a noisy place and you can never see all of it. So there is always uncertainty and ambiguity in the input, and strictly symbolic AI doesn't deal with uncertainty very well. Fuzzy logic tries to patch this, but it doesn't seem to scale very well. What's needed is to mix in some of statistics, which exists to cope with ambiguous data. One of the newer AI inference technologies does this quite nicely; it's called Bayesian Inference. (My work involves Bayesian inference, and I don't claim to be unbiased. It's works, and it feels right.)
Second, AI needs some knowledge about how the world works -- a knowledge base -- in order to reason about its input, and building a knowledge base is usually a major project. Writing a knowledge base is essentially programming in a wierd special-purpose language, and the rest of the project may be as simple as writing I/O wrappers for a generic inference engine. (If you want to try it, get clips, an expert-system shell, built around a rule-based inference engine.) Building knowledge bases is gradually getting easier, just as programming got easier after people had been doing it and building better tools for a couple decades. And machine learning techniques can do some amazing things, but with some caveats. How well it learns, and what it learns, depends on how you set up the learning problem and how you package the input. And I don't know how well the techniques scale.
Cye is cute, but it has nothing to do with state-of-the-art AI. I'd call it a pretty impressive example of minimalist engineering. The gear-shaped wheels both give traction and minimize navigation uncertainty, and they use the motor fan blades as shaft encoders. IIRC, it has one board and two moving parts. I don't see why it costs more that $75.
This is a very good point, and not just for software and system configuration. The same principle holds for policy about what's classified (proprietary, trade secret, whatever). It's seemed to me for quite a while that US government policy is to classify as much as possible, so that the really valuable information is a needle in a classified haystack. This leads people to lose respect for the classification.
Yet, we see in the NYT article that even the systems that reside on the Secret / Most Secret "air gap" networks are running MS software. WHY IN THE HELL ARE THEY DOING THAT?!?!?
Microsoft Office. It's not just the number-crunching that needs to be classified -- the resulting statistical analysis, report, and presentation are classified too. In fact, many people with classified data on their computers -- probably a majority -- have it there for communication only.
For an amazing fraction of the people in this country, these kinds of tasks imply use of MSOffice. And the rest of us have to communicate with them.
As I understand it, creating clone embryos, or even genetically modified embryos, is much easier than creating the sort of gestation tank that is required for scenarios like Brave New World and Cyteen. If you don't have automated gestation, or at least gestation by a non-human mother, you'll have human mothers who are attached to their childern and who won't like the idea that the kids they bear are will be born to be slaves. Surrogate mothers have done this already, an several famous court cases have been fought about it.
This is much more important to those scenarios than cloning, because if you do can gestate (and raise) the kids without an opportunity for anyone to have an emotional attachment to them, then it will be dangerously practical to raise slaves, and it won't matter whether they're clones or not.
The application that comes to my mind right away is viewfinders for digital cameras -- with these things, you could have an SLR style through-the-lens viewfinder. (They're certainly the right size, anyway.) And perhaps they use enough less power than the usual 2" LCD that that the batteries would last a while.
Really, though, is there any point to continuing to make screens smaller after they need a magnification system larger than they are?
This is a forced sale, not a theft. Because it's an auction, MS gets a chunk of money out of it. Theoretically, other companies will bid approximately the present value of the stream of profit they expect from having the license. In a competitive market, this would be about equal to MS decrease in profit, so it's arguably pretty fair.
Mind you, MS stockholders will probably lose some on the deal. Because the market is not competitive, the profit stream is bigger in MS hands than it would be in someone elses hands after the sale. And a fair price would be a big chunk of MS's market cap -- who could come up with that much cash?
The analogy someone made above to eminent domain seems pretty apt -- fair in theory, but mighty unpleasant. And I'm not convinced it will improve the world one bit. It certainly doesn't address the office-suite monopoly, which is more important now than the OS monopoly.
Slashdot Mirror
The ION-ETM system adds the IONTM Eye Control Software to the basic software for head control, and allows full control of a computer with your eyes. Two tiny cameras in the headset observe both your eye and the beacon on your monitor, allowing the computer and the IONTM Eye Control Software to determine where you are looking on the screen. As with ION-HTM, the ION-ETM system also senses
intentional blinking, and uses that for clicking and dragging.
The real root of our car problem is cheap fuel and sprawl development. In the long run, IVHS won't reduce accident rates or congestion anywhere near as effectively as a $10/gallon hike to the gas tax.
In his message, he asserted that America Online is using a programming error that has created a security flaw -- one not found in Microsoft's clone program -- to detect the Microsoft Messenger program.
If they show up at your door and want to search your computer, you'll be lucky if they let you watch from across the room. More likely, the officers at your door will not be computer-competent, and they'll take your hardware with them so they can have someone who knows more look at it -- no chance to make a backup, never mind if you need it to work, and your "receipt" documenting what they took might just be a picture of the box.
Personally, having them get a warrant try to crack in bothers me a lot less than a physical search. At least this way you keep physical control of your box and you have a chance to defend it or detect the intrusion. And you can be sure that the officers doing the search are at least as competent at a keyboard as your average script kiddie.
I don't like the idea of having my privacy invaded this way; I don't imagine anyone does. It would be better if you were notified of the warrant after the fact -- permanently sealed records keep voters from knowing about patterns of abuse. But of all the proposals I've heard for helping law enforcement cope with computers, I'd say this is most benign.
I don't understand how anyone could say that natural selection is in any doubt. Yes, there are varying ideas about the patterns of change in evolution caused by natural selection, notably punctuated equilibrium vs. uniformitarianism. But natural selection itself is simple. The necessary assumptions can be observed in everyday experience; that it can produce signifigant change can be seen in a human lifetime and demonstrated in a laboratory to produce appreciable changes (e.g. DDT resistant mosquitos and antibiotic resistant bacteria) and demonstrated in a laboratory (varieties of broccoli which produce mature seeds in 18 days). It is awfully hard to deny. In fact, the creationists who make the decision in Kansas specifically included "micro-evolution", which is just natural selection.
Here's what you need to have natural selection
- individuals aren't all the same
- some of these differences are hereditary
- some of the hereditary differences affect fitness (defined in terms of offspring produced)
Can anyone deny that real populations have these properties? Given these properties, an anyone deny that natural selection doesn't happen?Most creationists have had to admit that natural selection happens. What they won't accept is that they, and all the species they see around them, descended from a common ancestor.
While it might be rather difficult to manage the rpm database with a text editor, it isn't hard at all to manage it from the command line. I can't stand RedHat's GUI front ends to rpm, but I get along with rpm itself just fine. Try something line "rpm -qfi /bin/ls".
Here's the kind of thing they're worried about: imaging a virus or worm whose payload is a packet sniffer. Mostly it spreads as quietly as it can, but when a copy finds itself on a host in the target domain, it starts sniffing for l/p pairs and other critical information. When it has them, it sends out an http request. I'll leave the contents of the response to your imagination.
Is this so implausible? Could it do serious damage to your organization? Do you know how to prevent it?
Symbolic (traditional) AI got into trouble for two substantial reasons. First, the real world is a noisy place and you can never see all of it. So there is always uncertainty and ambiguity in the input, and strictly symbolic AI doesn't deal with uncertainty very well. Fuzzy logic tries to patch this, but it doesn't seem to scale very well. What's needed is to mix in some of statistics, which exists to cope with ambiguous data. One of the newer AI inference technologies does this quite nicely; it's called Bayesian Inference. (My work involves Bayesian inference, and I don't claim to be unbiased. It's works, and it feels right.)
Second, AI needs some knowledge about how the world works -- a knowledge base -- in order to reason about its input, and building a knowledge base is usually a major project. Writing a knowledge base is essentially programming in a wierd special-purpose language, and the rest of the project may be as simple as writing I/O wrappers for a generic inference engine. (If you want to try it, get clips, an expert-system shell, built around a rule-based inference engine.) Building knowledge bases is gradually getting easier, just as programming got easier after people had been doing it and building better tools for a couple decades. And machine learning techniques can do some amazing things, but with some caveats. How well it learns, and what it learns, depends on how you set up the learning problem and how you package the input. And I don't know how well the techniques scale.
Cye is cute, but it has nothing to do with state-of-the-art AI. I'd call it a pretty impressive example of minimalist engineering. The gear-shaped wheels both give traction and minimize navigation uncertainty, and they use the motor fan blades as shaft encoders. IIRC, it has one board and two moving parts. I don't see why it costs more that $75.
This is a very good point, and not just for software and system configuration. The same principle holds for policy about what's classified (proprietary, trade secret, whatever). It's seemed to me for quite a while that US government policy is to classify as much as possible, so that the really valuable information is a needle in a classified haystack. This leads people to lose respect for the classification.
Microsoft Office. It's not just the number-crunching that needs to be classified -- the resulting statistical analysis, report, and presentation are classified too. In fact, many people with classified data on their computers -- probably a majority -- have it there for communication only.
For an amazing fraction of the people in this country, these kinds of tasks imply use of MSOffice. And the rest of us have to communicate with them.
This is much more important to those scenarios than cloning, because if you do can gestate (and raise) the kids without an opportunity for anyone to have an emotional attachment to them, then it will be dangerously practical to raise slaves, and it won't matter whether they're clones or not.
Really, though, is there any point to continuing to make screens smaller after they need a magnification system larger than they are?
This is a forced sale, not a theft. Because it's an auction, MS gets a chunk of money out of it. Theoretically, other companies will bid approximately the present value of the stream of profit they expect from having the license. In a competitive market, this would be about equal to MS decrease in profit, so it's arguably pretty fair.
Mind you, MS stockholders will probably lose some on the deal. Because the market is not competitive, the profit stream is bigger in MS hands than it would be in someone elses hands after the sale. And a fair price would be a big chunk of MS's market cap -- who could come up with that much cash?
The analogy someone made above to eminent domain seems pretty apt -- fair in theory, but mighty unpleasant. And I'm not convinced it will improve the world one bit. It certainly doesn't address the office-suite monopoly, which is more important now than the OS monopoly.