Let me clarify one other point about the gold bead technology:
The limitations on our ability to detect RNA binding to a microarray (or DNA, but PCR is more sensitive) is NOT the minimum amount of RNA that we can detect from it's fluorescent tag. The problem is non-specific annealing to the target sequence; this is also a major problem in PCR. Every spot on the array "lights up" with a certain background intensity and if the RNA you're looking for is present in lower concentration than that, it is below the detection thresh-hold. So, even if those gold beads have a lower absolute detection thresh-hold, which I doubt but it's certainly possible, it doesn't matter because it's the background noise, which they do not promise to reduce, that actually limits your ability to detect signal.
You can buy antibody coated strips (for detecting anthrax) for $20, as mentioned in a previous post. Simple organic extraction gives you good enough samples to work with, but that's simple = an undergrad can do it not simple = a box sitting on a desk can do it internally and automatically; recall that we're talking about desktop detectors, here.
I want to make a distinction between efficient and sensitive. Yes, nothing is more sensitive than PCR. However, PCR does not always work; the unexplained failure rate is actually rather high and the technique can be exquisitely sensitive to contamination by similar sequences.
I cannot concieve that the gold bead technology is more sensitive than PCR. It's absolutely going to be faster.
Re:desktop not what you think it means
on
Desktop Biodetectors
·
· Score: 2, Interesting
As I continuously repeat, often hopping up and down and foaming at the mouth, to people who do not (I do) work with DNA:
DNA IS NOT THE MOST EFFICIENT WAY TO DETECT ANYTHING. IT IS INSIDE THE CELLS! THE CELL WALLS ARE STUDDED WITH ANTIGENS (things to make antibodies) AND THEY ARE ON THE OUTSIDE. THE CHOICE IS CLEAR.
I'll wait until I see hard numbers on the technology with the beads; if it really is more sensitive than the photodetection on present day DNA microarrays, than that's very promising. However, I'd like to see an independent group assay the technique's sensitivity before making any judgements on it's utility - I strongly suspect that it will turn out to be actually less sensitive, based on the (scientifically irresponsible) comments made by the crackpots and or sleaze at the company that produces it.
Poisons for example. I didn't read anything about that, how can they be detected?
Short answer: They can't.
Long answer: Some can, but you need specific methods to deal with each one. Most highly lethal airborne toxins are organic; in order to detect them, they need to be differentiable from the organic compounds that we human beings (and paper, and cleaning supplies, and so on and so on) put off into the environment all the time. Most airborne toxins can be differentiated, but the method to do so is different for each and every one. Any method that reliably detects a wide range of airborne toxins falls into the "false positive" problem - it's going to be going off continuously as things which are not toxins fool it for some reason or another.
So, you could build a machine that detects Serin nerve gas (which has, I'm just making this up, a very different UV absorption spectrum from what you usually have in the air; you take a sample periodically and check it's UV absorbance.) Okay, there are a certain number of harmless compounds, or combinations of harmless compounds, that duplicate the UV absorption and fool the assay. If you expand the assay to include mustard gas and cyanide, the number of harmless compounds that give false positives goes up. Anything that detects a wide range of harmful compounds is going to be going off continuously because your cup o ramen sent vaporised Yellow #5 floating through the air.
So, if you allready have reason to believe that there is Serin in the air, you can detect it. If you want something that sits on a table and detects the most likely/common poison gases, you can do that, with only occasional false positives. However, someone smart enough to make Serin is probably smart enough to get a copy of your list of common nerve gasses and make something else.
Finally, some never gases are lethal in such tiny quantities that there really is no way of "abiotically" detecting them without allready knowing that they're there (biotic detection = people keeling over dead.) Fortunately, they're difficult and hazardous to make.
Since the antibodies are to compounds exposed on the cell wall, you don't need to lyse the cells (see my previous post).
Also, antibodies are sufficiently specific in their binding that you'll get negligible false positives.
However, the detection thresh-hold is very high. These devices are very useful for (say) detecting anthrax in powdered form that was mailed to you. However, if someone put a bunch of anthrax spores in an air vent and let it blow around a building, and you had one of these things stuck up to a wall; well, aside from the fact that the antibodies degrade fairly quickly, it isn't sensitive enough to detect the minimum lethal amount of anthrax blowing around in the air.
Quick assays exist that are more sensitive, but they generate false positives, and as such are more useful for spreading panic. More sensitive assays exist that also employ antibody technology, but they're time consuming, not portable and they're *very* expensive.
Re:Too bad we don't have these things today?
on
Desktop Biodetectors
·
· Score: 3, Interesting
What are there now, somthing like under 30 cases of Anthrax so far? Sent to a few high-profile companies? Sure, that's a MASSIVE INCREASE over previous levels, but statistically you're still in pretty good shape. Don't let the media hype get you worked up.
Very true. Also one of the (many) reasons why these things won't work. I'll stick with the three simplest.
Firstly, in order to detect bacterial DNA you have to lyse (break open) the bacteria. This means you have to filter them out of the air, dissolve them in lysing buffer and apply them to your detector, at the bare minimum. Lysis buffer is expensive, and I doubt there technology actually works without doing more than lyse the cells (removing the cellular protein, much of which binds indiscriminately to DNA, would be a good start. Doing that in reasonable time requires a ~12,000 RPM centrifuge, precipitant compounds and a column, at the minimum.)
Secondly, DNA from other organisms is going to bind to your probe (including these gold bead things) with a certain frequency (this binding is called "base pairing," which is largely driven by hydrogen bonding but is not called a bond.) So, your background noise from that is going to be more than enough to drown out the signal from an anthrax concentration high enough to kill you, especially if you're standing across the room from the detector holding an envelope; the anthrax concentration drops as the square of the distance from the contaminant source unless there's a wind.
Thirdly, the reason the previous poster mentioned. In order for these things to be useful the false positive rate has to be on the same order of magnitude, or smaller, than the actual positive rate. Even if you use practical techniques instead of this absurdity with DNA, that's never going to happen. More practical techniques depend on markers on the bacterial cell walls (so you don't need to lyse the bacteria); when the CDC people report "preliminary results" indicating anthrax this is what they're talking about - these results are preliminary because there are many other, more common and harmless bacteria that have the same factors in their cell walls, and because the experiment to detect the stuff occasionally goes wrong for no apparent reason.
Even in a laboratory setting, if you want to detect the DNA from this stuff you have to *culture* it. The idea of a desktop machine, as opposed to a highly trained scientist with a lab full of sophisticated equipment, being able to detect the DNA from the amount of this stuff that is actually floating in the air is patently and absolutely absurd.
A handheld is even more so, since it wouldn't even be able to dissolve the bacteria it filtered.
These devices are either a pipe dream, a scam, or both. Either they'll just report A-OK all the time or, even worse, they'll periodically start an unjustified anthrax scare by giving off an alarm.
In this case, a primary instrument of that leverage may be something that has gotten relatively little attention in the hype surrounding the new operating system: a Web browser integrated in Windows XP called MSN Explorer.
Relatively little hype, huh? Is the reporter brain damaged? Is he incapable of remembering anything that happened more than a few weeks ago?
One question I have: I'm on TW's high speed internet service. Am I going to be shunted onto AOL?
They had something like this in each and every door (and, of course, Marvin.) Someone should take a boat, and make a mockup of Zaphod's ship.
Also, take heart. This means that the board of directors of Sony/Toyota will be first up against the wall, come the revolution (according to a copy of Encyclopedia Galactica beamed from the future.)
It's a common misconception that the Nevada desert is a wasteland. Guess what: It's not. There is a rather intricate ecosystem. Covering the desert wipes out this system.
This is true. Furthermore, there's the power transmission problem. Until we get room temperature superconductors, only California can benefit from Nevada's ecological destruction. Beaming the power from space is about as feasible as killing people with orbiting laser satellites.
The only real wasteland is in the open ocean. It doesn't look real different from other stretches of ocean but if there's no algae, you can cut off the sun (it only penetrates about 40 meters anyway). You float the things on the water. It's the only ecologically sound place to put them, and there are stretches of this sort of ocean comparatively close to all of the world's coastlines, which is where most people live.
Re:No, this is scary, not funny. I mean that.
on
RIAA to DoS Pirates?
·
· Score: 1
Picture: J Random Musiclover, uses WinMX and KaZaA, until they bog down terribly slowly.
In order for this to work, they're going to need to do denial of service attacks against Gnutella. Sure, our professors have all sold out. Sure, they're just another brick in the wall. However, do you think the former hippies who constitute the nation's CS faculty are going to take it sitting down when "the man" starts doing DoS on *their* (heavily infiltrated by KaZaA, not to mention, good lord, gnutella and freenet) computers? *Especially* when they don't even have to *do* anything, just look the other way while their students do (at my undergrad institution they do this allready, for chrissakes) their very worst?
The RIAA has enough hubris to try it anyway. They could be intelligent and not pick on anyone who might concievably fight back, but I don't think they've gotten that much smarter, and I don't think they understand at all what causes people who aren't empty suits to act or react the way they do.
I've filed a patent on any internetwork protocol ("IP") driven discussion forum centering on issues related to, but not limited to: communications technology, free speech issues, computer science, intellectual property policy, sociological interpretation of trends in technology and technology law, and recreational interests (science fiction, video games, anime et. al.) commonly associated with the "technologically savvy".
CmdrTaco, consider yourself served! I'm still waiting for all that money you owe me for continued use of your DNA.
Also, I've ammended my slashdot user agreement to include an arbitration clause (persuant to the broadest possible interpretation of the Federal Arbitration Act) so I won't actually be seeing you in a "court", per se.
They made up a student association at our U. and called on behalf of it. They faxed a message (on university letterhead) inviting the representative to come and discuss the issue with their organisation; of course, he didn't, but he certainly answered all of their mail.
Of course, they could have accomplished the same thing by joining CalPIRG; but, on one level, ten organisations with one tenth the membership seem bigger than one organisation of everybody.
So, everybody who cares on slashdot (I live in Manhattan) start faxing letters around to everybody in your district, and send in a single copy, with everyone's signature on it. Or, have everybody call on behalf of the organisation. This is what the christian right does, and it works fairly well - they heavily influence (even liberal) representatives in their districts. So, where nerds are concentated (in the NE and West Coast, I've always assumed - take a look at those graphic representations of world computer traffic from a few days ago) we ought to be able to form a political force - if we could 1. agree on anything and 2. form local political organisations on that basis.
However, I don't think it will get all the way to a slashdot-stlye free-for-all where anyone can jump in.
No, I don't think so either. While, certainly, slashdot-like forums will continue to multiply in the sciences (this is one, after all, for computer science,) I don't think they will ever generate much prestige. However, if I had an article (as opposed to a comment) which slashdot carried, I'd put it on my cv (for novelty's sake, if naught else.)
I think that the 'next wave' of this is going to involve a backlash against the diluted power of academic senates vis a vis boards of directors at major universities. We need to find new ways to exert control over the process of research and the nature of research related discourse - control which is increasingly being co-opted by non academic (or nominally academic) administrators and lawyers.
If I send those to a slashdot-style forum, my boss will laugh. They'll count something less than a one-line quote in a small town newspaper.
This is true. Even though I supported the public library of science initiative (reported earlier on slashdot, and also here , I'm about to submit an article to TIBS. An article on GenomeWeb just won't cut it.
However, that is not going to be the case forever. As data mining techniques become more available and sophisticated, that is to say, when real data mining (as opposed to just text matching) becomes a major way in which academics access content, articles in "free" journals are going to be *more* visible to your colleagues, and as important discoveries are coordinated using such techniques, citations will rise, and those journals will rise in prestige. This is going to aggravate what is allready a real schism within the academic community, of int. property versus the pursuit of truth. I think it's going to be a struggle - for the very soul of academia which is really under threat here - but I think we're going to win, because cutting out the middle-man makes for a more efficient way of sharing information, because history isn't over and materialism isn't really the driving force behind human creativity, and because people who love science for it's own sake are better at it.
An earlier poster said that the current system drives the smartest people out of academia and into industry. I couldn't disagree more - I don't think people in (the biotech) industry are very smart at all. They're kind of pathetic, mostly. I'll agree that there are a lot of frankly stupid "scientists" doing terrible work at supposedly public universities on the private dollar, and that the people who jump ship into industry are often a bit smarter than they are; but the smartest, most devoted people are still in the public sector.
Looky. A little-considered effect of all this inter discplinary convergence is that we're going to start running out of acronyms. Some poor slob has been trying to use T&A to mean "theory and application".
Well, it hasn't been good for gene research here in Manhattan. All my colleagues still look like they've been shot in the gut. Seriously. I don't really feel like going to work myself.
What do you think is the limiting ingredient in molec. bio. research? It isn't money - I hate to admit it, but we have plenty. It's competent and capable people who are interested in the field. Eight years down the line we might see more people entering this field particularly because of an interest in bioterrorism (although, if they work for the US government I worry about how much of what they do will really be "defensive".) I think it's more likely that it will draw our best and brightest into the military, intelligence, etc. Sure, we'll refocus our efforts to some extent on molec. bio. that is applicable in anti-bioterrorism, but that comes at an opportunity cost. Do you think that this knowledge is really of more benefit to mankind than plant genetics? AIDS or cancer research? Would it be of more benefit if there were no lunatics with anthrax bombs? I really don't see a silver lining to this.
If you want a really good treatment of this question (offline,) I strongly suggest Laurie Garrett's not-so-new-anymore book "Betrayal of Trust". Do your part to help the economy recover and BUY it at amazon.
Anyway, Laurie is not so confident in the efficacy of our smallpox vaccine stock as the CDC. Believe whom you will.
Why don't you read the article instead of looking out the window for black helicopters????
I really don't want to troll, but am I mistaken or is that addressed to me? I mean, I posted a link to a comic strip. Do I need to bracket it in rows of smiley faces? Meet me halfway, here.
Once someone says "it has nothing to do with encryption" aren't all future messages to the same effect redundant? Just because other people keep going on about encryption doesn't mean you have to keep upmoderating anybody who repeats that.
On a more serious note, if they really are just like us geeks on slashdot, why do they work at the NSA? Are geeks who work at the NSA happy with their jobs? I wouldn't want to solicit anything classified (since I post with my e-mail address:)) but do we know who at the NSA arranged this? It's a nice piece of work and I'm wondering how friendly an environment the NSA really is for people who want to do open source work. Is this the brainchild of one particular linux-friendly higher-up or are we going to see a lot of open source stuff coming out of the NSA?
This is not some new scheme to control the population... No doubt the people working on this are just geeks, whom are much like many of us here on slashdot.
Yeah, and this has nothing to do with encryption. They use steganography. I found the phrase "Help us! We've been captured by the NSA!" embedded in one of the header files. Story here .
At UCSC, where I was an undergrad, we had fairly numerous breadth requirements, which was excellent. The original poster seems to think that four year colleges should be more like trade schools - they should not. Four year college programs do not exist to churn out techs to work in industry - although there is certainly an effort to subvert them to do that. A graduate of such a program is _supposed_ to be an active participant in the intellectual life of society, whatever they may do for a living. So, yeah, if you have a BA in computer science you should know something about sociology, philosophy and art, not the least of how the CS industry is actually run at every level. Idea is - you need this information to make moral judgements about how to use your expertise, which is one of the things that a BA is supposed to do. If you really don't care, that's fine, I'm not criticizing you, but our educational institutions shouldn't be reduced to a rubber stamp so that people with no social conscience or inherent curiosity can get better paying jobs! Go to law school, for chrissakes.
The idea with teaching "big picture" thinking is that you will better understand the details of what you're doing if you know why you're doing them. The Poster's courses sound painfully stupid (decision theoretrics with no stat? taught by a little balding guy with pointy hair?) but that isn't a problem with breadth.
That doesn't mean that the pursuit or release of knowledge should be restricted in any way.
As a scientist I am more concerned with what other scientists are *doing* that with what they are *developing*. Our colleagues who developed the techniques to clone DNA into plant cells (a number of whom I know personally) did nothing wrong, and should not have delayed publication because of the "ethical" consideration of what someone else could do with it. The people who are genetically altering corn to make it increasingly resistant to chlorinated organics (roundup) are *doing* something unethical; and they are the ones, largely highly intelligent people, whom we need to reach and educate. Some of the things I'm attempting to do could have direct, terrible applications in germ warfare - but they could also be a great boon to medical research. The resolution of that dillemma is clear: we cannot call a halt to scientific progress because of fear.
Other scientists, and some people may draw an increasingly meaningless distinction and call them engineers, are actually applying these developments to do things that shouldn't be done. Biopreparat doesn't exist anymore, but I'm sure biological weapons research continues. The people who nerve gassed the Tokyo subway where highly educated. These people are doing more damage with their own scientific expertise than laymen ever can, or will, with something you release.
Ethics requirements at graduate schools should be specific, factual and tailored to the particular focus of the student. Individuals who want to go into plant genetics should take courses in the political economics of third world agriculture - the same ones that pol sci students take. Courses in "ethics" are substanceless exercises in sophistry (say that 10 times fast) that don't teach the consequences of the particular actions a student might actually take.
While relatively uneducated terrorists can make certain uses of publically released technologies like culturing eukaryotic cells or near unbreakable encryption, the *real* danger, and it is a real danger, is when the scientists ourselves are actually setting out to do harm; or applying these technologies in ignorance for our own economic gain.
> Besides, in any case you dumb shit, all the
> terrorists have to do is keep their noses clean!
Well, the fact is, they didn't keep their noses clean and seldom do. Since "classified" evidence can be applied to foreign nationals in immigration situations, without disclosing the evidence to the immigrant, if we could figure out who people were we could deny them entrance to the country on thin evidence, hearsay, etc. Which is not exactly an enlightened practice but it would have kept all of the people involved in the WTC out of the country, along with a lot of their totally innocent friends and acquaintances.
> We already have many forms of identification. ID's which can easily be counterfited.
This is true but isn't really a rational argument against having a single ID. A single ID, _if_ managed properly could be much harder to hack, or counterfeit. That would reduce the number of fake IDs in circulation, which would reduce crime on a number of levels. In fact, if we were willing to invest the resources needed to do so, we could ensure that hacking the database was effectively impossible; the only way to get a fake ID would be to get someone on the "inside" to help, and if the system were set up properly we could catch those people if and when any falsified ID they had personally certified was used.
It could be made to work, and, while I agree that their president is a grade A pompous asshole, Oracle really does have the technology to accomplish all of this.
Which is not to say that I think we should (I agree with previous posters about the need for a spate of privacy controls) but we definitely could and it could absolutely work.
The threat of a lawsuit may be more effective than an actual law. Sure, the guy who cracked the SDMI watermarks almost certainly would have won the resultant lawsuit (unless he got the idiot judge from the DeCSS case) but he couldn't afford to be sued... actually he was getting all Sun Tzu on the RIAA/SDMI, but still.
Since the NSA allready thinks that any cryptographer who doesn't work for them is some kind of rogue lunatic, I don't think they'd shy away from bringing charges against people they didn't like who published a development that they would rather keep secret. Sure, eventually they'd lose, but until they actually did they could keep people in prison for months or years awaiting appeals! The threat of being prosecuted for publishing such a description (which, as I said, is tantamount to a device under a hostile interpretation of the DeCSS ruling) is pretty serious. Also, while I'm sure we'll win DeCSS sooner or later, it's more than likely that the final ruling will be vague or waffling enough not to close out future prosecutions.
My understanding had been (I find an indirect reference to this so maybe the reporter is wrong) that the ruling included the Object code (in whatsoever language it was written, my friend wrote a compiler that turned the first page of King Lear into DeCSS - much more clever than a pretty picture), and the Object code _is_ just a description. Fine, so we suppose there's a difference between a description and a description-that-some-automated-routine-somewhere- can-understand. Sure, whatever.
While you _can_ just publish a truly abstract description of the algorithm, especially in a pure math journal, applied math journals generally want to see you do it. Although, I'm sure the reviewers would be sympathetic if that were _illegal_. Regardless, we know that researchers who come up with something like that are going to want to code it to test their theories - is that going to be illegal? Are they going to have to add the backdoor (even if that somehow jams up there whole gig) to make the program legal? Doesn't that mean that they'll have to be *given* the back door, whatever on earth it actually is, so that they can make sure it decodes their message? Will they have to turn the source code over to the government before publication so that the government can add the backdoor? I'm running on too long here but am I missing some obvious solution to this seeming logical disjoint, or does this outlaw any innovation in encryption at all?
Slashdot Mirror
Let me clarify one other point about the gold bead technology:
The limitations on our ability to detect RNA binding to a microarray (or DNA, but PCR is more sensitive) is NOT the minimum amount of RNA that we can detect from it's fluorescent tag. The problem is non-specific annealing to the target sequence; this is also a major problem in PCR. Every spot on the array "lights up" with a certain background intensity and if the RNA you're looking for is present in lower concentration than that, it is below the detection thresh-hold. So, even if those gold beads have a lower absolute detection thresh-hold, which I doubt but it's certainly possible, it doesn't matter because it's the background noise, which they do not promise to reduce, that actually limits your ability to detect signal.
You can buy antibody coated strips (for detecting anthrax) for $20, as mentioned in a previous post. Simple organic extraction gives you good enough samples to work with, but that's simple = an undergrad can do it not simple = a box sitting on a desk can do it internally and automatically; recall that we're talking about desktop detectors, here.
I want to make a distinction between efficient and sensitive. Yes, nothing is more sensitive than PCR. However, PCR does not always work; the unexplained failure rate is actually rather high and the technique can be exquisitely sensitive to contamination by similar sequences.
I cannot concieve that the gold bead technology is more sensitive than PCR. It's absolutely going to be faster.
As I continuously repeat, often hopping up and down and foaming at the mouth, to people who do not (I do) work with DNA:
DNA IS NOT THE MOST EFFICIENT WAY TO DETECT ANYTHING. IT IS INSIDE THE CELLS! THE CELL WALLS ARE STUDDED WITH ANTIGENS (things to make antibodies) AND THEY ARE ON THE OUTSIDE. THE CHOICE IS CLEAR.
I'll wait until I see hard numbers on the technology with the beads; if it really is more sensitive than the photodetection on present day DNA microarrays, than that's very promising. However, I'd like to see an independent group assay the technique's sensitivity before making any judgements on it's utility - I strongly suspect that it will turn out to be actually less sensitive, based on the (scientifically irresponsible) comments made by the crackpots and or sleaze at the company that produces it.
Poisons for example. I didn't read anything about that, how can they be detected?
Short answer: They can't.
Long answer: Some can, but you need specific methods to deal with each one. Most highly lethal airborne toxins are organic; in order to detect them, they need to be differentiable from the organic compounds that we human beings (and paper, and cleaning supplies, and so on and so on) put off into the environment all the time. Most airborne toxins can be differentiated, but the method to do so is different for each and every one. Any method that reliably detects a wide range of airborne toxins falls into the "false positive" problem - it's going to be going off continuously as things which are not toxins fool it for some reason or another.
So, you could build a machine that detects Serin nerve gas (which has, I'm just making this up, a very different UV absorption spectrum from what you usually have in the air; you take a sample periodically and check it's UV absorbance.) Okay, there are a certain number of harmless compounds, or combinations of harmless compounds, that duplicate the UV absorption and fool the assay. If you expand the assay to include mustard gas and cyanide, the number of harmless compounds that give false positives goes up. Anything that detects a wide range of harmful compounds is going to be going off continuously because your cup o ramen sent vaporised Yellow #5 floating through the air.
So, if you allready have reason to believe that there is Serin in the air, you can detect it. If you want something that sits on a table and detects the most likely/common poison gases, you can do that, with only occasional false positives. However, someone smart enough to make Serin is probably smart enough to get a copy of your list of common nerve gasses and make something else.
Finally, some never gases are lethal in such tiny quantities that there really is no way of "abiotically" detecting them without allready knowing that they're there (biotic detection = people keeling over dead.) Fortunately, they're difficult and hazardous to make.
These things really work, by the way.
Since the antibodies are to compounds exposed on the cell wall, you don't need to lyse the cells (see my previous post).
Also, antibodies are sufficiently specific in their binding that you'll get negligible false positives.
However, the detection thresh-hold is very high. These devices are very useful for (say) detecting anthrax in powdered form that was mailed to you. However, if someone put a bunch of anthrax spores in an air vent and let it blow around a building, and you had one of these things stuck up to a wall; well, aside from the fact that the antibodies degrade fairly quickly, it isn't sensitive enough to detect the minimum lethal amount of anthrax blowing around in the air.
Quick assays exist that are more sensitive, but they generate false positives, and as such are more useful for spreading panic. More sensitive assays exist that also employ antibody technology, but they're time consuming, not portable and they're *very* expensive.
What are there now, somthing like under 30 cases of Anthrax so far? Sent to a few high-profile companies? Sure, that's a MASSIVE INCREASE over previous levels, but statistically you're still in pretty good shape. Don't let the media hype get you worked up.
Very true. Also one of the (many) reasons why these things won't work. I'll stick with the three simplest.
Firstly, in order to detect bacterial DNA you have to lyse (break open) the bacteria. This means you have to filter them out of the air, dissolve them in lysing buffer and apply them to your detector, at the bare minimum. Lysis buffer is expensive, and I doubt there technology actually works without doing more than lyse the cells (removing the cellular protein, much of which binds indiscriminately to DNA, would be a good start. Doing that in reasonable time requires a ~12,000 RPM centrifuge, precipitant compounds and a column, at the minimum.)
Secondly, DNA from other organisms is going to bind to your probe (including these gold bead things) with a certain frequency (this binding is called "base pairing," which is largely driven by hydrogen bonding but is not called a bond.) So, your background noise from that is going to be more than enough to drown out the signal from an anthrax concentration high enough to kill you, especially if you're standing across the room from the detector holding an envelope; the anthrax concentration drops as the square of the distance from the contaminant source unless there's a wind.
Thirdly, the reason the previous poster mentioned. In order for these things to be useful the false positive rate has to be on the same order of magnitude, or smaller, than the actual positive rate. Even if you use practical techniques instead of this absurdity with DNA, that's never going to happen. More practical techniques depend on markers on the bacterial cell walls (so you don't need to lyse the bacteria); when the CDC people report "preliminary results" indicating anthrax this is what they're talking about - these results are preliminary because there are many other, more common and harmless bacteria that have the same factors in their cell walls, and because the experiment to detect the stuff occasionally goes wrong for no apparent reason.
Even in a laboratory setting, if you want to detect the DNA from this stuff you have to *culture* it. The idea of a desktop machine, as opposed to a highly trained scientist with a lab full of sophisticated equipment, being able to detect the DNA from the amount of this stuff that is actually floating in the air is patently and absolutely absurd.
A handheld is even more so, since it wouldn't even be able to dissolve the bacteria it filtered.
These devices are either a pipe dream, a scam, or both. Either they'll just report A-OK all the time or, even worse, they'll periodically start an unjustified anthrax scare by giving off an alarm.
Here's my favorite:
In this case, a primary instrument of that leverage may be something that has gotten relatively little attention in the hype surrounding the new operating system: a Web browser integrated in Windows XP called MSN Explorer.
Relatively little hype, huh? Is the reporter brain damaged? Is he incapable of remembering anything that happened more than a few weeks ago?
One question I have: I'm on TW's high speed internet service. Am I going to be shunted onto AOL?
They had something like this in each and every door (and, of course, Marvin.) Someone should take a boat, and make a mockup of Zaphod's ship.
Also, take heart. This means that the board of directors of Sony/Toyota will be first up against the wall, come the revolution (according to a copy of Encyclopedia Galactica beamed from the future.)
It's a common misconception that the Nevada desert is a wasteland. Guess what: It's not. There is a rather intricate ecosystem. Covering the desert wipes out this system.
This is true. Furthermore, there's the power transmission problem. Until we get room temperature superconductors, only California can benefit from Nevada's ecological destruction. Beaming the power from space is about as feasible as killing people with orbiting laser satellites.
The only real wasteland is in the open ocean. It doesn't look real different from other stretches of ocean but if there's no algae, you can cut off the sun (it only penetrates about 40 meters anyway). You float the things on the water. It's the only ecologically sound place to put them, and there are stretches of this sort of ocean comparatively close to all of the world's coastlines, which is where most people live.
Picture: J Random Musiclover, uses WinMX and KaZaA, until they bog down terribly slowly.
In order for this to work, they're going to need to do denial of service attacks against Gnutella. Sure, our professors have all sold out. Sure, they're just another brick in the wall. However, do you think the former hippies who constitute the nation's CS faculty are going to take it sitting down when "the man" starts doing DoS on *their* (heavily infiltrated by KaZaA, not to mention, good lord, gnutella and freenet) computers? *Especially* when they don't even have to *do* anything, just look the other way while their students do (at my undergrad institution they do this allready, for chrissakes) their very worst?
The RIAA has enough hubris to try it anyway. They could be intelligent and not pick on anyone who might concievably fight back, but I don't think they've gotten that much smarter, and I don't think they understand at all what causes people who aren't empty suits to act or react the way they do.
I've filed a patent on any internetwork protocol ("IP") driven discussion forum centering on issues related to, but not limited to: communications technology, free speech issues, computer science, intellectual property policy, sociological interpretation of trends in technology and technology law, and recreational interests (science fiction, video games, anime et. al.) commonly associated with the "technologically savvy".
CmdrTaco, consider yourself served! I'm still waiting for all that money you owe me for continued use of your DNA.
Also, I've ammended my slashdot user agreement to include an arbitration clause (persuant to the broadest possible interpretation of the Federal Arbitration Act) so I won't actually be seeing you in a "court", per se.
They made up a student association at our U. and called on behalf of it. They faxed a message (on university letterhead) inviting the representative to come and discuss the issue with their organisation; of course, he didn't, but he certainly answered all of their mail.
Of course, they could have accomplished the same thing by joining CalPIRG; but, on one level, ten organisations with one tenth the membership seem bigger than one organisation of everybody.
So, everybody who cares on slashdot (I live in Manhattan) start faxing letters around to everybody in your district, and send in a single copy, with everyone's signature on it. Or, have everybody call on behalf of the organisation. This is what the christian right does, and it works fairly well - they heavily influence (even liberal) representatives in their districts. So, where nerds are concentated (in the NE and West Coast, I've always assumed - take a look at those graphic representations of world computer traffic from a few days ago) we ought to be able to form a political force - if we could 1. agree on anything and 2. form local political organisations on that basis.
However, I don't think it will get all the way to a slashdot-stlye free-for-all where anyone can jump in.
No, I don't think so either. While, certainly, slashdot-like forums will continue to multiply in the sciences (this is one, after all, for computer science,) I don't think they will ever generate much prestige. However, if I had an article (as opposed to a comment) which slashdot carried, I'd put it on my cv (for novelty's sake, if naught else.)
I think that the 'next wave' of this is going to involve a backlash against the diluted power of academic senates vis a vis boards of directors at major universities. We need to find new ways to exert control over the process of research and the nature of research related discourse - control which is increasingly being co-opted by non academic (or nominally academic) administrators and lawyers.
If I send those to a slashdot-style forum, my boss will laugh. They'll count something less than a one-line quote in a small town newspaper.
This is true. Even though I supported the public library of science initiative (reported earlier on slashdot, and also here , I'm about to submit an article to TIBS. An article on GenomeWeb just won't cut it.
However, that is not going to be the case forever. As data mining techniques become more available and sophisticated, that is to say, when real data mining (as opposed to just text matching) becomes a major way in which academics access content, articles in "free" journals are going to be *more* visible to your colleagues, and as important discoveries are coordinated using such techniques, citations will rise, and those journals will rise in prestige. This is going to aggravate what is allready a real schism within the academic community, of int. property versus the pursuit of truth. I think it's going to be a struggle - for the very soul of academia which is really under threat here - but I think we're going to win, because cutting out the middle-man makes for a more efficient way of sharing information, because history isn't over and materialism isn't really the driving force behind human creativity, and because people who love science for it's own sake are better at it.
An earlier poster said that the current system drives the smartest people out of academia and into industry. I couldn't disagree more - I don't think people in (the biotech) industry are very smart at all. They're kind of pathetic, mostly. I'll agree that there are a lot of frankly stupid "scientists" doing terrible work at supposedly public universities on the private dollar, and that the people who jump ship into industry are often a bit smarter than they are; but the smartest, most devoted people are still in the public sector.
Looky. A little-considered effect of all this inter discplinary convergence is that we're going to start running out of acronyms. Some poor slob has been trying to use T&A to mean "theory and application".
Well, it hasn't been good for gene research here in Manhattan. All my colleagues still look like they've been shot in the gut. Seriously. I don't really feel like going to work myself.
What do you think is the limiting ingredient in molec. bio. research? It isn't money - I hate to admit it, but we have plenty. It's competent and capable people who are interested in the field. Eight years down the line we might see more people entering this field particularly because of an interest in bioterrorism (although, if they work for the US government I worry about how much of what they do will really be "defensive".) I think it's more likely that it will draw our best and brightest into the military, intelligence, etc. Sure, we'll refocus our efforts to some extent on molec. bio. that is applicable in anti-bioterrorism, but that comes at an opportunity cost. Do you think that this knowledge is really of more benefit to mankind than plant genetics? AIDS or cancer research? Would it be of more benefit if there were no lunatics with anthrax bombs? I really don't see a silver lining to this.
And there is indeed a smallpox vaccine.
If you want a really good treatment of this question (offline,) I strongly suggest Laurie Garrett's not-so-new-anymore book "Betrayal of Trust". Do your part to help the economy recover and BUY it at amazon.
Anyway, Laurie is not so confident in the efficacy of our smallpox vaccine stock as the CDC. Believe whom you will.
Actually, present techniques are more sensitive than a DNA screen. Read about them at the cdc.
Basically, its easier to detect the bacteria themselves than to do a screen on their DNA.
I really don't want to troll, but am I mistaken or is that addressed to me? I mean, I posted a link to a comic strip. Do I need to bracket it in rows of smiley faces? Meet me halfway, here.
Once someone says "it has nothing to do with encryption" aren't all future messages to the same effect redundant? Just because other people keep going on about encryption doesn't mean you have to keep upmoderating anybody who repeats that.
On a more serious note, if they really are just like us geeks on slashdot, why do they work at the NSA? Are geeks who work at the NSA happy with their jobs? I wouldn't want to solicit anything classified (since I post with my e-mail address :)) but do we know who at the NSA arranged this? It's a nice piece of work and I'm wondering how friendly an environment the NSA really is for people who want to do open source work. Is this the brainchild of one particular linux-friendly higher-up or are we going to see a lot of open source stuff coming out of the NSA?
Yeah, and this has nothing to do with encryption. They use steganography. I found the phrase "Help us! We've been captured by the NSA!" embedded in one of the header files. Story here .
At UCSC, where I was an undergrad, we had fairly numerous breadth requirements, which was excellent. The original poster seems to think that four year colleges should be more like trade schools - they should not. Four year college programs do not exist to churn out techs to work in industry - although there is certainly an effort to subvert them to do that. A graduate of such a program is _supposed_ to be an active participant in the intellectual life of society, whatever they may do for a living. So, yeah, if you have a BA in computer science you should know something about sociology, philosophy and art, not the least of how the CS industry is actually run at every level. Idea is - you need this information to make moral judgements about how to use your expertise, which is one of the things that a BA is supposed to do. If you really don't care, that's fine, I'm not criticizing you, but our educational institutions shouldn't be reduced to a rubber stamp so that people with no social conscience or inherent curiosity can get better paying jobs! Go to law school, for chrissakes.
The idea with teaching "big picture" thinking is that you will better understand the details of what you're doing if you know why you're doing them. The Poster's courses sound painfully stupid (decision theoretrics with no stat? taught by a little balding guy with pointy hair?) but that isn't a problem with breadth.
That doesn't mean that the pursuit or release of knowledge should be restricted in any way.
As a scientist I am more concerned with what other scientists are *doing* that with what they are *developing*. Our colleagues who developed the techniques to clone DNA into plant cells (a number of whom I know personally) did nothing wrong, and should not have delayed publication because of the "ethical" consideration of what someone else could do with it. The people who are genetically altering corn to make it increasingly resistant to chlorinated organics (roundup) are *doing* something unethical; and they are the ones, largely highly intelligent people, whom we need to reach and educate. Some of the things I'm attempting to do could have direct, terrible applications in germ warfare - but they could also be a great boon to medical research. The resolution of that dillemma is clear: we cannot call a halt to scientific progress because of fear.
Other scientists, and some people may draw an increasingly meaningless distinction and call them engineers, are actually applying these developments to do things that shouldn't be done. Biopreparat doesn't exist anymore, but I'm sure biological weapons research continues. The people who nerve gassed the Tokyo subway where highly educated. These people are doing more damage with their own scientific expertise than laymen ever can, or will, with something you release.
Ethics requirements at graduate schools should be specific, factual and tailored to the particular focus of the student. Individuals who want to go into plant genetics should take courses in the political economics of third world agriculture - the same ones that pol sci students take. Courses in "ethics" are substanceless exercises in sophistry (say that 10 times fast) that don't teach the consequences of the particular actions a student might actually take.
While relatively uneducated terrorists can make certain uses of publically released technologies like culturing eukaryotic cells or near unbreakable encryption, the *real* danger, and it is a real danger, is when the scientists ourselves are actually setting out to do harm; or applying these technologies in ignorance for our own economic gain.
> Besides, in any case you dumb shit, all the > terrorists have to do is keep their noses clean! Well, the fact is, they didn't keep their noses clean and seldom do. Since "classified" evidence can be applied to foreign nationals in immigration situations, without disclosing the evidence to the immigrant, if we could figure out who people were we could deny them entrance to the country on thin evidence, hearsay, etc. Which is not exactly an enlightened practice but it would have kept all of the people involved in the WTC out of the country, along with a lot of their totally innocent friends and acquaintances. > We already have many forms of identification. ID's which can easily be counterfited. This is true but isn't really a rational argument against having a single ID. A single ID, _if_ managed properly could be much harder to hack, or counterfeit. That would reduce the number of fake IDs in circulation, which would reduce crime on a number of levels. In fact, if we were willing to invest the resources needed to do so, we could ensure that hacking the database was effectively impossible; the only way to get a fake ID would be to get someone on the "inside" to help, and if the system were set up properly we could catch those people if and when any falsified ID they had personally certified was used. It could be made to work, and, while I agree that their president is a grade A pompous asshole, Oracle really does have the technology to accomplish all of this. Which is not to say that I think we should (I agree with previous posters about the need for a spate of privacy controls) but we definitely could and it could absolutely work.
The threat of a lawsuit may be more effective than an actual law. Sure, the guy who cracked the SDMI watermarks almost certainly would have won the resultant lawsuit (unless he got the idiot judge from the DeCSS case) but he couldn't afford to be sued... actually he was getting all Sun Tzu on the RIAA/SDMI, but still.
Since the NSA allready thinks that any cryptographer who doesn't work for them is some kind of rogue lunatic, I don't think they'd shy away from bringing charges against people they didn't like who published a development that they would rather keep secret. Sure, eventually they'd lose, but until they actually did they could keep people in prison for months or years awaiting appeals! The threat of being prosecuted for publishing such a description (which, as I said, is tantamount to a device under a hostile interpretation of the DeCSS ruling) is pretty serious. Also, while I'm sure we'll win DeCSS sooner or later, it's more than likely that the final ruling will be vague or waffling enough not to close out future prosecutions.
My understanding had been (I find an indirect reference to this so maybe the reporter is wrong) that the ruling included the Object code (in whatsoever language it was written, my friend wrote a compiler that turned the first page of King Lear into DeCSS - much more clever than a pretty picture), and the Object code _is_ just a description. Fine, so we suppose there's a difference between a description and a description-that-some-automated-routine-somewhere- can-understand. Sure, whatever.
While you _can_ just publish a truly abstract description of the algorithm, especially in a pure math journal, applied math journals generally want to see you do it. Although, I'm sure the reviewers would be sympathetic if that were _illegal_. Regardless, we know that researchers who come up with something like that are going to want to code it to test their theories - is that going to be illegal? Are they going to have to add the backdoor (even if that somehow jams up there whole gig) to make the program legal? Doesn't that mean that they'll have to be *given* the back door, whatever on earth it actually is, so that they can make sure it decodes their message? Will they have to turn the source code over to the government before publication so that the government can add the backdoor? I'm running on too long here but am I missing some obvious solution to this seeming logical disjoint, or does this outlaw any innovation in encryption at all?