Return On Investment. It is cheaper to solve the problems of dangerous and inept drivers. This will eliminate tens, if not hundreds, of times as many deaths in a single year as the 9/11 attacks, which were a solitary, unrepeated incident in which the terrorist organization involved showed - through plots of unbelievable naivety - that the hijacking and destruction of the twin towers was sheer unbridalled luck on their part.
Pouring in money into a counter-terrorism outfit that hasn't yet succeeded in foiling a single plot (the plots that have been foiled have all been foiled by traditional policework) in the hopes that there'll be some dramatic payoff is utter lunacy.
Nationalizing healthcare would be another good way to save lives. The next four leading industrial nations pay something like 1/200th the amount Americans do, have lower obesity, better preventative care and lower preventable deaths. Money saved =AND= lives saved.
In these tough economic times, burning money on a white elephant that does nothing except deceive those who can't be bothered doing the basic arithmetic isn't helping the nation. Frankly, America would have done far better to ignore the whole Twin Towers episode entirely. All America has succeeded in doing is to have destabilized some nations, achieved a far stronger alliance between Iran and North Korea than had ever been thought possible (given one's rabidly anti-religious and the other is rabidly anti-atheist) and blown up Gadhaffi's CD collection. My, what an achievement. And it's put the US in the hole by about a trillion and a half.
Do you know how many lives could have been saved, in the past decade, if the government had spent even half of that in making sure public schools served healthy meals, the poor had affordable healthcare, the mentally ill were off the streets and in psychiatric care?
Hell, since most of the home-grown terrorists have been either deprived or insane, what do you think it would have done to the homegrown terror movement? Erased it, in all probability, totally out of existance.
THAT is how to spend your money wisely. THAT is what we need, not some paranoid manifesto by the psychos for the psychos.
You mean, aside from the attempted bombing of Portland, Oregon? There's been a few others as well. I'd consider the Koran burning by that fringe church a terrorist event, given that it was designed purely to control policy through fear and intimidation.
Deaths, that's different from attacks. Has it prevented any deaths? Since we can't know what would have happened without the law, it's hard to be certain. And how do you define "deaths" in this context anyway? The number killed as a direct result of the incident? Do you include indirect as well? And how indirect is permitted? Do you adjust for the number who died in the incident but would have died within some reasonable margin (a year and a day is normal) from some other cause anyway?
Also, what do you define as terrorism? I'd consider the bombing of the Olympic games, the flying of an aircraft into the IRS building, the bombing of Federal offices, the Unabomber campaign, all the racial lynchings and murders that have gone on over the decades, etc, to all be acts of terrorism. Most of these happened prior to the Patriot Act, but:
(a) Can you honestly say the Patriot Act would have stopped a single one of them? (b) You are aware that the total deaths due to these acts of terror outnumber those that happened in 9/11?
I'd also consider the NRA and NORAID funding of the IRA - these days, the "Real IRA" as the rest stopped the killing - to be acts of terrorism. Can you name me those who have been tried for terrorist-related activities under the Patriot Act?
They are, however, bloody annoying and make spotting anything important that much harder.
If you know legit connections will be in certain IP blocks, ban all others.
If not but you've decent FTP software, have it block access from any IP that tries to access specific accounts (such as Administrator, samba, nagios or whatever ones that both don't exist and are being targeted on your box).
If you prefer something more entertaining, use Linux' packet munging code. What you want to do is detect inbound connections that aren't legit then rewrite the destination IP to somewhere else. Make it someone who is unlikely to be at any real risk but is likely to kick up a fuss.
That's not the bit that scares me the most. The bit that scares me the most is that anyone with an ounce of skill in reverse engineering can identify the security flaws used, and anyone with an ounce of skill in assembly can disassemble Stuxnet, alter what it targets, and launch the new variant.
By banning the talk, the DHS is preventing US industries from protecting themselves against economic warfare. Plenty of nations (China and Russia especially) are investing in cyber-warfare. There's plenty of amateurs out there with axes (albeit often as delusionary as the DHS') to grind. It is simply not excusable for the US to be placed in this kind of danger.
For what purpose? Siemans can't get a worse rep than to be accused of having worked with virus writers. The consumers can't exactly switch from SCADA to Infiniband or other rival networking technologies. The exploit is public knowledge.
...doesn't the existance of a virus that can attack such devices make this a zero-day flaw? The hack is public, since anyone can disassemble the virus that's in the wild and see how it works.
And, frankly, I don't see it being awfully difficult for any Black Hat with a mind to to rip out the prior payload and install one that can attack a wider range of devices. Surely it is in the interests of security for corporations to understand what they can do to mitigate the risk of this.
The DHS, IMHO, is acting in a manner that directly threatens US interests and US corporations by preventing those at risk from knowing as much as those who pose a risk. This argument has been had out before, with regards to CERT and when it should post alerts. It was accepted that there would be a reasonable pause to allow a fix. The virus was first discovered in July 15 2010. So the vulnerabilities have been zero-day for 10 months now.
Decoupling pre-dates object-oriented programming by some ways, but that increases the importance of the point. It is a requirement of sound software engineering that you have appropriate abstraction.
I'd be more surprised if this didn't result in an upsurge in multicasting. Streaming 11 people ptp would break the limit of 10 "performances". Multicasting once to 1,000 people in one go would be a single "public performance". Guess which becomes the more attractive. (It'd also slash bandwidth usage, freeing up the Internet for things it's meant for, like prawns.)
It is specified in a blueprint (the nucleic DNA and the epigenome), designed by evolution and built by Von Neumann machines which constitute your actors. Von Neumann demonstrated, via cellular automata, that the actor and blueprint could be contained within a single entity. I pity those who are 60-70 years behind on mathematics, but it really isn't my problem.
The brain isn't designed to think independent of context. It's built to be part of a social system. Yes, people need to think for themselves, but to mandate that requires us to break the architecture of our minds. It can't work as the primary solution. Indeed, this current study only shows that what used to be two points ("mob thinking" and "collective intelligence") are just two points on an entire continuum. The problem is that humanity prefers to slide to the lower end of the spectrum rather than rise to its potential. THAT is what you need to solve. The details of who thinks and how then become incidental. Mere implementation details.
Once they've filtered the cockpit voice recordings. There'll be all kinds of noise, especially during a major storm. Towards the end, there'll be a multitude of sirens, klaxons, buzzers and alarm clocks going off. But you can't just filter any old noise, you have to filter out the noise that adds nothing but keep in all the noise that is important. That's harder than just applying a basic filter.
Try transcribing the dialog off a movie without rewinding it. You'll find it's hard. Takes longer than the movie actually runs for. Now try doing the same with a jet engine stuffed in one ear and a hundred games consoles wailing at the other. It's probably going to take a LOT longer than it did when things were quiet.
The text depends on whether the recording is time-continuous or is triggered by sound. It's probably the former, since you want to record even slight fluctuations in engine sounds, or other subtle disturbances, but I don't want to assume. Recording change is cheaper than recording a constant. If there's any nonlinearity to the recording, working out when things were said would be harder.
And, finally, the sounds of the controls being operated has to be time-synchronized with the data recorder's recording of those events taking place, so allowing the instrument readings (as far as are recorded on the data recorder) to be correlated with the speech.
You are bombarded because you haven't developed these bubbles yet.
I dislike the terms "information bubble" or "filter bubble". These are much more akin to virtual private networks or virtual circuits, in that the information is physically in the same space but logically seperated. Only those belonging to that specific virtual network can observe what is on it, with most people belonging to just one virtual network.
Most Slashdotters will have, at the very least, set up a switch. A switch prevents your LAN from being flooded with packets by constricting the information that enters the LAN. Nobody thinks of it as a bubble, though.
In the real-world, there's way too much information out there. You'd be drowned in the stuff if you were exposed to it all. So, you set up mental versions of switches, firewalls and VPNs to keep the traffic down to tolerable levels.
This is a Good Thing, if configured correctly. The problem is that extremists, fundamentalists and dogmatists are very good at teaching people to misconfigure their mental networks, whereas the places that are supposed to teach you how to configure brain WANs correctly (schools and civic organizations) either teach you nothing at all or are so badly infiltrated by extremists, fundamentalists and/or dogmatists that they reinforce the very worst practices. This is a Bad Thing.
Precisely. The only solution I can think of is for education and culture to hammer away at the walls. If the barriers never get a chance to solidify, but always remain at least a little fluid, then other mechanisms for reconciling beyond petty rejection must be developed. You don't need to come up with the perfect solution, you need only force the brains of others to do so.
I absolutely agree with you regarding the incremental approach vs revolution thing. A portable lab, combining the best techniques for developing cultures, the best compact microscope, and even the best in rapid DNA analysis (although the rapid systems are not that accurate yet) since all you need is a centrifuge, a few ezymes (one to multiply the DNA, one to chop the strands into bits) and some relatively compact kit for reading the DNA fragments. There may be a few other bits and pieces you'd want, too, but you could do it and be as accurate as any fixed lab in any nation.
I prefer the revolutionary approach over the evolutionary one (what I consider "invention" versus "innovation") - particularly for an X-Prize type contest - for a few reasons. First, evolution of ideas is going to happen anyway. You don't need a big incentive or a contest to ensure that what's inevitable is inevitable a little bit quicker. Second, there's not much spin-off on evolving ideas, since each step is a relatively small increment compared to the whole. What I call an "invention" is, if not 100% of the whole (since there will be some leverage of existing work) nonetheless many orders of magnitude larger in increment size, with all that that implies for the potential for spin-offs and secondary development.
The third reason is admittedly somewhat egotistical. Ok, very egotistical. I regard the level of intelligence and geekery required for truly novel thought to greatly exceed that required for straight innovation. In that regard, I very much recognized, agreed-with and totally approved of the fictionalized version of John Nash in "A Beautiful Mind" where he wants his project to be truly original and scoffs at the pedestrian work of his rivals. And, yes, I know that he was also barking mad at the time (both in reality and in the movie) and also had a very long lean time when those pedestrian rivals were highly successful. However, all bright people are at least a little mad and you do have to search a LOT of blue sky to find anything worth the finding.
So, in the short-term, certainly, evolution is the way to go. Also, evolution is also absolutely necessary since no idea is perfect when first developed and no idea that was perfect can remain so when the environment it is in is changing.
I will, however, always hold that the original ideas are the long-term best hope of avoiding evolutionary dead-ends. Which happen and happen rather a lot in both nature and technology. (It's impossible for evolution to not have the majority of paths be dead-ends sooner-or-later, since evolution is an exponential process. The danger is the same one you run into when trying to step-wise find the maximum of a function - local maxima. Out-competes the alternatives then dies. It's why herustics always allow tracing backwards, so that you can undo such mistakes. Technology's a bit more awkward. Replacing silicon, for example, has been so hard because there's too much dependence on silicon specifically for you to simply trace back and follow a different path.)
The drawback to original ideas is that it costs orders of magnitude more to try an idea out, ideas are much less likely to succeed at all (since there's less history to work with in understanding what is wanted or how something might be made to work), there's no economy of scale because there's no scale and indeed no economy for the idea, and theory will outstrip what can be done in practice by a good few centuries. In the case of robotics, over two millenia. By the time the supporting technology exists, the theories the original idea relied on may no longer be considered true or a completely different approach may have been decided upon.
Ok, that's more than one drawback.
The key, though, is to bounce the original ideas around rather than keep them closetted away.
A totally general problem may well be impossible. I'll accept that. The solution I describe would be capable of going through a catalogue of known molecules (which presumbaly would only contain ones you care about) and telling you which ones were present. This would be sufficient to identify any well-known variant of any well-known pathogen with greater accuracy than a traditional doctor, provided the power budget issue could be solved. Since the requirement was to out-perform a traditional doctor, not out-perform Bones, this is sufficient.
Oh, I wouldn't go so far as to say it's "feasible", merely that you're limited in remote sensing. To determine if a technique works, you start by assuming it will and then work along that chain until you either meet a contradiction (proving that the technique won't work) or have a successful implementation. So, to test whether a tricorder is, indeed, a possibility you start by assuming that a tricorder could exist in principle, deduce what must automatically follow (remote sensing relies on radiation of some sort, for example) and work from there.
I have no objection to someone saying "well, if we continue along this chain, we reach this problem" as you have done. The problem can then be discussed and solved or discussed and shown to be unsolvable. That is an entirely rational, sound approach. My problem is with people who start from either the a-priori assumption that something which contradicts their preconceived notions HAS to be wrong, or who start from the a-priori notion that pleasent discussion is for wimps. Such people do not deserve the dignity of being called human.
Yes, there are billions of molecules involved. To falsify the technique, you need only show that no matter how many absorption lines you know, you can never reach a diagnosis that is as accurate as a traditional doctor using traditional techniques. So, given the number of molecules, you need only show that there is no statistical liklihood that, no matter how good your device and how extensive your knowledge, you can't rule in or out enough possibilities for the device to be useful.
That would be an entirely fair argument.
It's not a simple argument, though. Any given type of cell will have a number of unique molecule chains. What's more, of the byproducts generated by the cell, some may be unique to that type of cell (or, at least, very unusual). Because you're not looking for a molecule in isolation, but rather for enough of a signature from each and every element of the set of molecules that would go together for any given pathogen, I'm not convinced that the complexity of the problem is enough to falsify the method.
Is it enough to say the technique can't be used today, as per your final argument? Ah, now THAT you can most certainly argue. Remember, remote sensing is going to have to require extreme accuracy with very low noise levels across the entire radio spectrum (forget IR - way too short a wavelength, even microwave is dubious as by that time you're dealing with chemical bonds and not the molecule as a whole, making it useless for this kind of work). Because it's totally dependent on background noise being missing, the background noise has to be there in the first place. It's also highly dependent on you collecting enough of a signal to be able to say that the background noise is indeed missing.
In short, a decent-sized parabolic dish operating with a very long baseline, with analogue components kept supercooled, otherwise there's no possibility with what's currently out there of detecting a damn thing.
Now, onto the database. Let's say you've a nice, round billion molecules to contend with, each of which has (for the sake of argument) ten lines you can look for. These numbers aren't intended to be realistic, they're merely intended for illustrative purposes. I'm not concerned with the signatures of the chemical bonds (except when the resolution you can work at makes it otherwise impossible to distinguish between two molecules) and I'm certainly not concerned with atoms or electrons, so a fair chunk of overlap should (note: should, not will) fall away. Actually, ten lines is probably too many, since you've only three axes you can rotate the molecule around but that gives you room to expand into chemical bonds if you need.
In the worst case, that gives you ten billion lines to look for. One of the largest SETI experiments involved a billion DSPs, so it's not impossible to get that kind of compute power in one place. That's only ten baselines worth of data collection, which is acceptable tim
Your first mistake is to combine "government" and "logic" in the same thought. Your second mistake is to confuse spending rationally versus the normal procurement method which is to spend the least.
Ah, well, to weight correctly, you would create a pool for every possible combination and select randomly the correct number from each pool. At first I took it that they tried to do exactly that but it wouldn't take a month to generate a valid sample with such a method. Which means that they're randomly selecting from everyone and then seeing if the sample meets the criteria. I can think of no other way it can take so long.
Which means the problem exists between keyboard and IDE.
The only jackasses here are you and the guy you replied to.
One thing is absolutely certain - no matter how wrong I am, I'm more right than you could even dream of being by the mere fact that I looked for a solution. What's more, a solution to the correct problem. Something you and the parent post were incapable of identifying.
If you can't identify the problem, DON'T harass another's identification of the solution space.
Hmmm. IR for a long, complex chain molecule, given that even reasonably short and simple chain molecules are all in the radio spectrum. Methinks you are an idiot.
NMR... And how the hell do you propose to make a bloody 3T magnet portable??? And what the friggin' hell is the power source going to be??? Are you a complete imbecile or just practicing?
Given that the challenge is clearly intended to produce a device that can be taken to patients who can NOT be taken to a hospital with static medical devices, and given that said people are very likely in areas where there is FUCK ALL for transportation, power or repair shops, and where diseases are likely to be things like Marburg, Ebola or XDR-TB, where the devices will likely be in the hands of someone who last week had been in tryouts for witch-doctor of the month (ie: not someone you would trust with a syringe or scalpel), what the hell kind of device do you think the challenge is going to want??!!
The last thing a prize like this needs is some prize idiot like you to come up with some insane idea that is wholly unworkable in such environs. The popular phrase is "beating tractors into ploughshares".
Your utter drivel of a post is worth replying to only in that it is so utterly sickening that you could have had enough brain cells to type and yet insufficient brain cells to comprehend what is required.
Slashdot Mirror
Return On Investment. It is cheaper to solve the problems of dangerous and inept drivers. This will eliminate tens, if not hundreds, of times as many deaths in a single year as the 9/11 attacks, which were a solitary, unrepeated incident in which the terrorist organization involved showed - through plots of unbelievable naivety - that the hijacking and destruction of the twin towers was sheer unbridalled luck on their part.
Pouring in money into a counter-terrorism outfit that hasn't yet succeeded in foiling a single plot (the plots that have been foiled have all been foiled by traditional policework) in the hopes that there'll be some dramatic payoff is utter lunacy.
Nationalizing healthcare would be another good way to save lives. The next four leading industrial nations pay something like 1/200th the amount Americans do, have lower obesity, better preventative care and lower preventable deaths. Money saved =AND= lives saved.
In these tough economic times, burning money on a white elephant that does nothing except deceive those who can't be bothered doing the basic arithmetic isn't helping the nation. Frankly, America would have done far better to ignore the whole Twin Towers episode entirely. All America has succeeded in doing is to have destabilized some nations, achieved a far stronger alliance between Iran and North Korea than had ever been thought possible (given one's rabidly anti-religious and the other is rabidly anti-atheist) and blown up Gadhaffi's CD collection. My, what an achievement. And it's put the US in the hole by about a trillion and a half.
Do you know how many lives could have been saved, in the past decade, if the government had spent even half of that in making sure public schools served healthy meals, the poor had affordable healthcare, the mentally ill were off the streets and in psychiatric care?
Hell, since most of the home-grown terrorists have been either deprived or insane, what do you think it would have done to the homegrown terror movement? Erased it, in all probability, totally out of existance.
THAT is how to spend your money wisely. THAT is what we need, not some paranoid manifesto by the psychos for the psychos.
You mean, aside from the attempted bombing of Portland, Oregon? There's been a few others as well. I'd consider the Koran burning by that fringe church a terrorist event, given that it was designed purely to control policy through fear and intimidation.
Deaths, that's different from attacks. Has it prevented any deaths? Since we can't know what would have happened without the law, it's hard to be certain. And how do you define "deaths" in this context anyway? The number killed as a direct result of the incident? Do you include indirect as well? And how indirect is permitted? Do you adjust for the number who died in the incident but would have died within some reasonable margin (a year and a day is normal) from some other cause anyway?
Also, what do you define as terrorism? I'd consider the bombing of the Olympic games, the flying of an aircraft into the IRS building, the bombing of Federal offices, the Unabomber campaign, all the racial lynchings and murders that have gone on over the decades, etc, to all be acts of terrorism. Most of these happened prior to the Patriot Act, but:
(a) Can you honestly say the Patriot Act would have stopped a single one of them?
(b) You are aware that the total deaths due to these acts of terror outnumber those that happened in 9/11?
I'd also consider the NRA and NORAID funding of the IRA - these days, the "Real IRA" as the rest stopped the killing - to be acts of terrorism. Can you name me those who have been tried for terrorist-related activities under the Patriot Act?
Require client certs?
They are, however, bloody annoying and make spotting anything important that much harder.
If you know legit connections will be in certain IP blocks, ban all others.
If not but you've decent FTP software, have it block access from any IP that tries to access specific accounts (such as Administrator, samba, nagios or whatever ones that both don't exist and are being targeted on your box).
If you prefer something more entertaining, use Linux' packet munging code. What you want to do is detect inbound connections that aren't legit then rewrite the destination IP to somewhere else. Make it someone who is unlikely to be at any real risk but is likely to kick up a fuss.
Multicasting is native on the Internet backbone. Here is a list of ISPs that support it to the home.
That's not the bit that scares me the most. The bit that scares me the most is that anyone with an ounce of skill in reverse engineering can identify the security flaws used, and anyone with an ounce of skill in assembly can disassemble Stuxnet, alter what it targets, and launch the new variant.
By banning the talk, the DHS is preventing US industries from protecting themselves against economic warfare. Plenty of nations (China and Russia especially) are investing in cyber-warfare. There's plenty of amateurs out there with axes (albeit often as delusionary as the DHS') to grind. It is simply not excusable for the US to be placed in this kind of danger.
For what purpose? Siemans can't get a worse rep than to be accused of having worked with virus writers. The consumers can't exactly switch from SCADA to Infiniband or other rival networking technologies. The exploit is public knowledge.
Who, then, is going to be protected?
...doesn't the existance of a virus that can attack such devices make this a zero-day flaw? The hack is public, since anyone can disassemble the virus that's in the wild and see how it works.
And, frankly, I don't see it being awfully difficult for any Black Hat with a mind to to rip out the prior payload and install one that can attack a wider range of devices. Surely it is in the interests of security for corporations to understand what they can do to mitigate the risk of this.
The DHS, IMHO, is acting in a manner that directly threatens US interests and US corporations by preventing those at risk from knowing as much as those who pose a risk. This argument has been had out before, with regards to CERT and when it should post alerts. It was accepted that there would be a reasonable pause to allow a fix. The virus was first discovered in July 15 2010. So the vulnerabilities have been zero-day for 10 months now.
Decoupling pre-dates object-oriented programming by some ways, but that increases the importance of the point. It is a requirement of sound software engineering that you have appropriate abstraction.
I'd be more surprised if this didn't result in an upsurge in multicasting. Streaming 11 people ptp would break the limit of 10 "performances". Multicasting once to 1,000 people in one go would be a single "public performance". Guess which becomes the more attractive. (It'd also slash bandwidth usage, freeing up the Internet for things it's meant for, like prawns.)
It is specified in a blueprint (the nucleic DNA and the epigenome), designed by evolution and built by Von Neumann machines which constitute your actors. Von Neumann demonstrated, via cellular automata, that the actor and blueprint could be contained within a single entity. I pity those who are 60-70 years behind on mathematics, but it really isn't my problem.
The brain isn't designed to think independent of context. It's built to be part of a social system. Yes, people need to think for themselves, but to mandate that requires us to break the architecture of our minds. It can't work as the primary solution. Indeed, this current study only shows that what used to be two points ("mob thinking" and "collective intelligence") are just two points on an entire continuum. The problem is that humanity prefers to slide to the lower end of the spectrum rather than rise to its potential. THAT is what you need to solve. The details of who thinks and how then become incidental. Mere implementation details.
Saints don't let saints post on Facebook
Once they've filtered the cockpit voice recordings. There'll be all kinds of noise, especially during a major storm. Towards the end, there'll be a multitude of sirens, klaxons, buzzers and alarm clocks going off. But you can't just filter any old noise, you have to filter out the noise that adds nothing but keep in all the noise that is important. That's harder than just applying a basic filter.
Try transcribing the dialog off a movie without rewinding it. You'll find it's hard. Takes longer than the movie actually runs for. Now try doing the same with a jet engine stuffed in one ear and a hundred games consoles wailing at the other. It's probably going to take a LOT longer than it did when things were quiet.
The text depends on whether the recording is time-continuous or is triggered by sound. It's probably the former, since you want to record even slight fluctuations in engine sounds, or other subtle disturbances, but I don't want to assume. Recording change is cheaper than recording a constant. If there's any nonlinearity to the recording, working out when things were said would be harder.
And, finally, the sounds of the controls being operated has to be time-synchronized with the data recorder's recording of those events taking place, so allowing the instrument readings (as far as are recorded on the data recorder) to be correlated with the speech.
There I agree absolutely. The switch should never be capable of overriding your preferences, no matter what others deem.
You are bombarded because you haven't developed these bubbles yet.
I dislike the terms "information bubble" or "filter bubble". These are much more akin to virtual private networks or virtual circuits, in that the information is physically in the same space but logically seperated. Only those belonging to that specific virtual network can observe what is on it, with most people belonging to just one virtual network.
Most Slashdotters will have, at the very least, set up a switch. A switch prevents your LAN from being flooded with packets by constricting the information that enters the LAN. Nobody thinks of it as a bubble, though.
In the real-world, there's way too much information out there. You'd be drowned in the stuff if you were exposed to it all. So, you set up mental versions of switches, firewalls and VPNs to keep the traffic down to tolerable levels.
This is a Good Thing, if configured correctly. The problem is that extremists, fundamentalists and dogmatists are very good at teaching people to misconfigure their mental networks, whereas the places that are supposed to teach you how to configure brain WANs correctly (schools and civic organizations) either teach you nothing at all or are so badly infiltrated by extremists, fundamentalists and/or dogmatists that they reinforce the very worst practices. This is a Bad Thing.
Bubbles tend to be spherical, but you require something more fractal in nature.
Precisely. The only solution I can think of is for education and culture to hammer away at the walls. If the barriers never get a chance to solidify, but always remain at least a little fluid, then other mechanisms for reconciling beyond petty rejection must be developed. You don't need to come up with the perfect solution, you need only force the brains of others to do so.
I absolutely agree with you regarding the incremental approach vs revolution thing. A portable lab, combining the best techniques for developing cultures, the best compact microscope, and even the best in rapid DNA analysis (although the rapid systems are not that accurate yet) since all you need is a centrifuge, a few ezymes (one to multiply the DNA, one to chop the strands into bits) and some relatively compact kit for reading the DNA fragments. There may be a few other bits and pieces you'd want, too, but you could do it and be as accurate as any fixed lab in any nation.
I prefer the revolutionary approach over the evolutionary one (what I consider "invention" versus "innovation") - particularly for an X-Prize type contest - for a few reasons. First, evolution of ideas is going to happen anyway. You don't need a big incentive or a contest to ensure that what's inevitable is inevitable a little bit quicker. Second, there's not much spin-off on evolving ideas, since each step is a relatively small increment compared to the whole. What I call an "invention" is, if not 100% of the whole (since there will be some leverage of existing work) nonetheless many orders of magnitude larger in increment size, with all that that implies for the potential for spin-offs and secondary development.
The third reason is admittedly somewhat egotistical. Ok, very egotistical. I regard the level of intelligence and geekery required for truly novel thought to greatly exceed that required for straight innovation. In that regard, I very much recognized, agreed-with and totally approved of the fictionalized version of John Nash in "A Beautiful Mind" where he wants his project to be truly original and scoffs at the pedestrian work of his rivals. And, yes, I know that he was also barking mad at the time (both in reality and in the movie) and also had a very long lean time when those pedestrian rivals were highly successful. However, all bright people are at least a little mad and you do have to search a LOT of blue sky to find anything worth the finding.
So, in the short-term, certainly, evolution is the way to go. Also, evolution is also absolutely necessary since no idea is perfect when first developed and no idea that was perfect can remain so when the environment it is in is changing.
I will, however, always hold that the original ideas are the long-term best hope of avoiding evolutionary dead-ends. Which happen and happen rather a lot in both nature and technology. (It's impossible for evolution to not have the majority of paths be dead-ends sooner-or-later, since evolution is an exponential process. The danger is the same one you run into when trying to step-wise find the maximum of a function - local maxima. Out-competes the alternatives then dies. It's why herustics always allow tracing backwards, so that you can undo such mistakes. Technology's a bit more awkward. Replacing silicon, for example, has been so hard because there's too much dependence on silicon specifically for you to simply trace back and follow a different path.)
The drawback to original ideas is that it costs orders of magnitude more to try an idea out, ideas are much less likely to succeed at all (since there's less history to work with in understanding what is wanted or how something might be made to work), there's no economy of scale because there's no scale and indeed no economy for the idea, and theory will outstrip what can be done in practice by a good few centuries. In the case of robotics, over two millenia. By the time the supporting technology exists, the theories the original idea relied on may no longer be considered true or a completely different approach may have been decided upon.
Ok, that's more than one drawback.
The key, though, is to bounce the original ideas around rather than keep them closetted away.
That's so incredibly sad. It should be a requirement that people writing such algorithms read and understand such books as "The Unfinished Game".
A totally general problem may well be impossible. I'll accept that. The solution I describe would be capable of going through a catalogue of known molecules (which presumbaly would only contain ones you care about) and telling you which ones were present. This would be sufficient to identify any well-known variant of any well-known pathogen with greater accuracy than a traditional doctor, provided the power budget issue could be solved. Since the requirement was to out-perform a traditional doctor, not out-perform Bones, this is sufficient.
Oh, I wouldn't go so far as to say it's "feasible", merely that you're limited in remote sensing. To determine if a technique works, you start by assuming it will and then work along that chain until you either meet a contradiction (proving that the technique won't work) or have a successful implementation. So, to test whether a tricorder is, indeed, a possibility you start by assuming that a tricorder could exist in principle, deduce what must automatically follow (remote sensing relies on radiation of some sort, for example) and work from there.
I have no objection to someone saying "well, if we continue along this chain, we reach this problem" as you have done. The problem can then be discussed and solved or discussed and shown to be unsolvable. That is an entirely rational, sound approach. My problem is with people who start from either the a-priori assumption that something which contradicts their preconceived notions HAS to be wrong, or who start from the a-priori notion that pleasent discussion is for wimps. Such people do not deserve the dignity of being called human.
Yes, there are billions of molecules involved. To falsify the technique, you need only show that no matter how many absorption lines you know, you can never reach a diagnosis that is as accurate as a traditional doctor using traditional techniques. So, given the number of molecules, you need only show that there is no statistical liklihood that, no matter how good your device and how extensive your knowledge, you can't rule in or out enough possibilities for the device to be useful.
That would be an entirely fair argument.
It's not a simple argument, though. Any given type of cell will have a number of unique molecule chains. What's more, of the byproducts generated by the cell, some may be unique to that type of cell (or, at least, very unusual). Because you're not looking for a molecule in isolation, but rather for enough of a signature from each and every element of the set of molecules that would go together for any given pathogen, I'm not convinced that the complexity of the problem is enough to falsify the method.
Is it enough to say the technique can't be used today, as per your final argument? Ah, now THAT you can most certainly argue. Remember, remote sensing is going to have to require extreme accuracy with very low noise levels across the entire radio spectrum (forget IR - way too short a wavelength, even microwave is dubious as by that time you're dealing with chemical bonds and not the molecule as a whole, making it useless for this kind of work). Because it's totally dependent on background noise being missing, the background noise has to be there in the first place. It's also highly dependent on you collecting enough of a signal to be able to say that the background noise is indeed missing.
In short, a decent-sized parabolic dish operating with a very long baseline, with analogue components kept supercooled, otherwise there's no possibility with what's currently out there of detecting a damn thing.
Now, onto the database. Let's say you've a nice, round billion molecules to contend with, each of which has (for the sake of argument) ten lines you can look for. These numbers aren't intended to be realistic, they're merely intended for illustrative purposes. I'm not concerned with the signatures of the chemical bonds (except when the resolution you can work at makes it otherwise impossible to distinguish between two molecules) and I'm certainly not concerned with atoms or electrons, so a fair chunk of overlap should (note: should, not will) fall away. Actually, ten lines is probably too many, since you've only three axes you can rotate the molecule around but that gives you room to expand into chemical bonds if you need.
In the worst case, that gives you ten billion lines to look for. One of the largest SETI experiments involved a billion DSPs, so it's not impossible to get that kind of compute power in one place. That's only ten baselines worth of data collection, which is acceptable tim
Your first mistake is to combine "government" and "logic" in the same thought. Your second mistake is to confuse spending rationally versus the normal procurement method which is to spend the least.
Other than that, you're entirely correct.
Ah, well, to weight correctly, you would create a pool for every possible combination and select randomly the correct number from each pool. At first I took it that they tried to do exactly that but it wouldn't take a month to generate a valid sample with such a method. Which means that they're randomly selecting from everyone and then seeing if the sample meets the criteria. I can think of no other way it can take so long.
Which means the problem exists between keyboard and IDE.
The only jackasses here are you and the guy you replied to.
One thing is absolutely certain - no matter how wrong I am, I'm more right than you could even dream of being by the mere fact that I looked for a solution. What's more, a solution to the correct problem. Something you and the parent post were incapable of identifying.
If you can't identify the problem, DON'T harass another's identification of the solution space.
Hmmm. IR for a long, complex chain molecule, given that even reasonably short and simple chain molecules are all in the radio spectrum. Methinks you are an idiot.
NMR... And how the hell do you propose to make a bloody 3T magnet portable??? And what the friggin' hell is the power source going to be??? Are you a complete imbecile or just practicing?
Given that the challenge is clearly intended to produce a device that can be taken to patients who can NOT be taken to a hospital with static medical devices, and given that said people are very likely in areas where there is FUCK ALL for transportation, power or repair shops, and where diseases are likely to be things like Marburg, Ebola or XDR-TB, where the devices will likely be in the hands of someone who last week had been in tryouts for witch-doctor of the month (ie: not someone you would trust with a syringe or scalpel), what the hell kind of device do you think the challenge is going to want??!!
The last thing a prize like this needs is some prize idiot like you to come up with some insane idea that is wholly unworkable in such environs. The popular phrase is "beating tractors into ploughshares".
Your utter drivel of a post is worth replying to only in that it is so utterly sickening that you could have had enough brain cells to type and yet insufficient brain cells to comprehend what is required.