Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Security Government Privacy United States Your Rights Online

Bruce Schneier: Why Collecting More Data Doesn't Increase Safety 149

Posted by timothy from the but-it's-only-wafer-thin dept.
Jeremiah Cornelius writes "Bruce Schneier, security expert (and rational voice in the wilderness), explains in an editorial on CNN why 'Connecting the Dots' is a 'Hindsight Bias.' In heeding calls to increase the amount of surveillance data gathered and shared, agencies like the FBI have impaired their ability to discover actual threats, while guaranteeing erosion of personal and civil freedom. 'Piling more data onto the mix makes it harder, not easier. The best way to think of it is a needle-in-a-haystack problem; the last thing you want to do is increase the amount of hay you have to search through. The television show Person of Interest is fiction, not fact.'"
This discussion has been archived. No new comments can be posted.

Bruce Schneier: Why Collecting More Data Doesn't Increase Safety More

Bruce Schneier: Why Collecting More Data Doesn't Increase Safety

Comments Filter:

  • Fiction, not fact. (Score:5, Insightful)

    by Black Parrot ( 19622 ) on Saturday May 04, 2013 @02:21PM (#43630395)

    Good luck if he thinks he convince the American public that televised fiction isn't fact.

    • Re: (Score:3)

      by Mitreya ( 579078 )

      Good luck if he thinks he convince the American public that televised fiction isn't fact.

      Indeed. From what I understand almost everyone believes TV shows as documentaries.

      "24" convinced people that beating the crap out of suspects is often the only (and effective) way

      "CSI" convinced people that the crappiest image can be enhanced up to a perfectly clear picture in a few clicks.

      • Re:Fiction, not fact. (Score:5, Informative)

        by auric_dude ( 610172 ) on Saturday May 04, 2013 @03:12PM (#43630717)
        Television dramas that rely on forensic science to solve crimes are affecting the administration of justice via http://www.economist.com/node/15949089 [economist.com]

      • Re: (Score:1)

        by Anonymous Coward

        ... believes TV shows as documentaries.

        Whatta ya mean? Next you'll be saying chimps aren't monkeys. Or bombs work without the red wire. Or that corpses don't mummify naturally.

      • Re: (Score:2)

        by icebike ( 68054 )

        Good luck if he thinks he convince the American public that televised fiction isn't fact.

        Indeed. From what I understand almost everyone believes TV shows as documentaries.

        "24" convinced people that beating the crap out of suspects is often the only (and effective) way

        "CSI" convinced people that the crappiest image can be enhanced up to a perfectly clear picture in a few clicks.

        Oh, yes, please heap some more insult on Americans. Don't bother with a citation, just dig deep into your sack of bullshit and hurl away.
        Ask yourself who is dumber and more gullible, the guy who watches entertaining make believe-drama (and knows its make-believe drama), or the clown who assumes all americans believe the make-believe drama, simply because someone told him so.

        • Re: (Score:2)

          by Mitreya ( 579078 )

          Oh, yes, please heap some more insult on Americans. Don't bother with a citation, just dig deep into your sack of bullshit and hurl away.

          I didn't say all Americans, but the effect is common and well known. Here's some references for you if you'd like to educate yourself (the CSI thing has a Wiki article for a while now)
          http://en.wikipedia.org/wiki/CSI_effect [wikipedia.org]
          http://www.worlddialogue.org/content.php?id=460 [worlddialogue.org]

          • Re: (Score:2)

            by icebike ( 68054 )

            Did you read past the second sentence of your first link:

            While this belief is widely held among American legal professionals, some studies have suggested that crime shows are unlikely to cause such an effect.

            As for the second link, pure rubbish, which never once seriously suggests or offers any evidence that ANYONE believe the torcher aspect of the show.

            • Re: (Score:2)

              by cusco ( 717999 )
              Meant to reply to this the other day and didn't get around to it.

              I work in the security industry, and the 'CSI Effect' is very well known. Ask any salescritter in this industry and they'll tell you stories of multiple prospective clients who want to be able to do truly ridiculous things with their security cameras. The good ones gently correct them, the bad ones say, "Of course we can do that!"

      • Comment removed (Score:5, Informative)

        by account_deleted ( 4530225 ) on Saturday May 04, 2013 @06:25PM (#43631661)
        Comment removed based on user account deletion

      • "CSI" convinced people that the crappiest image can be enhanced up to a perfectly clear picture in a few clicks.

        Nah... we've been convinced of that since Blade Runner at the latest. Probably much earlier.

        • Re: (Score:2)

          by ppanon ( 16583 )
          To be fair, Blade Runner was a science fiction movie where they had flying cars. Deckard didn't just blow up a small part of the picture, he actually blew up something that was hidden from the standard point of view. The only way that made any sense was if it was some kind of 3D interference hologram that somehow didn't come across in the 2D movie projection. i.e. it's not a photograph as we know it Jim, not as we know it.

    • Good luck if he thinks he convince the American politicians that televised fiction isn't fact.

      Yes that really is better.

    • Re: (Score:2)

      by slick7 ( 1703596 )

      Good luck if he thinks he convince the American public that televised fiction isn't fact.

      The real question is...To whom do the benefits go? If you believe the Gutterment is concerned about your safety, then why the NDAA?
      It's obvious that the confiscation of firearms is virtually impossible but, buying all the ammunition they can purchase with your tax money will make ammunition for those same firearms very expensive. This in turn will cause the remaining ammo to skyrocket in price, guess who's going to sell that ammunition to you? Would you buy it, trust that it will work?

    • ... presses the enhance button repeatedly to view the person from another angle and 20ft away, using satellite images.

  • Person of Interest (Score:1)

    by Anonymous Coward

    "The television show Person of Interest is fiction, not fact." - I'd more characterize it as an fs*cking fairy tale, not just fiction.

    • Re: (Score:2)

      by cas2000 ( 148703 )

      i'd classify it as propaganda to get people to like the idea of a total surveillance police state because there's good-guy superheroes (including an ex-spook and a philanthropic billionaire. and don't forget the dog. doggies are nice and good guys are nice to dogs, it's the easiest way for you to know that they're good guys) protecting people who need protection from bad guys.

  • ... the collection of data helps after the fact, i.e., once someone is caught. The additional data allows a more solid case to be built, and makes it easier to find co-conspirators.

    • Re:After the fact... (Score:4, Insightful)

      by khasim ( 1285 ) <brandioch.conner@gmail.com> on Saturday May 04, 2013 @02:36PM (#43630491)

      The additional data allows a more solid case to be built, and makes it easier to find co-conspirators.

      Yep. So the "compromise" could be lots of data collected but only kept for a short time (weeks, not years).

      On the other hand, the frequency of any threats is so rare that do we really want to erode our liberties like this? Is regular police work just not capable of "connecting the dots" without this kind of surveillance?

      Fascism begins when the efficiency of the Government becomes more important than the Rights of the People.

      • So the "compromise" could be lots of data collected but only kept for a short time (weeks, not years).

        Or requiring a warrant to access the data.

        • Not a good idea, 9/11 onwards, they have to always include a clause along the lines of: "In cases where the officer writes down any of these phrases 'terrorist', 'child pornography', or 'national security', the officer may then immediately access the information, and then, if they want to, may later apply to the FISA court where a warrant will automatically be granted."

        • Re: (Score:1)

          by Anonymous Coward

          Given the government's track record for such things, I'd rather not have the data be collected at all.

      • Actually, the fascists really don't give a damn about any efficiency as long as they can run the world.

        The MIC creates "programs" (Bush mentioned that many times), which in reality are the creation of haystacks so that taxpayer money can be spent on these MIC "programs" looking for the needles in the created haystacks that contain no needles. But the MIC gets the money anyway. And since there are no needles in the fabricated haystacks, (damn, that "program" was not funded enough, we need more money, we

    • Re:After the fact... (Score:4, Insightful)

      by DarkOx ( 621550 ) on Saturday May 04, 2013 @03:06PM (#43630685) Journal

      .. the collection of data helps after the fact, i.e., once someone is caught. The additional data allows a more solid case to be built, and makes it easier to find co-conspirators.

      I'll buy that. Once you know who you can go back and sift through logs, security camera footage, peoples cell phone snaps, phone records, etc and find evidence. I don't Bruce would argue otherwise.

      But...Where mass murder and terrorism is concerned what is our objective? Make sure we can punish the guilty or prevent attacks?

      So far I am not aware of any revelation that has come out of all the surveillance that would have helped us 'prevent' the bombing. Plenty of things we might have done, but all things we already knew we could be doing but had rejected for reasons of civil liberties, cost, character of our nation etc.

      Its also entirely possible that something that helps us identify and punish the guilty after the fact harms our ability to detect and prevent in terms of to much hay.

      • Re: (Score:2)

        by icebike ( 68054 )

        But...Where mass murder and terrorism is concerned what is our objective? Make sure we can punish the guilty or prevent attacks?

        So far I am not aware of any revelation that has come out of all the surveillance that would have helped us 'prevent' the bombing. Plenty of things we might have done, but all things we already knew we could be doing but had rejected for reasons of civil liberties, cost, character of our nation etc.

        You are exactly right, there is nothing that would be effective in "preventing" the bombing which would not render the country a total police state.

        Yes, we lost a few lives, and yes lots of people were hurt.
        We lost a hell of alot more lives building a nation where we have the right to walk down town with a backpack without being stopped and questioned. (Except perhaps if you are Black and live in some portions of NYC).

        There are simply not enough police to tail every miscreant or potential felon in the coun

        • Re: (Score:2)

          by DarkOx ( 621550 )

          The certainty of being caught is the best weapon we have right now.

          So what you are saying is that we have no effective weapon. The 911 terrorists did not even expect to survive if successful. I don't think they had much concern about being caught.

          It appears the Boston bomber killed his older brother rather than allowing him to be taken alive. When your would be attackers don't value their own lives punishment capital or otherwise is not an effective deterrent.

          So back to the original questions. Is the data collection itself turning our society into something different

    • You REALLY think suicide bombers are bothered by the fact that you'll know they did it after they did it?

      • Replying to myself, I know, but...

        You don't need surveillance to find that out, you just have to solve the puzzle...

      • Re: (Score:2)

        by icebike ( 68054 )

        You REALLY think suicide bombers are bothered by the fact that you'll know they did it after they did it?

        In the Boston case, they apparently had no stomach for suicide, and were gullible enough to believe they would get away with it (they didn't even try to leave town). Only after they saw their pictures all over the TV did they try to steal a car.

        These were not real bright guys.

        And neither are your average suicide bomber from what I have been reading. The "true believers in the cause" have pretty much been expended and the terror masters now prefer the weak minded and gullible with nothing to lose.

        • Re: (Score:2)

          by cusco ( 717999 )
          If they had bothered to disguise themselves AT ALL they would have gotten away with it too, and the FBI would have had to frame someone for the attack. A blond wig, big sunglasses, a modicum of makeup (even inexpertly applied), a new change of clothes, and a wide-brimmed hat would almost certainly of made pretty much all of their photo evidence useless.

          nothing to lose.

          The FARC and possibly the PLO have been known to find deeply-indebted people with terminal diseases and offer to pay off their debt a

  • A lack of concern for freedom. (Score:4, Insightful)

    by Anonymous Coward on Saturday May 04, 2013 @02:28PM (#43630457)

    The main problem here is that people just don't seem to care about freedom if they believe that something will keep them safe (or at least makes them feel safe). Even if it were true that the TSA, ubiquitous government surveillance, free speech zones, the Patriot Act, and warrantless surveillance in general kept people safe, that wouldn't make them any less wrong. Indeed, the main problem is that people seem to generally be spineless cowards who give up freedom for safety and are easily manipulated (especially after a disaster).

    • then they deserve freedom nor liberty.

    • Sort of, the right seems to care more about their right to bear arms than to rights that are actually meaningful in day to day life. If we ever get to the point where private ownership of firearms is going to make a difference, we've got more serious problems on our hands.

      Focusing on things like real trials rather than show trials and actually having an independent judiciary would make a much larger difference in the long run.

      • Re: (Score:2)

        by jedidiah ( 1196 )

        Guns just represent liberty and self-reliance. They are the start of the slippery slope that includes spray paint, decongestant, pressure cookers, and sharp kitchen knives.

        The gun grabbing mentality preaches that we are all helpless and need to wait for the nanny state to sort things out should anything go wrong. It implies a contempt for the electorate that should be more obvious to more people.

    • So sayeth the Anonymous Coward.

  • Uh (Score:2)

    by trifish ( 826353 )

    Needle-in-a-haystack problem? Really? Seriously, in the post-PC era... data mining gets more difficult as the amount of data increases? uh... I've always thought that to gain any meaningful stats, you need a large enough sample...

    • Re:Uh (Score:4, Insightful)

      by maxwell demon ( 590494 ) on Saturday May 04, 2013 @02:56PM (#43630617) Journal

      This is not a problem of statistics, this is a problem of identifying individual terrorists. Even if you could determine exactly how many terrorists there are, it would help you absolutely nothing to prevent the next terror act. You have to know who the terrorist is.

      You can stare at the weather statistics of the last ten centuries as much as you want, it won't help you much when trying to predict when and where the next lightning will strike.

      • I don't disagree with your point, but I think the analogy is flawed.

        If you can stare at weather statistics with sufficient data to see what circumstances resulted in lightening strikes, then you can accurately predict where extra effort needs to be taken to avoid them. In fact that sounds really useful and much like what we do with tornados for example. The statistics don't tell you that there will be a tornado at a certain place at a certain time, but the do tell you when they are likely enough to sound th

    • The opposite. (Score:5, Insightful)

      by khasim ( 1285 ) <brandioch.conner@gmail.com> on Saturday May 04, 2013 @03:01PM (#43630641)

      uh... I've always thought that to gain any meaningful stats, you need a large enough sample...

      That works for trends. Not for the actions of individuals.

      From TFA:

      Rather than thinking of intelligence as a simple connect-the-dots picture, think of it as a million unnumbered pictures superimposed on top of each other.

      He's a bit wrong there. It isn't a million unnumbered pictures. It's one picture per person in the country at the time. That's over 300 million pictures. Each one overlapping millions of other pictures.

      uh... I've always thought that to gain any meaningful stats, you need a large enough sample...

      And after a certain point you are just amplifying the "noise". And enough "noise" can appear to be a pattern.

      It is only after an event that the "noise" can be filtered out and the extraneous pictures discarded.

    • As the sample set's size tends to infinity, so does the computational power and/or the time required for effective mining (ceteris paribus, of course).

    • Re: (Score:1)

      by Anonymous Coward

      Until you are awash in data. Then you run into a practicality wall. Not to mention limitations on what variables are actually meaningful (i.e. What actually is meaningful? How meaningful? What variables don't we know about? What about counter indicators? etc). After all, these systems are well known for disturbingly large false positives and false negatives.

    • In statistics you want a large sample so that outliers will be obscured and the overall trends can be discovered. In security it's the outliers you care about, and the trends of what the general population is doing don't really matter. Thus a large sample will be counterproductive.

    • The problem is that you have to store and process all of that data. And much of that data isn't in a form that the computer system can process. Which means that they're storing tons of data that they'll never be able to use, or will at most be able to use it after they've determined whom to arrest. But, in terms of prevention, which is what safety is about, it doesn't do you any real good.

      Remember that statistics can talk about populations accurately, but if you try and take that description and apply it to

    • " uh... I've always thought that to gain any meaningful stats, you need a large enough sample..."

      Don't think. You weaken the nation. (seriously)

    • Re:Uh (Score:4, Insightful)

      by phantomfive ( 622387 ) on Saturday May 04, 2013 @03:49PM (#43630925) Journal
      Your sample should be large enough to have what you're looking for, but no larger. He mentions that the FBI has over 700,000 people on its watch list. They don't have enough people to investigate them all. If they could narrow down that list to 500 serious potential terrorists, their job would be a lot easier.

      How to accomplish that? The simplest way is to catch them right before they are about to attack. For example, we could read the minds of individuals who are experienced in seeing the future, call them pre-cogs. Then when they are in agreement, we can catch the terrorist with our future crime force, lead by Tom Cruise.

      Just kidding. Bruce Schneier doesn't give an plan on how to stop future terrorists, his point is that there's no reason to shred even more civil liberties in order to try to catch terrorists, especially since it probably won't help.
    • yes for some ML problems more data is best
  • I suppose that it is always easier to collect data than to intelligently analyse the amount of data gathered, just because sensors are less complex devices (to put it simple: compare the time you need to collect a cluster of data vs. the time you need to find interrelations). And, obviously, the problem gets worse with growing data (since the days of cluster analysis).

    Obvious, but Bruce does good marketing.

    CC.

    • It is always possible to collect data, and simply save it. Nobody has to search, nobody has to listen.
      Until, maybe a year or two later, when a PERIOD of Interest is identified, which reduces what is to be searched immensely.

  • The lies from monsters like Schneier (Score:1)

    by Anonymous Coward

    The purpose of intelligence agencies in powerful nations has NOTHING to do with threats from 'criminals' and 'terrorists'. No, it is about giving the true masters of these societies the most perfect control possible of the 'mob'.

    Today, the internet allows those that rule you to get feedback in real-time that explains the effectiveness of ANY governmental PR campaign disseminated by the mass media. If the response of the sheeple is 'wrong', the message can be immediately re-engineered and broadcast again. Re

    • What good did all the control and data for the USSR and East-Germany? The latter really had perfected the art of its citizens spying on its citizens and still: It just collapsed. You can't really control a population. There are just too many people and when they decide to do something all your control is moot.

      That's actually the point Schneier tries to make: From a certain point on collecting more and more data the ROI (and eroding civil liberties is one of those "investments") just isn't there anymore. You

  • Didn't the FBI say something similar (Score:4, Interesting)

    by Art Challenor ( 2621733 ) on Saturday May 04, 2013 @03:32PM (#43630821)
    Didn't the FBI make a similar comment after it was revealed that they had questioned the Boston bomber in the past? Something to the effect that they could not follow up on every suspicisous character without turning the country into a police state.

  • False positives (Score:3)

    by gmuslera ( 3436 ) on Saturday May 04, 2013 @03:38PM (#43630849) Homepage Journal
    Is not just collection. You face consequences for false positives [theblaze.com]. And anything you said could be used against you, even if a joke in a private mail (if you ever said something they didn't like).

    So you are walking in thin ice, you could get big charges for something that you don't see as a crime (or see it as a joke or a prank). And people do weird things in that kind of situations,

  • It was never about increasing safety. Loss of civil rights is a feature, not a bug.

  • I suppose this is a problem, but we have to be realistic here. The FBI has failed repeatedly to complete large software projects. They have trouble handling even clear-cut data. http://en.wikipedia.org/wiki/Virtual_Case_File [wikipedia.org]

  • Bruce Schneier facts (Score:5, Funny)

    by KugelKurt ( 908765 ) on Saturday May 04, 2013 @09:04PM (#43632263)

    Bruce Schneier doesn't need to hide data with steganography - data hides from Bruce Schneier
    Bruce Schneier knows who the Anonymous Coward is
    Bruce Schneier can recite pi. Backwards.
    Bruce Schneier can securely wipe any hard drive by shaking it like an etch-a-sketch.
    Bruce Schneier knows Chuck Norris' private key.
    Bruce Schneier can write a recursive program that proves the Riemann Hypothesis. In Malbolge.
    Bruce Schneier can read captchas.
    Hashes collide because they're swerving to avoid Bruce Schneier.
    Bruce Schneier is the root of all certificates.
    Bruce Schneier intercepts all your internal monologues by a man-in-the-middle attack.

    http://www.schneierfacts.com/ [schneierfacts.com]

  • While Collecting More Data Doesn't Increase Safety, authorities might not have been able to track down the bombers in Boston before they made it to New York. I am not advocating big brother, but if your in a public place you really can not expect any privacy. There are just too many idiots out there. Americans simply rely on the government to take care of them.

  • The question is if the needle to hay ratio is better in the added hay.
    If there was no needles in the original haystack, adding more hay may add a needle.

Slashdot Top Deals

One man's constant is another man's variable. -- A.J. Perlis

Close