Big Brother Builds Better Bandwidth, But Will There Be A Benefit?

The NSA is building a new “data center” in Bluffdale, UT outside of Salt Lake City. And while plenty of people think their acronym really stands for Never Say Anything, one of their former employees has said plenty that informed a March article about the new facility at wired.com. The long and short of it: our security-minded friends will be able to tap into and store pretty much every email, text, phone call, bank transfer and retail transaction. As in all of them. Yottabytes of information compiled at exaflop speed. The question is, will they actually be able to sift through all of the chaff (think billions of grocery store purchases, bar tabs  and online porn membership transactions not to mention your daughter’s texts about Jimmy’s hair and how cute Leslie’s boots were) to get to any kernels of actionable data? That’s what Holman Jenkins is wondering after the Aurora massacre . . .

In other words, once this $2 billion dollar facility is up and collecting, the potential for harm is obvious. But could it actually do anyone any good – as in identifying the next James Holmes out there?

The Colorado shooter, Mr. Holmes, dropped out of school via email. He tried to join a shooting range with phone calls and emails going back and forth. He bought weapons and bomb-making equipment. He placed orders at various websites for a large quantity of ammunition. Aside from privacy considerations, is there anything in principle to stop government computers, assuming they have access to the data, from algorithmically detecting the patterns of a mass shooting in the planning stages?

Oh right, privacy considerations. Yeah well those have pretty much gone the way of the Oldsmobile. And while few will mourn the Cutlass Ciera, plenty of folks like their privacy – or what’s left of it.

But back to the matter at hand.

After the Aurora theater massacre, it might be fair to ask what kinds of things the NSA has programmed its algorithms to look for. Did it, or could it have, picked up on Mr. Holmes’s activities? And if not, what exactly are we getting for the money we spend on data mining?

The Bluffdale facility will go online some time next year. But even knowing that all that high desert computing power is being shepherded by a small army of the best bunch of boffins tax dollars can buy, we think we’ll still keep a mohaska strapped to our waist. As Americans have discovered time and time again, relying on the competence and efficiency of our intelligence community is a great way to get yourself killed.

19 Responses to Big Brother Builds Better Bandwidth, But Will There Be A Benefit?

  1. avatarchris says:

    I saw this movie before… didn’t turn out too well for Shia LaBeouf…

  2. avatarPhil says:

    I seriously doubt the NSA has enough computing power and time to analyze every piece of information that the article mentions. I don’t doubt that they do analyze lots of information, but the amount we put out there grows every day and there is no way they can keep up, so determining someone likely to commit a massacre is almost impossible, especially since that list would be miles long.

  3. avatarrosignol says:

    The question is, will they actually be able to sift through all of the chaff (think billions of grocery store purchases, bar tabs and online porn membership transactions not to mention your daughter’s texts about Jimmy’s hair and how cute Leslie’s boots were) to get to any kernels of actionable data?

    Do they have sufficient processing power and clever enough algorithms? Yes. NSA are the best there is.

    Did it, or could it have, picked up on Mr. Holmes’s activities?

    Very unlikely. Domestic lone-nuts are difficult to find, what most intelligence agencies try to do is intercept communications between members of an organizational cell. Lone-nuts aren’t part of a cell, thus, no comms to intercept.

  4. avatarIdahoPete says:

    I am shocked! shocked! to see you put disparaging comments in your website concerning one of our greatest, finest, most wonderful intelligence agencies! These selfless workers spend hours and hours reviewing billions of conversations online, and it is not their fault that many of the conversations from terror groups are in weird languages they don’t understand. Fortunately, it is much easier for them to intercept potentially harmful plots discussed in English by enemies of the Obama administration (oops, sorry, that should be “enemies of the people”). And why would they want to intercept the communications of a brilliant academic like James Holmes? After all, he was funded by the Obama administration’s NIH, right? So he must have been thoroughly screened by the government. Just like that Army major [don't call him a Muslim terrorist] who killed the 13 soldiers in another highly effective Gun Free Zone at Ft. Hood, Texas. (Ooops, that should have been “allegedly” killed 13 soldiers.)

    Anyway, we should all love the government’s ability to monitor every word we type, print, speak and think. After all, it’s for our own good. Oh yeah, and it’s FOR THE CHILDREN! So there.

    OOMMMMMMMMMMMMMMMMMMmmmmmmmmmmm

  5. avatarjwm says:

    yeah, all the intel people got caught flat footed by the fall of the berlin wall and the collapse of communism in russia. and that was a huge event that they should’ve seen coming. based on a couple of contacts with intel people in the distant pass i really don’t put a lot of faith in the breed or their handlers. i mean cia are supposed to be covert,right. but every one of them carried a browning hi power or a variant on the karl gustav. not hard tells to spot.

  6. avatarMatt in FL says:

    RF, your link to wired.com in this post goes to the NYT story related to the Chicago post above this one.

  7. avatarMark says:

    NSA stands for “no such agency”.

  8. avatarAharon says:

    “The question is, will they actually be able to sift through all of the chaff to get to any kernels of actionable data?”

    Good question. For a number of years they have been sweeping up and storing emails, cell phone calls, etc. I don’t know how interactive and advanced their systems are regarding seeking out and filtering the information to pro-actively move on it or if it is only intended to compile info in a central location for government snooping on specific targets.

    It is also possible they have software that will track and copy every online post and comment we make.

  9. avataruncommon_sense says:

    It does not surprise me that our government will record every e-mail, text message, voice call (whether from land lines or cell phones), blog entry, bulletin board entry (social networking sites), and electronic purchase. All of that data will be available for data mining. (Voice recognition software will convert recorded audio to written transcripts.) And people can write algorithms to search through that data for whatever they want.

    I doubt that the facility will be able to prevent much from happening for two reasons:

    First of all, lots of people and organizations have no idea what they are going to do from one moment to the next much less over the course of weeks or months — the common complaint that “their left hand doesn’t know what their right hand is doing”. All the computing power in the world pales in comparison to the human brain. If people “in the know” don’t know what is going to happen, how would a computer know what is going to happen?

    Second, actual people — analysts — would review potentially troublesome activity that the algorithms spew out. There are only so many analysts and they can only do so much.

    I think the main application of the facility and data will be twofold. It will enable law enforcement to build a case for prosecution after a suspect committed a crime and law enforcement has a suspect. And it will enable law enforcement to proactively monitor high profile people that they have reason to believe — from traditional methods — will commit illegal activity.

    Anything much beyond that is science fiction. Yes, computers can crunch data much better than humans. But when it comes to evaluating and judging human behavior, computers will never have that capability.

  10. avataruncommon_sense says:

    Important F.Y.I.

    Unless you use your computer at a free public Wi-Fi hotspot that doesn’t have surveillance cameras, anything you do with your computer is traceable back to your home or business where you used your computer.

    The only communication methods and transactions that are absolutely secure are face-to-face conversations and written messages on paper or memory storage devices that are hand delivered to the recipient … assuming of course that no one is eavesdropping on you!

    • avatarmatt says:

      Unless you use your computer at a free public Wi-Fi hotspot that doesn’t have surveillance cameras, anything you do with your computer is traceable back to your home or business where you used your computer.

      You can always use your neighbors WiFi. WEP can be bruteforced, and WPA has rainbow tables available. High gain antennas would allow you to go further. And there is always Tor and proxies in foreign/hostile nations.

  11. avatarSanchanim says:

    Ok well I can see the gun relationship here, but I digress.
    Grab a cup of coffee and pull up a chair. this might be a bit long.

    I personally work for a company which does diagnostics data correlation for certain devices. I won’t say which and how, but needless to say I have knowledge in that arena. I also worked for the better part of a decade for government contractors. I know what a SCIF is and and can throw acronyms with the best of them.

    At first glance this seems like the TV show person of interest doesn’t it. Big brother watching and trying to figure out ways to stop crime from happening. In reality I don’t think this is the case.

    First lets think like Dan has stated, we are talking about yottabytes of data. That is A CRAP LOAD! Even with the inbound band width you would need to store it somewhere. The amount of hard drives space, or probably more practically memory needed to hold this is massive. Power and facility space needed to hold all this is a computer nerds wet dream. Even with the inbound bandwidth that they are talking about it will have a hard time keeping up.

    As an example lets say they parse information from RAM “memory” so they reduce needed IO on hard drives and can quickly weed out “I luv you” text messages etc. The amount of computing power needed to correlate this amount of data, and then store long term to make anything of use is almost unrealistic, except this is the government. Your average teenager sends 3000 text messages a month just to get a starting point.

    Any long term data storage would need to happen on disk, so you would need a complex system, maybe something like IBM’s Netezza appliance only on massive steroids to hold it all. Now lets say you get it on disk because of a key word search or some other algorithm like source. Then the question becomes what is long term, an hour, a day, week, month? Lets say something manages to actually get pulled and stored, we then need HUMINT to actually check it out, follow it to ground.

    What would be a reasonable response time? How would someone know what is a real threat and what isn’t? Would we wind up with our LEO’s being over tasked running after people who just might have had a fight with their girl friend, went and got some ammo because they were going hunting next week, filled up their car with gas, and got some fertilizer at OSH for their garden? Let’s not forget this needs to be correlated by date as well, and given the specific issue with Aurora, the items were purchased over months, not days, so we would need to be able to store items for long term. If there is lag time from the time of purchase to being posted, and then having it actually parsed by what ever back end system they are using, the thought of trying to run through that much data is giving me a headache!

    As we all know many terrorist activities like 9/11 took months, even years to plan and execute. The amount of data that would need to be parsed in order to find any viable correlation would be astronomical, unless we profile by racial origin, travel destination, known affiliations, etc. This still probably wont solve the loan wolf issue, only a high profile organized activity. In this case we would wind up with a secret list of folks, I am sure we here at TTAG would all be on it, to watch. Let’s say you can then cut your watch list by 70%, then we can actually try to manage all this. How many folks as an example searched for how to make a bomb after the Aurora incident. this might cause the system to bonk for lack of a better word from false positives.

    Now we have to look at things like, cell phones, maybe they aren’t under the persons name, or are a burn phone, or let’s say credit cards are used under multiple names, or cash is used. How do you track that? We know we can use multiple proxy servers to mask our communication origin, so that would be another level of information you would need to sift through. Email accounts, phones, social networking sites don’t require to verify who you are. Mix that in with using some sort of code worded conversations and well this whole facility is pretty much useless, unless you have an algorithm for matching up unusual conversations. Also note that to record, or otherwise monitor voice communications en mass and then have some system to translate every known language, and dialect known to man kind would require more technology and space than this facility. It would all have to be parsed real time and then flagged or stored.

    Now this brings me to my current job with diagnostics. We only use 200 or so KPIs and correlate that across thousands of devices. The amount of power required to do this, and time is pretty high. You have to realize this is a tiny amount of data, less than 1 Mb. Now most electronic transactions are pretty small as well, however you are talking about a few orders of magnitude larger in sample size. The time it would take to parse all this would be ridiculously long even with super computers at the helm. To realistically parse this kind of data you would have to diversify your collectors across the US to pre-process information. What would go back to the facility would only be items of interest based on one or more positives. So unless the NSA is placing black boxes in every data center across the US I don’t think this experiment will go very well for them.

    You then also have an issue where unless legislation is passed to gather all this and every credit card company, cell phone maker, brick and motor store hand you an information packet regarding privacy this is a law suit waiting to happen on a massive scale. The first time someone gets swatt’d for a false positive, this system will come under such scrutiny. Even now the government is formulating an online privacy and communication act. Part of this is to protect us from prying eyes. Unless there is a line the precludes the NSA which I doubt they won’t get a special hall pass. What I could see is the use of this facility if a warrant is granted to track specific individuals, but this may or may not help.

    My only major fear is that with the current administration things are being done without Congress by executive order, so our Constitution is being thrown out the window. The question is how much of our rights and liberties are going to be trampled before folks stand up and fight to put a stop to it?

    If ya got this far thanks for reading!

  12. avatarAccur81 says:

    Dear Big Brother,

    If you could do a better job of securing the border, that would be super. I’ll secure my own property, at least until (the rest) of the police arrive.

  13. avatarLevi B says:

    Big brother is already here and watching.

  14. avatarOHgunner says:

    In the words of someone I talked to a few days ago…

    “You know, I remember when people would talk about this stuff they were crazy conspiracy theorists. Nowadays, it’s just the news.”

Leave a Reply

Please use your real name instead of you company name or keyword spam.