— How I’d Sift Through Your Phone Calls — It’s Conspiracy Time!

Well, the wide-scope security program dubbed PRISM, Edward Snowden, and the National Security Administration (NSA) have been in the news lately.  But not because Government wanted to discuss PRISM, Snowden, or the NSA.  Not because they were being transparent but, because a whistle-blower informed everyone of what was going on.  Government first denied it, until they realized the story wasn’t going away, and then they ultimately had to fess-up that ‘perhaps, they ARE collecting your data; but that they are not “listening” to your actual phone calls (despite the fact they are recording them).  Government may know when you’re on your phone, where you’re at when you’re on your phone, and whose number called you [metadata].  They are archiving videos, actual phone calls, social media activity, and emails [which they claim you don’t have an expectation of privacy for].  But, they remind you that they are not “LISTENING” to every single phone call.  They joke that its impossible because they don’t have enough staff to do such a thing.  So that should provide a measure of comfort… right?!?!

Not So Fast

HYPOTHETICALLY of course!  Let me explain what I would do If I was in charge of a government and didn’t particularly care about people’s liberty and I wanted to know what was being said on a large portion of phone calls that I recorded but didn’t have the ability to have a human listen to.  I’d use technology. The same technology that allowed me to collect the data in the first place actually.  Here’s How I’d do it in 3 steps:

— STEP 1) Record the calls to storage memory

— STEP 2) Convert the calls from Voice-To-Text

— STEP 3) Search for Keywords

Pretty simple, and all of the tools are commercially available.  But just for poops-and-grins, let’s look a bit closer.

[Step 1] – Get Access

I’d pressure (but I’d call it “Partner”) with companies like Verizon, AT&T, and other owners of Cellular Towers and get access,. I’d use a FISA (“Foreign” Intelligence Security Act) Court to give the partner companies top-cover so they don’t get sued into bankruptcy.  After I had access to that data, I’d Bluff in UtahArchive/Record it.  Obviously, I’d need large data storage capacity to hold all of those recordings, and since there isn’t one, I’d build it.  It’d become the largest running “Department Of Defense” construction project EVER. I’d pick a place, kinda isolated in hopes people don’t realize how big it is and just hope no media picked up the story.  I’d also appropriate the surrounding area and build it close to a military base to limit the growth around it for the foreseeable future.

I’d give the facility enough storage to maintain all of the phone calls that have been made since the telephone has been invented. I’d ensure that it had enough power to supply it and I’d give it 6 or 7 generators and enough fuel storage to keep it running for 4 months in order to ensure power to the battery back-up systems in the case of any long-term external power disruptions.

[Step 2] – Throw Software and Processing power at it

The most important part of my plan would require me to figure out how to sift through so many phone calls when doing it manually would require hundreds of years (even if all phone calls were to stop now) with tens-of-thousands of humans.  Hmmm!  How would I do that?

I got it!  I’d partner with companies who have been utilizing the technology of turning voice into text for years.  Yes, I’d partner with Apple and learn how Siri works. Siri is the App on  millions of IPhones that utilize statistical analysis to determine what words a human is saying, yada yada yada.  It turns your words into text and utilizes those words to complete tasks such as writing emails, finding directions, or sending text messages.  While partnering with Apple, I’d also partner with Nuance.  Their Dragon Dictation program would work perfectly for my need to turn voice into text.  I would reverse engineer it and run it 24 hours a day, seven days a week, searching for those suspicious words I fear.

[Step 3] – Sort the Text

With all the words converted into text, I’d sift them for keywords or algorithms that I believe will lead to suspicious activity.  Obviously, this will never lead me to a terrorist, because terrorists don’t use the words, “terrorist”, “terrorism”, “bomb”, “explosive”, “jihad”, “car-bomb”, “target”, or “secret” (or any other words that are causing the NSA to flag this page)!  If the computers DID find something suspicious (which they would not), they’d be tagged and scheduled to a human operator.

Luckily, I’ve Sequoia_ibmbeen building Supercomputers for many years in conjunction with the Department of Energy and Lawrence Livermore National Laboratory.  As a matter of fact, the NSA, has 2 computers on the list of the Top 10 fastest Supercomputers on the planet with Sequoia (#3) and Vulcan (#8).   I’d buy 2  BlueGene/Q computers based on the highly energy efficient ‘Sequoia and house them in two separate basements at my newly built facility.  I’d use them to translate “targets” (phone call recordings) and they’d be kept running all night and all day.

Yes, I know I do technically need a judge to grant permission for me to actually listen to any recorded calls, but the fact that the computers are listening means that a human, is not!  If a computer finds a call suspicious… that means there is a reasonable suspicion for me to investigate and listen to those calls further, so I wouldn’t actually need to get a warrant from the FISA court.  This is for national security afterall!!

But hey, What Do I know?  That’s just how I’d do it!

Interesting:  DARPA Wants You To Transcribe, And Recall All Of Your Messages.

Now They Get It.  How the NSA Converts Spoken Word To Text

Kali Pinckney

What's On Your Mind??

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s