NSA, The Technology Future, and Where It Is

cowmix writes "It was weird watching 60 Minutes II last week when the head of the NSA was complaining that his organization was totally behind in technology. Further, he told of stories of the organization's horrible inefficiencies and even went into how at the first of January 2000 all the computers in the NSA were down for three days. The thing that really shocked me was seeing pictures of the inside of one of the NSA headquarters and also SEEING people decoding telephone conversations. I didn't know what to make of it."

Comments

What are they up against?

[dragons_flight]

Bullshit warning: I'm about to pull a lot of numbers out my ass. I hope to be semi-reasonable and conservative, but it's guesswork nonetheless.

Let's suppose for the sake of argument the NSA can in fact intercept any transmission and beyond that can convert any spoken words in any language to flawless text.

5 minutes of phone time per person per day worldwide
6 billion people
at least 1 word every 3 seconds
2 people in the typical conversation
8 character average word length (w/ space)
= 2.4 Terabytes per day

200 important daily newspapers
50,000 words per issue
= 80 Mbytes per day

5,000 magazines / periodicals
median time of 2 weeks
100 pages on average
average 400 words per page
= 114 Mbytes per day

15,000 worldwide radio stations
35% of time is spoken
1 word every 2 seconds in spoken segments
= 1.8 Gigabytes

7 million new webpages a day (source)
10k average size
= 70 Gigabytes per day

500 million email users
average 0.5 email sent per user per day
18k average email size (source)
= 4.5 Terabytes per day

Total = 7 Terabytes per day

If the NSA really were out to track everything, suffice it to say, it's one monster of a computer engineering problem. We are generating more information than ever and don't have the same kinds of well defined enemies. And how many actual analysts are required to make any sense of all that? Is it any wonder they might be falling behind?

Of course I'm sure there are lots of sources of information, such as TV, that I haven't even covered.

Re:Are we supposed to believe this?

[The_Steel_General]

Believe it. What he said (bureaucracy, office politics, inefficiency) was exactly my impression when I worked there not so long ago.

It is, furthermore, the real reason why I get disgusted by NSA's anti-crypto stance: It's about protecting their jobs exactly as they are today. There's this expectation of entitlement, that because they've always been able to decrypt some significant percentage of messages, they should always be able to do so. Adapting to changes in technology? Hey, that's for the rest of the world, not us. Focus on weak links, traffic analysis, other techniques forced upon us in the past? C'mon, there's only 8 hours in a day -- we'll just outlaw anything that would make our work more difficult.

It's resulted in absurdities like encryption jobs (and know-how!) moving to other countries, CSS's unusually easy-to-break "encryption," and t-shirts classified as munitions. Way to go, guys.

I will certainly agree that it might cost more, but I, too, would like some assurance that Congress isn't paying them to remain clueless bureaucrats. I don't insist that they open up every line item throughout their budget -- just some acknowledgment of their new, post-Cold War situation. I would love for DIRNSA to get in front of Congress and say "Okay, we can't count on being able to break the encryption on any message out there, so we're changing the focus of our efforts to X, Y, and Z. We'll continue encryption research, try to figure out the best way to crack existing schemes, but our efforts have to take into account the rising tide of encryption technology use. But for that to be successful, we'll need more money because..."

Would that be so hard?

TSG

The NSA is always at war = always deceptive

[Infonaut]

Think of it this way - the NSA has to always operate under the assumption that it is in a state of war. The more information it can gather, the more knowledge about potential threats it can accumulate, the better. The weaker it appears to its potential enemies, the better.

Attila the Hun actually almost never outnumbered his opponents. He won using carefully-crafted deception plans and sheer terror to demoralize his enemies.

The Allies were able to intercept and decrypt a huge chunk of Nazi messages throughout WWII as a result of their ongoing effort to crack Enigma. These decrypts probably shortened the war in Europe by months if not years, but they had to use the intercepts wisely, so as not to tip off the Germans.

During the 1950s, the Russians talked about atomic bombs 'rolling off the assembly lines like sausages', when they actually had a very limited stockpile.

The point is that sometimes you deceive your enemy into thinking that you're stronger than you are, and at other times you make them think you're weaker than you actually are.

Intelligence agencies are any nation's first and last line of defense. They're the ones that tip off leaders about potential dangers, well before they surface on CNN or in the pages of the Washington Times. They're also the ones who can provide the necessary misdirection so that critical programs are not detected by the intelligence resources of other nations.

Case in point: The F-117 Stealth Fighter. Remember when Testor's came out with a plastic model of what they thought the Stealth looked like? The Pentagon freaked out on Testor's and tried to keep them from selling the model kit. Of course, when it was revealed a few years later that an F-117 group had actually been flying *operationally* for several years, and that the Stealth fighter looked nothing like the model, we could all see the depth of the deception effort.

If the NSA releases its doors to the television cameras, *particularly* to 60 Minutes (which has a long history of not having a clue about defense-related matters), it's part of an extensive deception plan.

They're just doing their job.