Anyone interested in the proper application of probability theory to catching terrorists, and how this affects TIA-style data-mining programs should read this article:

http://www.lewrockwell.com/orig7/rudmin1.html

If you need a nice primer on Bayes', check out:

http://yudkowsky.net/bayes/bayes.html

And here a blogger applies bayes' theorem to judging whether such systems are better suited for catching terrorists or sweeping up political dissidents:

http://scat-he-g4.sunderland.ac.uk/~harryerw/blog/index.php?entry=entry060524-164934

(oh, go ahead, guess which one!)

But, but, but Tiberius saw it work exactly that way on "24", so it HAS to be true!

Indicates that an agency had previous interest in his activities -- probably an intelligence agency.

Brings to mind the "previous to the event" government interest in Sirhan Sirhan and Lee Harvey you-know-who.

... after a debilitating injury. I would have had to sign a waiver understanding that on request, the clinic would turn over all its records on me to the Secret Service or Homeland Security.

Didn't know I was paranoid until that moment. :)

Such security schemes are universally ineffective and spectacularly vulnerable to abuse. -- prunes

This was discussed many months ago, on the old Unclaimed Territory, and other database mining experts made the same point. Since then, unfortunately, their point -- and yours -- has more or less gone by the wayside in our discussions.

As I remember it, what was said was that the enormous number of false positives alone, even when using the best heuristic algorithms now available, made such systems essentially useless for the purpose which they were ostensibly built to serve, and that it was unlikely that future developments in these algorithms would substantially improve their effectiveness.

On the other hand, if you know someone named Joe Blow, and you want to find out if he ever contributed to the Democratic party, or overdrew his checking account, or rented a porno videotape, or was treated for a chronic disease, they are ideal.

It makes one wonder, does it not, if these systems weren't built for precisely that purpose, and not the ones stated. Not to protect the public, in other words, but to reduce it to chattel.

Had to go out to JFK airport in NYC yesterday to meet someone. At the first airport monorail stop, a group of 6 or 7 young soldiers got on, dressed in desert camo (at JFK?), berets at a jaunty angle, nice big black M-16s in their hands. They didn't make eye contact with any of us civilians, but were having some sort of argument among themselves and didn't seem very happy. It was a very odd feeling, looking at these kids, who -- if they hadn't been in uniform with rifles -- I would have ignored, but realizing that if I or anyone else in the train did something these soldiers didn't like...well, we could be detained, arrested, even shot, if they got really nervous. I had no faith in these soldiers and their judgment and in fact felt very insecure, especially because my husband is dark and is frequently taken for a Middle Easterner. In fact, I was scared (and thinking about how Iraqis must feel, who don't speak the soldiers' language and don't understand their customs). It was extremely unpleasant and very sobering. Everyone in the train was silent until the soldiers got off.

Big brother is watching us, and his little brother has guns.

WT: As I remember it, what was said was that the enormous number of false positives alone, even when using the best heuristic algorithms now available, made such systems essentially useless for the purpose which they were ostensibly built to serve, and that it was unlikely that future developments in these algorithms would substantially improve their effectiveness.

The technical term for this is "garbage in — garbage out". The only technological advance with the new systems is "a very large amount of garbage in — a very large amount of garbage out.

On the other hand, in addition to being able to tell what Joe Blow did with his money, you can also fairly quickly produce a list of everyone who contributed to, say, PETA or the ACLU. But what use could that possibly be to a government that has already demonstrated its willingness to use the machinery of law enforcement for partisan political purposes? Oh, ye of little faith.

Ok, accepting Mtv's (advertisers) statistic that Half of all college students become so depressed at some time that they cannot function, what happens when 60% of all college students show signs of depression?

It seems to me that we have two options. We can essentially put everyone on antidepressants, or let them be depressed and declare happy people to be abnormal. If everyone is either on controlled substances, or shows signs of "abnormal levels of contentedness", then we are approaching the 100 percent threat level. Everyone is a suspect. We are going to need some really big databases to keep tabs on everything about everyone.

Better keeping tabs on everything about everyone than having another 9/11.

[to Glenn]: Touchy today are we? Glad to hear it. Based on your other hysterical rants about the boogey man government I wasn't sure how far your fear would take you.

Yeah, Glenn's got a thing about people that put words in his mouth (or thoughts in his head). He puts a lot of effort into his writings to make it clear what he's saying.

But if you'd wanted to be presumptuous, and guess at his views on something he'd not covered, perhaps it would have been best to use the evidence at your disposal (as you seem to have done above; see bolded) on collateral matters and use that as a basis for forming an opinion. Then you wouldn't have looked like as much of a jerk.

But, if it gives you any solace, you'll find in me (a distinct minority of one, as I speak only for myself) a person who does think that a registry of gun ownership is not a bad idea. We do it for cars. Why not guns? The mere fact of ownership should not be particularly annoying (or furtive, perhaps?). I think that it would be even better if we had a physical means to tie the bullets fired to the guns they came from. Seems like an eminently useful thing to police that are trying to do their job and find out the whys and wherefores of a bullet lodged in some stiff's thorax.

I think everyone also agrees that one big problem is that criminals get guns. If we have a means of finding out how they got them (and where), don't you think we might be able to find out how to keep some of them out of the hands of people that most rational observers say shouldn't have them? <*PSSSST* "gun shows" *>

Cheers,