HAL 9000 Meets Big Brother?
What if the core issue in the warrantless wiretapping arguments currently raging is one that nobody has mentioned? What if the real secret that has yet to be exposed is a logical next step in technological wiretapping capabilities, but one that our legal system has never been faced with before? What if -- in essence -- computers are the ones deciding which calls to tap, rather than NSA agents or judges?
Before I get started here, let me state for the record that everything in this article is pure and utter speculation. I do not have access to anything other than publicly available documents. I have no secret sources. I do not claim anything I say here is the truth -- merely rampant speculation and educated guesses.
Mostly, this means paying attention to what is not now being said by all concerned. The Boston Globe has a good article which outlines the current debate in Congress, and what the Bush White House is pushing for RIGHT NOW, before Congress' summer vacation.
The elephant in the room nobody's talking about, though, may be a seemingly natural technological step but an enormous legal chasm which must somehow be bridged: what legal rights and responsibilities does a machine have as a part of our justice system?
As it is written into law now (including the Foreign Intelligence Surveillance Act or FISA), the basic process is thus: legally acquired evidence comes to the attention of a law enforcement agent. The law enforcement agent outlines the evidence, and applies to a court for a search warrant (for a wiretap, for instance). The judge considers the evidence, and either issues the warrant or declines. The law enforcement officer, warrant in hand, then taps the person's phone and gathers evidence against the authorized person. Such evidence is legally obtained, and can be used as evidence in a court case charging the target with a crime.
That's it in a nutshell. There are two human beings, the agent and the judge, who both assumably make measured, rational decisions about the possibility that someone has committed a crime or is still in the process of committing a crime.
Now, ever since the warrantless wiretapping story originally broke, the discussion has focused (for good reason) on whether it is being used against foreigners or Americans, and where each is standing when the wiretap takes place -- in America or on foreign soil.
But what if those aren't really the sticking points? What all the smoke and mirrors surrounding President Bush's wiretapping may really be obscuring is the fact that they're trying to introduce a new paradigm in American law. Here's what Bush may actually be asking Congress for the right to do: A system of computers vacuums up vast amounts of data traffic, uses sophisticated programs to analyze and filter these communications, and spits out the ones that the software has been programmed to flag for human attention.
If that's true, that would require massive changes to the legal system. Because it is then the computer who decides what could be evidence and what probably is not. If the computer is programmed to think a certain communication is fishy in any way, then it spits the message out for a human being to read. But here's the massive legal problem with this -- the evidence has already been collected. The only evidence to take to a judge to get a wiretap warrant may indeed come from a wiretap. And that wiretap may come from a technological net cast very wide indeed, which is technically supposed to be illegal. And illegally-gained evidence cannot be used to get a warrant, or used in a court case.
This may be the problem that even FISA court judges are having problems with. In general, FISA is a pretty lenient court, approving almost all petitions the government brings before it. It has been called a "rubberstamp court" more than once, for this very reason. But even a FISA court judge can't rule in favor of Pandora's curiosity, because they know that once that box is open, there'll be no shutting it.
This may also be the reason why the White House is arguing for changes in the law rather than asking the FISA court for warrants for whatever it is doing. The changes in the law which are needed may have nothing to do with what is reported in the media, but in fact be more fundamental and far-reaching.
There is a bigger question than "Is it legal?" or "How can we change the law to make it legal?" which must also be addressed: "Should it be legal?" Due to the secrecy surrounding the law in Congress (national secrecy sometimes means closed sessions and secret laws), this debate may have to happen among the public, since the lawmakers may be unable to legally discuss the issue.
This argument will fall along the fault line of those who strongly believe in civil liberties (even if it makes intelligence and law enforcement work harder), and those who believe in security at all costs (even if it means their phone will occasionally be tapped by a computer).
But it's not that simple a debate, which is why I wish this debate would actually start. Because a computer is at the center of the debate, it significantly changes the outlook. The vast majority of communications handled by such a computer system would be rejected and filtered out, meaning a human being never sees them in any way.
Civil libertarians may be more comfortable with a computer tapping their lines than an NSA gnome. Security-at-all-costs people may be more comfortable with the implied efficiency of such a system, since it would free up agents to concentrate on the suspicious messages.
Such a gigantic fishing expedition is currently illegal, to the best of my knowledge (these days, it's hard to know exactly what the NSA or the White House considers "illegal" -- if anything -- when it comes to wiretapping).
But should it remain illegal? Should the FISA law be changed to permit it? For foreign-to-foreign calls? For foreign-to-America calls? For domestic American calls?
These issues need debating. Since the FISA law is up for vague and secret changes, perhaps now is the time to have this debate.
[Today's Tom Toles cartoon is worth checking out on this subject.]
Cross-posted at The Huffington Post
-- Chris Weigant
Chris,
While these are interesting questions, this is already occurring... ie... it isn't speculation.
How do we know?
We know that AT&T set up a "way-station" where all calls and data are transmitted directly to the NSA from the AT&T whistleblower.
We know that the sheer volume of these communications precludes humans from sifting each one.
Ergo... we know that the "sifting" is being done BY computers programed to "hit" on key words... and then THOSE communications are pulled for the law enforcement agent.
We also know that the REST of the communications are stored in a database (data-mining).
In essence, you have one program doing at least two activities that are, by our current laws, illegal, and was the logical next step of all the other spy programs we've seen throughout the years.
It's even worse than it seems.
Automated text- and voice-recognition systems are in their Nth generation since 1960 and NSA is always at least 2 gens ahead of commercial products. What new systems can do, based on NSF projects being granted these days, is parse all such communications for "meaning", which means throw their contents into bins based on what the author says. The fear is not that Big Brother monitors our speech and text, it's that Big Brother will allocate who you are politically, based on the content of your written and spoken word.
Once done, it is an easy matter to tie that together with voter caging, automated voting mechiness, and get control of the entire Electorate, combined with highly controlled localized and directed media, helping fine-tune the 'message of the day', etc.
I was asked on another forum to watch some PBS (or A&E, I don't recall) specials about issues such as this, especially the "Las Vegas Incident"..
For those of you unfamiliar with the case, let me briefly outline it. Federal authorities had received reliable intelligence that during a New Years Eve celebration (2004's, I think), there was going to be a terrorist attack. Federal agents moved in and started "data mining" the area. They obtained hotel records from all the hotels on their guests, looking for anyone that "fit the profile".
As it turned out, the intelligence was false. No attack was planned..
But one segment of this special dealt with this one couple who were completely indignant that the Federal Government would "violate" their privacy by obtaining their hotel records.
I simply cannot understand this for one simple and basic reason..
Let's play the scenario in reverse.. Let's say that there HAD been a terrorist attack in Las Vegas. Hundreds were killed. During the investigation to determine who was responsible, Federal agents obtain all the records of the local hotels, trying to find anyone that "fit the profile". No one could object to this. It is only good police work. It's a natural effort of the investigative process...
So, this begs the question. If people do no have a problem with this so-called "violation of privacy" AFTER the fact, *why is it a problem BEFORE the fact??* People are saying that they don't mind the so-called invasion of privacy to INVESTIGATE the brutal murder of hundreds of innocent men, women and children. But they get all indignant and scream "GESTAPO!!!" if it is done to SAVE the lives of hundreds of innocent men, women and children..
Herein lies the hypocrisy of the America Bashers. Personally, I don't have any problem whatsoever with my phone being tapped or my emails being read or my actions being monitored. And this is for two reasons. One, because having been in the business, I know the necessity of such actions, when warranted. And two, I have nothing to hide.
Now, tieing this into CW's article, I think he hit the entire issue out of the park with this statement:
"Civil libertarians may be more comfortable with a computer tapping their lines than an NSA gnome. Security-at-all-costs people may be more comfortable with the implied efficiency of such a system, since it would free up agents to concentrate on the suspicious messages."
But what CW failed to mention is what I see as the best way to go..
FISA judges will have approve **the computer program** that will mine the data. And THAT is where, I believe, the consternation lies. I see a NEW court becoming an offshoot (or replacing) the FISA court. A court that will deal specifically in computer operated data mining operations. The Judges in this court will be knowledgeable in computer software. Knowledgeable enough to be able to rule on the constitutionality of what a computer can and can't do.
Questions like this are already being asked. For the first time, armed autonomous robots are being used in Iraq, despite the famous "Three Laws"....
http://blog.wired.com/defense/2007/08/httpwwwnational.html
One of the comments after the article sums up the dilemma perfectly.
"If the robot kills an innocent civilian, who is at fault?"
Technology is joining the war on terror with a vengeance.. It is past time that questions such as the one asked about the use of armed robots be asked. Debates such as the one CW proposes MUST be done and be done soon...
If we do not, then the decisions will be made for us.. And they may not be to our liking...
Michale.....
Michale -
Very interesting comments. While we're not going to agree (as I have previously stated, I am an absolutist on the Fourth Amendment), I would like to respond to some of the points you brought up.
You logic is good, Spock would approve. But taken to its logical end, if the government is allowed to vacuum up things like hotel records (I saw the same show, it was on PBS, possibly either "Frontline" or "America at a Crossroads"), to prevent crimes from happening in the first place, then why shouldn't they collect such records all the time? There may be terrorist plots which the government is unaware of which could be stopped by such practices, after all.
Unfortunately, this is where George Orwell's name inevitably arises.
Your point on the FISA court hiring some computer-literate judges is right on the mark. When technology and law collide, it is best to get some expert opinions on the matter! In the initial draft to this story, I was going to write something like:
========================
If that's true, that would require massive changes to the legal system. Because it is then the computer who decides what could be evidence and what probably is not. If the computer is programmed to think a certain communication is fishy in any way, then it spits the message out for a human being to read. Or, more accurately, it is the computer programmer who creates the software, and whoever inputs the keywords used to search. While the programmer and the data inputter are indeed human, the program runs on its own -- so who is really responsible?
======================
I was wondering if anyone would catch that. We are so conditioned to imagining "computers" as "sentient beings" when in fact they're not. Any Turing machine (that is, every computer on the market today) just does what it is told by a human programmer. I even anthropomorphized them myself by saying "If a computer is programmed to think..." when that is completely inaccurate.
But I left the issue out, as it just added a layer of complexity that wasn't entirely necessary to the conversation I was provoking.
Your further cite of the Asimov's Three Laws of Robotics shows you understand the deeper issue all too well. The Three Laws would be absolutely impossible to actually program into a computer brain, at least as computers are manufactured today. I think Asimov even admitted this, later in his life. Of course, a "positronic brain" may be different, which leaves it an open issue....
But in any case, we are seeing in Iraq the very beginnings of robot warfare, you are right. Now, you can argue that guided missles, or even the first crude guided torpedos, were the birth of such warfare, but the new flying drones are truly the beginning of a new era in warfare. I'm not talking about human-controlled drones, where someone with a joystick flies the thing by remote control, I'm talking about planes which can fly themselves, carry out a mission, and return on their own -- with no human intervention.
We're not quite at "Terminator" level of robot warfare, but anyone who looks for milestones in technology should be aware of this development.
As for your final question, just like with Air Traffic Control computers (if two planes collide in midair as a result of a computer bug, who is responsible?) I would personally answer "the programmer and the software tester" are responsible for the deaths. Not everyone would agree with me, but that's the way I see it.
-CW
I love the meat that these types of issues allow us to knaw on..
Most of your points are, as you say, of the "agree to disagree" variety.. I don't think that the 4th Amendment is a suicide pact. If the government has to over step some bounds to save lives, well.. In that case, the end surely justifies the means..
However, with your comments on "robot warfare".. I think we are a lot closer to "Terminator" style warfare than you want to admit.. Missiles and drones are "robots" of a sort, but they are simply instruments of a *human's* decision to kill..
In the case of the Iraq robots I posted, they are simply sent out on patrol and the actual decision to kill (albeit a programmed 'if A happens and you are B, then you will do C' decision) is made by the robot itself.. At least, that is my understanding of the underlying principles...
And I also agree with you regarding the "Three Laws".. The fiction that is "Three Laws Safe" (as popularized in the recent Will Smith hit, I ROBOT) is simply not possible with todays technology..
But, I do believe that such a concept in AI is worth striving for.. Especially in light of the idea that we may very well see an I, ROBOT or BICENTENNIAL MAN (Robin Williams) type society in our lifetime...
Michale.....
Since we have approached this issue with the idea of "Movies" in mind and given CW's statement:
"....then why shouldn't they collect such records all the time? There may be terrorist plots which the government is unaware of which could be stopped by such practices, after all."
..... I decided to throw into the mix for discussion the moral, ethical and legal ramifications put forth in the movie, THE MINORITY REPORT...
For those who are not into movies as much as others, TMR is a sci-fi style movie dealing with a futuristic police force (Pre-Crime) that can view the future and uses such viewings to apprehend murderers before they can commit the murder.
As I write this, the parallels between that movie and the real life Las Vegas incident are becoming more compelling and relevant.
Towhit, would we want authorities to use such an obvious and gross invasion of privacy to to save innocent lives?? What if the lives saved were you or your family??
Michale.....