Sousveillance redux.
This essentially humorous column by Matt Smith has been taken quasi-seriously by John Gilmore and is kicking up a ruckus among the usual suspects. —Myself, I think one participant nailed it when he pointed out that the profound discomfort people are feeling at the idea of posting Poindexter’s personal information is precisely the point.
In related news, Salon has managed to interview a number of computer scientists who, facing the prospect of fat DARPA contracts, nonetheless manage to see some merit in the idea of a Total Information Awareness program. (It’s premium content, so if you don’t have a subscription, click through the little Mercedes commercial for your free daily pass.)
Jeffrey “Frankly, I don’t see any other way for us to survive as a civilization” Ullman, a database expert from Stanford, wrote a rambling piece on Islamic fascists and fundamentalists and warring on terror in the days after 911, including the difference between terrorism and state-sanctioned warfare (terrorists can’t parade their weaponry, a la missiles trooped up and down Red Square back in the day) and a story about the time Osama bin Laden’s nephew dissed his nephew at some toney Eastern college. Salon nipped this quote as a rallying cry of the pro-TIA faction:
Modern technology has given criminals and terrorists many new and deadly options. Just about the only defensive weapon to come out of the developments of the past 50 years is information technology: our ability to learn electronically what evils are being planned. If we use it wisely, we can keep our personal freedom, yet use information effectively against its enemies.
Sounds breathtaking, doesn’t it. We can learn what evils are being planned.
Well, no. We can create massive databases of seemingly trivial information and use it to search for patterns and act or not act on what we find. But how do we know what patterns presage evil? How do we differentiate them from ordinary, everyday activities that fit the (ominously unspecified) pattern? How do we deal with the innocent lives that will be disrupted and possibly ruined by false positives? —There will be false positives. To quote some sobering numbers from Bobby Gladd, a statistician who’s kicked up a ruckus about false positives in the War on Drugs: a pipe-dream TIA that’s 99.9% accurate would still finger 240,000 innocent people. Surely a little disruption in our everyday lives is worth preventing another tragedy, supporters will say; this logic would also have us ban automobiles. More to the point: think of the waste of time and effort on the part of the good folks at the Department of Homeland Security, running down false positives spat out by a clunky, unwieldy database running search algorithms we’re still in the process of, you know. Tweaking.
And hell: how do we define “evil” in the first place? Wave the hand of Potter Stewart over the whole mess and merely know it when we see it? You might want to talk to Greens, nuns and peace activists who’ve tried to catch a plane the past few months before you blithely sign off on someone else’s definition of what it is exactly we’re looking for and trying to stop.
It would be lovely, wouldn’t it? A system that could scan all this trivia and unfailingly find these patterns and bring them to our attention, protecting us from the Bad Stuff before it happens. But putting this system, with its insanely broad sweep, under the control of a secretive branch of the government with a nakedly partisan agenda and the ability to re-write the definition of the thing it’s looking for—
It’s not just un-democratic, un-American, un-free and irresponsible. It’s staggeringly stupid.
(We could, I suppose, call Poindexter, and ask him what he thinks about the abyss gazing back…)
Commenting is closed for this article.
First of all, my first name is Jeff, not James.
Second, the 0.1% false drops figure that Mr. Gladd quoted in the Salon.com article is simply a made-up number. We don't know what the false-drop rate has to be --- that's a matter for research to discover.
Incidentally, although I have no idea who you people are, I found this article, with NO FALSE DROPS, simply by quoting one sentence from my article to Google and asking to see all documents that had quoted the sentence (there were only 3). You might imagine that in all the Web someone would use the same sentence in another context, but that is not the case. Moreover, the technique seems fairly robust. I periodically use it to check who is making copies of my tutorials that I put on the Web, and I rarely get a false drop.