Monday, June 28, 2010

More on granularity

I keep on coming back to these couple of broad, abstract processes that I see referenced in all different kinds of news stories and events.

One is about differing beliefs on lenses, magnification levels, whether to take the 50,000 foot view or the ground level view, Stewart Brand's long now, or a Human Rights Watch reporter's yesterday, today and tomorrow. Or a smart synthesis, but you end up choosing one or the other as your basis for decisions when you have to make a decision.

The other broad process has to do with granularity and things that didn't used-to-be possible. Situations where old workarounds are no longer necessary, therefore things that used to be subsidized as a side effect of necessities of transmission are no longer subsidized. Matt Welch talks to Alex S. Jones on Bloggingheads and says, in healthier times for the newspaper business, people were buying the package. Probably in homes in the suburbs, they were subscribing. And the word "granularity" went through my mind. They bought it for the crossword or the comics, and they got the hard news, because taking it apart wasn't an option. That just wasn't how things were done. Today, I can spend all day on just ESPN or just Scott McCloud's website or the Comics Journal or Garry Trudeau's Doonesbury website on Slate. Okay, Slate runs all kinds of links all over the place, til the design looks like a freaking pinball machine - so maybe I get exposed to stories by accident just like I do with a paper paper when I go for the 3/4-page Macy's ad of the lady in the bra and panties, and stay for the Planning Commission story.

It's one example of the phenomenon that is brought about by increased capacity and the overcoming of speed bottlenecks and the reduced cost. Skype and Webex, remote access products, VOIP phone systems empower global workforces - so the labor standards of any one nation have a harder time. Employers, businesspeople used to get employees from around town because there was no other option - maybe even domestic long-distance calling was a big deal before cel plans.

Both of these things have ramifications for my main preoccupations/themes on this blog. The former, fundamentally, is the conflict of grand project versus casualties incurred in the course of that project. Can't make an omelette without breaking a few eggs, we've contextualized it as "an omelette," so a few elbows in the ribs are par for the course, now that we've defined what the course is. The difference between grand project focus and casualty focus is a similar question to where to put your scale/order of magnitude in assessing most any metric of the human condition over time.

The latter is maybe more about the widening palette of tools, ways and means, with which to do surveillance. You could even say that a tighter focus leads to higher-quality surveillance and fewer accidental false positives. Fine - this is where I am partially a civil efficacian. Show me some numbers. But you have to look at all aspects of costs and benefits. Like for instance, the unmanned drone strikes. They're still causing Afghan casualties, but enable the war to continue without PR disasters (at least this particular stripe of PR disaster) because there obviously will be no American soldiers killed in this particular avenue of the operations. I don't know. My bias is not to be over there at all. But take the drone strikes against a similar military operation in the 70s or 80s, and maybe it is "safer" as a result of technology, all other things being equal.

Another example would be big, honking databases with big honking SQL queries. I resent Acxiom and the obnoxious things they do. However, do faster computers lead to better TP/FP ratios over time? I had really better get empirical in this project. Anyone who is reading this, thank you for your indulgence, you can check out if you wish, because it is not polished, it's more like immediate notes as I think of them, to perhaps be arranged into something coherent at some point in the future. I suppose I take these issues about technology enabling Bill of Rights violations in a layered fashion. There is the Bill of Rights absolutist, and I like that a lot. But the center moved right after 2001. I'm pissed off, but would like to stay engaged anyhow. So you say, fine, the absolutist position isn't working because we elect people who don't take it. The next layer is the civil efficacian. If you're going to do it, for gods' sakes, do it well. This is where I am perhaps agnostic on how technology will influence false positives, mistaken identity stories and so on. My friend Scott Tipton once tipped me off to the fact that Charles Schulz had invented a minor Peanuts character in the seventies named "Five." I really liked this, and was a little disappointed to discover that it wasn't just absurdist, playing with words, just busting out with the Red Baron and root beer or the cat next door turning Snoopy's doghouse into shreds of wood, but a sort of classical conservative commentary and satire of chagrin, upon oldschool computerization, like with whirring great reels of 3M tape. The kid was so immersed in bureaucracy and depersonalized service, his name was Five. The point of this Schulz story is that the incorporation of computers in those days, maybe would have been seen as making things worse. I guess that has always been the case - cue one of my favorite words, "boffins." Boffins in white lab coats had it all figured out. Maybe that has changed, I'm not sure, but it seems like it has in, for instance, the consistent role of Google in providing semi-amazing, ubiquitous informational wonders to most everyone most every day. (Or Bing. Whatever. I don't care, I am vendor-agnostic.) They're doing it really well, like a killer app, or the adjective "winning" as in "a winning smile."

Okay, and I say that with chagrin as well. You have Siva V. (fill in his name properly here), talking about Street View, bands of citizens surrounding the big trucks so they couldn't shoot. And the thing with the wireless interception. They do seem a little out of control.

So maybe the efficacian in me is more angered and worried by technology to a degree that outweighs the enjoyment of all of the ubiquitous informational wonders on a daily basis.

(Side note: Let's also don't forget black swans. I'm a big Bob Wright fan, but I think we are getting more and more giant burst events which *eventually* screw up one's thesis about a net positive change over time, just in a spring-loaded fashion - bottomheavy if you will. It looks like nonzero relationships until the 11th hour, at which time a bomb goes off that not only fucks everything up, but retroactively changes the picture from the times when you thought things were good. Like waiting for the other shoe gonna drop.)

So does Street View make it worse or just different? Will they find exciting new ways to eviscerate the 4th Amendment when there are nanocomputers with atom-sized cameras and wings (that didn't used to be cost effective or technologically possible) that get into the airstream and we swallow them, and the ones that we don't swallow are shooting our houses? (Miniaturization seems important to my points here - like granularization taken to mean tinier atoms of information, more degrees of freedom in computer modeling, but also literal granules.)

Or - will the new cost-effectivity of the technologies mean they can afford to, let's say, use 1:1, census-style checking, Total Information Awareness, rather than statistical sampling in cases where sampling used to be used because of cost- and will that 1:1 mean actually fewer accidental false positives, unmonitored intrusion on a truly quantum scale, but only interfering with the lives of the true positives?

Even if the accidental false positive rates go down, you still have deliberate false positives to worry about. So for reasons of either AFP (if you're not satisfied with the rate - even one accidental dead person is too many) or DFP, there is a third layer. Beyond and below Bill of Rights absolutist is civil efficacian. Beyond and below civil efficacian, is encryptor. Screw them, says the encryptor, maybe their rates will be .00001 but I am not going to become a statistic. If I have something private to say, I will say it privately.

Beyond that, all you have is man-in-the-middle, and what else? I'm sure there is more - my knowledge of encryption isn't that great. Long post.

No comments:

Post a Comment