 |
||||||||||
 |
 |
 |
 |
 |
 |
 |
||||
|
||||||||||
|
|
|
|
|
|
|
||||
|
|
|
Let us try and understand the future by looking at history, shall we ? I know, I
know, we are so digitally advanced we don't need history. Nor do we need thinking,
come to think about it. What we need are more schmucks a la Robert Paterson to give
us more regurgitated crap about swarm intelligence.
As an aside, let me point out to all of the dumbass bloggers, and second generation "thinkers" contributing their own retard drool-bubbles to the problems of computers and networks, a bit of history, what else ? Back before all of you found yourself all "digital" all of a sudden, there was a horrid month of September. Before then, every September, stupid college kids would get email accounts at their respective colleges, and then a deluge of idiocy swamped the Internet (yes, the Internet existed before your newfound digital skin. Fancy that). The elders took a couple of months to hammer some semblance of responsibility and etiquette into the unwashed hordes your smelly fat wives produce with nauseating regularity, every 16 months after your signing of yet another mortgage agreement for yet more picket fence. It was easy at first. Only the brighter kids would care about this Internet thing, and here's the thing with intelligence : Not only is it easier to educate intelligent people, but there's fewer of them. So, the effort required to civilise any group varies with the square of the group's median intelligence. There's some scatter effects we won't get in, let's pretend the group is pretty homogeneous, with tiny deviations from that median. Then it got harder and harder to do. And eventually, it got impossible. That September never ended, and it was the birthdate of a new Internet. Not quite the crap you know now today, slightly better, but still. However, the Internet had existed before that, and for a while. And people irreducibly better than you spent years, decades, trying to figure this thing out. Before we get into that, let me explain what I mean when I say better than you. I mean simply this, you would never, no matter what you tried, or for how long, ever become something that could be described in the same system they are. If they were an oil tanker, you wouldn't be a smaller oil tanker. You wouldn't be a gas canister. You would be a paper boat. If they were a piece of paper, you wouldn't be a smaller piece. You wouldn't be a moldy scrap of toilet paper firmly ensconced on the perianal hairs of a burrito loving truck driver by the name of Earl. You would be a line. Are we clear ? Good. Now, these people spent years of their lives writing meaty articles about intelligence, artificial, programmable, expert or otherwise. It never went anywhere. Read that again. It never went anywhere. You thought I was just being inappropriate and crude above, did you. You know, moldy scrap of TP et all. Sadly, I was just being plain. Here's the thing : Minsky writes Pr(V) Pr(Fj|V) = Pr(Fj) Pr(V|Fj). You however start your computers, run various scripts (not that you know about it), to read other blogs, to add them to your own blogroll, to comment on them and trackback and everything. Gets you about the same place, in the end. With the notable exception that you have no idea where you are, or where it's going from there. Then again, you didn't even know your name all that well to begin with, so what's the complaint. Ever seen that instrument they show kids in school to teach them about probability distribution, a large rack with balls coming out from atop, like this :  Now, kids are smart enough to understand the math, or dumb enough not to. But their parents seem willing to curl up and fall from the top. Be the ball! Wheee. And then they call this swarm intelligence. I suppose, in a sense, it is. Much like what comes out of dung is aroma, bullshit has it's particular flavour, the end result of the political process is correctness and so forth. To summarize the facts, if the professors fail to understand the situation by looking at those balls, you will not succeed to understand anything about it by BEING the balls. You may understand something about the difference between steel balls and meatballs, but chances are you won't. And under no circumstances whatsoever will your pathetic attempt be considered "intelligent", at least by anyone who isn't you. And now, let's look at our historical exhibits. Exhibit A, search engines. Period 1. Search engines had little importance for the people playing on the Internet. They happily created documents that linked to other documents, according to contextual criteria (ie, something about the people, the documents, or their respective relationships) Period 2. A couple of grad students notice that the above contextual linking can be used to structure the entire Internet, and proceed to do so. Period 3. The structure thus created becomes so important that people create less and less links according to contextual criteria, and more and more according to meta-criteria (ie, something about what the effect on the structure is likely to be) This, in and by itself, is the death of Google. It is an engine that transforms contextual links into value. As a result of it's working, part of the fuel is transformed into useless exhaust fumes. When the concentration of fuel to exhaust reaches a certain level, the engine stops. Which is why we've already sold the Google stock. Exhibit B, wikis. Period 1. The wiki has little importance to anyone outside it's subject matter. The importance of first hand uses, such as reading articles, writing articles and editing articles is paramount. Period 2. As more and more people find out about the wiki, and it's importance grows, more and more users are interested in meta-uses, and the relative importance of first hand uses decreases. Things like edit wars, political clout within the editorial community, numbers of edits per author, editors tops and halls of fame become more important than the original, intended uses. This, in and by itself, is the death of the wiki. A wiki is fundamentally no different than a link farm, in it's structure. What could make it different is human contribution. It would seem that the more important the wiki is, the less first-degree user contribution, and the more meta-contribution it receives. Exhibit C, user-voted link lists. Period 1. The link list is not important to the Internet, so it receives virtually no spam, and and it's not important to it's contributors, so nobody can be bothered to make hundreds of sock puppet accounts and vote his own submissions up. Period 2. As the importance of the list grows, so grows the time and effort available to spam it, and the interest to do so. This will sooner or later kill the list, because really, who cares ? Exhibit D, blogs. Period 1. (To make it perfectly clear, this was before 1998, and you're not one of the cool kids.). People had webpages they put up for a variety of reasons, with a variety of content. Some were interesting, and other people read them. Period 2. People have blogs. They put up a variety of content, that is not really all that varied, just like any polynomial is not really all that different from all the other polynomials. The majority of the readership is other bloggers and robots. In a desperate attempt to increase "traffic", things like rss/atom/etc feeds are invented. These have the advantage that they leverage user interest into immense traffic numbers. For instance, one single inadvertent click on some rss feed reader, coming from one single guy in Turkey looking for gay porn is bound to send requests to hundreds of blogs, and lo and behold, there've been 500 "visitors" on 200 blogger's pages. Ho-hum. This sort of idiotic self-scripted traffic is growing, because most people don't have the apparatus to distinguish between it and actual interest, and even some who do don't want to. It would be obviously easier, cheaper and faster to just edit the blasted database, and store a hit number that is ego-adequate. But hey, why have sex when you can date. Blogs in period 2 are essentially indistinguishable from linkswarm and other cooperative schemes where a few people get together, click on each other's links for the purpose of having clicked on links, and then watch in amazement how indeed, their website got traffic! omg! The moral of all this is simple. As importance grows, the user base goes from intended usage to meta-usage, and sense shrinks. This would be what I call the fundamental corruption. I really like it, because it has a few remarkable qualities. Firstly, it's unavoidable. Anyone with at least some systems theory understanding can easily see why. Anyone else, maybe I'll explain it some time. Or maybe someone else will. In the long run, you're dead anyway. Secondly, it's highly ironic. The conversion from intended to meta-use, the increase in prevalence of spam and bots, that is, of users that are not human, and of users that, while biologically human, still won't pass a Turing test, it all has something of the solid beauty of a patch of weeds. You can almost see the remains of a gravel path, there's two solid poles remaining that were a barn, under all the mail.ru wild roses there's the vague foundation of what maybe was the house. That round thing might have been the well. But the Earth has reclaimed it, and just so, the Internet, which is a machine, made out of machines, is reclaiming it. It does not understand sense, and then sense has no place on it, other than temporarily. The first phase in our war against our machines has started. Amusingly, all the science fiction got it wrong. They aren't mean, and they aren't purposeful. They aren't out to get us. Asimov's rules would be useless. They will just gladly give us enough rope to hang ourselves, and they don't even know it. Swarm intelligence. 11 people care. |