jump to navigation

Dropped Frames August 4, 2007

Posted by malthusio in Uncategorized.

What happens when an irresistable force meets an unmovable object? Implosion, explosion, or possibly some other kind of ‘plosion perpendicular to either two (perpenplosion?). The problem with the accelleration of change is that it vastly exceeds our capacity to cope with it. Individual human capability is (currently?) linear whereas collective human capability is exponential, due to positive feedback loops present between human-human and human-technological relations.

The reason I say that individual human capability is linear becomes obvious when we consider what happens at potential branch points. As an individual, you can only do one thing, so you have to choose what branch of activity to follow. However for the human collective, different branches of a task can simply be forked off to different individuals (typicaly refered to as “co-operation”). While mob mentality is typicaly not known for enhanced capability, this is not what is at work when I speak of human collective operation here.

The problem with the mob is that it is a tightly coupled system, and where positive feedback pathways exist, they will become the dominant force. So the mob is simply a harmonic resonance that reverberates back through all of its members. Again, this is simply becuase the coupling of the mob is too tight. Unity is not actualy a good thing, becuase the propagation delays across the network are large, only the most common messages are fully propagated, because the frame buffers of individual nodes are overloaded, and you get dropped frames. This is what produces the positive feedback squeal that is the mob mentality.

 So as the rate of change accelerates, and individual human capability to absorb change stays the same, we get dropped frames, and so bizare as it sounds, the acceleration of change threatens to reduce us to a single simple harmonic resonance. In this way, modern information society is like a massive DOS attack against humanity. Under McLuhan’s terms, this shock causes us to amputate and extend ourselves to remove the source of the irritation, and replace the lost capability. McLuhan has also gone so far as to suggest that we have already amputated the entire cereberal cortex and produced its prosthetic extension outside of ourselves. Metaphoricaly, perhaps, but I don’t know that I put as much faith in this human creation as McLuhan seems to. The interface between man and the extended cereberal cortex of media is simply to low bandwidth, which makes it highly vulnerable to DOS attacks. The part of it that operates more or less independantly from individual humans is maybe less subject to this limitation, but I don’t think that open source mediational technologies existed when McLuhan wrote “Understanding Media: The extensions of man”.

Anyways, I’m probably getting off topic here. The point is that individual humans can only absorb so much information, and so much difference at once. Anything more, and you start getting dropped frames. These frames can be thought of in both conceptual and narative terms. If you drop enough frames of film, you compromise the narative unity of the story. When a router, switch of frame relay drops a frame, a message isn’t communicated. In both cases, the result is a cognitive disconnect with the environment. Either it isn’t understood, or the message simply doesn’t register.

The internet is interesting because it both excacerbates the problem and provides a solution at the same time. The combinatoric proliferation of information that the internet allows for through fine grain massively scaled collaboration (look at wikipedia or sourceforge for good examples.) amplifies the accereration of change (not sure what the amplification factor is, or where it’s applied, but it seems likely that it is either an additive or multiplicative factor of the base of the exponential). This of course produces more information at a faster rate than humans can possibly deal with it. However, at the same time, I think that humans are using the internet as a frame buffer, like an external hipocampus for the integration of long term memories stored outside of ourselves (And it’s possible that this is what McLuhan was talking about in my reference to him above). Asynchronous comunication provides a frame buffer, and if human cognitive capacities are being stretched in the way that I think they are, then this only makes sense. The alternatives are a kind of solopsism or mind death (as you become reduced to a single resonant frequency).

In his novel “Accelerando”, Charles Stross developes an idea he calls the “exocortex” which is a swarm of distributed (artificialy intelligent) agents which the user interacts with through a wearable apparatus (which isn’t actualy all that far off. voice interfaces still need some work, as well as some of the software, but mostly, no one has realy bothered to put something like this together. Steve Mann  has probably gone the farthest with wearables, but from what I understand, his are largely passive. Also check this out. Not sure if he’s behind it, but he has a glog there). Personaly, I think that people are using the internet something like an exocortex. With a simple google search you can recall things from the long term memory and collective unconciousness embodied by the internet. The problem is that there is still plenty of room for dropped frames across the low bandwidth interface of user and machine.

For myself, I find that I think and forget far more thoughts than I ever get written down. By choosing to write one, I’m forced to drop the other, usualy on the floor. When you get tired, the problem is worse. Never mind writing a coherent thought, it’s possible to drop so many frames that no thought has sufficient narative unity to be understood…. which is odd because you generated them….

Anyways, there are two solutions to this problem. Either we find some way to increase the bandwidth of interfaces/channels that we drop frames across, or we can simply decrease the bandwidth requirements. Personaly, I’m in favor of pursuing solution 2 over solution 1 (although these are not mutualy exclusive, so there is no reason we can’t do both so long as we are able), since will provide a solution that will be even more relevant if we were to increase the bandwitdth of these bottlneck interfaces. Simply throwing bandwidth and computational power at a problem is no subtitute for a good algorithm. In fact, believe it or not, having good algorithms (and design in general… simply focusing on algoritms is too linear an approach for my tastes) becomes even more important as computational capacity, bandwidth, memory etc increase.

Intuitively it seems like the opposite should be true, since you can brute force more and more. However, this approach runs hard up against the law of diminishing returns. I suppose what I’m basicaly arguing for is quality over quantity. I’m not saying we can’t have both, but where we have a choice… quality should be prefered. Otherwise, we start dropping frames in unexpected places. Like the environment for example… Did you know that most landfill sites are filled mostly with dropped frames? Or maybe I’m just getting carried away with the metaphor. Even I’m starting to drop frames in the connection to the analogy…



No comments yet — be the first.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: