Sunday, June 17, 2012

The Eliott Effect


One of the major problem I have in discussing the nature of the ongoing crisis and the fate of the industrial society is dispelling the myth of the "immaterial economy". It is, I must say, particularly widespread among French elites, perhaps because, as a nation, we tend to despise manual work. Not so long ago we had a debate in the Municipal Council about the digital economy in Nantes and only my boss raised the tantalium supply issue (as I said, he is my boss), as for our resident Greens, they push for the replacement of paper by computers and tablets.

The idea is that since work done and transmitted by a computer is not tangible, it is entirely the product of the human brain and therefore free. Digital economy could therefore fuel a limitless growth even in an era of widespread energy and raw material shortages.

Nothing could be further removed from the truth, as I was taught in a rather dramatic way.

I happen to be owned by a four-legged black-furred monster called Eliott who enjoys stealing chicken escalopes and leaping on tables at the most inconvenient moment. So that day, I was reading my mails while enjoying an excellent Iranian tea, when the little monster decided it was the right time to run across the table, toppling the mug in the process. The tea poured down on my laptop, which proved to have little taste for even light flavored Iranian blends. It fizzled then went dark, causing the collapse of my personal section of the global network and the loss of a considerable number of digital book, including a highly valuable Chukchi grammar.

To make things worse, I was barely recovering from a nasty and costly break up and the state of my finances was far from optimum, so to replace my computer I had to have an interesting, if somewhat tiring, talk with my banker – I needed a laptop for my political activities so doing without was not an option.

The first lesson of this story is that having a mug of tea too close to your computer when your cat is chasing an imaginary mouse is a bad idea. The second lesson is that the so-called immaterial economy is highly dependent upon very material devices and infrastructures and is very likely to crumble when those devices and infrastructures can no longer be built or maintained.

It is possible to build a computer with XVIIIth century technology, even if it will be costly. The concept was first described in 1795 by J. H. Müller, an engineer in the Hessian army and in 1822 Charles Babbage tried to build a mechanical computer – the difference engine - on behalf of the British government. The stated goal was to produce cheap tables, a time consuming and expensive job then. Unfortunately, the standards were so exacting that Babbage ended up spending twice the price of a ship of the line without producing a working prototype and the government finally killed the project.

Babbage then moved on to another, more complex, project : the analytical engine. The analytical engine, albeit entirely mechanical, was more advanced than the computers of the late forties. It had a “memory” of 16,7 kb and used a programming language akin to assembly language, which allowed for loops and conditional branching. The analytical engine was never built but recent experiments have shown that it could have been with the technology of the time.

The idea may have been even older. At the beginning of the twentieth century, a kind of small-sized mechanical computer, was discovered in the wreck of of a Greek ship from the fourth century BC. It was apparently designed to make astronomical calculations. It seems not to have been the only one of its kind.

Mechanical computers of various kind have been used up to the sixties, mostly for research, military or navigation purposes. Thus, the physicist Henry Fermi used a mechanical computer to model neutron transport and the economist William Phillips designed an hydraulic computer (the Moniac computer) to simulate a national economy.

Mechanical or hydraulic computers have their limitation and I seriously doubt you could play Europa Universalis III on one of them or store your collection of paleo-asiatic grammars in their memory. It is quite possible, however, to perform complex computations with them and that beats doing them by hand.

Mechanical computers are definitely one of those technology we should get through the coming Dark Age. They will be of great use to an ecotechnic society and relatively easy to manufacture in a resource-poor world.

As for build a world-wide-web with them, however... well, it could be possible to build a continent spanning network of canals carrying information for a collection of hydraulic computers, but I feel it would be somehow impractical on a planet where water tends to pour down the sky with very inconvenient regularity.

Digital computer are far more convenient in that matter. The problem is that they require a lot of rare material to be build. Silicium is arguably very common but to purify it to the level it can be seeded with germanium, you need a relatively high technology and a lot of resources. Should you manage to preserve that, you will still need tantalium for condensators and rare earth metals for screens and hard disks, and of course a lot of copper for an awful lot of wires.

To make things worse, your networks have to be maintained and your computer powered. Unlike, for instance, a hydraulic computer, digital computers need continuous inflow of high grade energy to remain useful. In a world where brownouts, then blackouts will become more and more common, this will make their use quite problematic and even outright impossible in some areas.

As the core limits, then withdraws, power availability in peripheral regions to preserve itself, computer use is bound to decline and the digital economy to retreats, until it becomes restricted to a few industrial and political centers. The rest of us will have to do with pens and paper and get the boardgame version of Europa Universalis out of its shelf yes, it was a boardgame at the beginning.)

Besides, neither satellites nor cables are eternal. They are bound to break down at some point and will need to be replaced or repaired. This will become increasingly difficult as the pool of resource and the energy supply available to the society shrinks. When the Soviet Union collapsed, for instance, the Soviet equivalent of the Global Positioning System, Glonass, was neglected by a government, which struggled to pay for its own survival. It quickly fell in disrepair and in 2001 only 6 satellites out of 24 were still in working order. The system has since recovered, has Russia has done, but, but the fate of Glonass shows what happens to complex infrastructure once the society, which has built it can no longer maintain it.

You can expect a similar evolution with the internet. First, the network will become unable to keep up with the rising demand, and price measures will be implemented to limit said demand – there already has been attempts to do so, but governments, afraid of the impact upon the public opinion, have managed to thwart them, so far.

Then, as our ability to maintain our infrastructures declines, accidents will happen and the Eliott effect will fragment the Internet, one cable at the time. My opinion is that the transoceanic cables will be the first to go, isolating whole regions and forcing them to go off-line indefinitely or to organize their own continental version of the Internet; Those localized Internets will then fragment further, due to economic or political crisis, war or accident, losing usefulness at every step of the process. At the end, the Internet will become a collection of unconnected regional networks, with little added value compared to a standard library, and will morph into a mere administrative tool for whatever remains of the local governments.

At some points, most such network fill fall in disrepair and be terminated, as the Minitel (a French precursor of the Internet) will be in a few days from now. Others may linger on, in particularly stable and rich areas until the last digital computers die. Those networks will, however, very different from those we are accustomed to, text-rich rather than pictures rich, and you will be quite unlikely to find your favorite movie on them – yes, even that kind of favorite movie.

For my French readers, it will be reminiscent of the Minitel (and yes, it may include that kind of service) and of the text-based network featured in Avalon. Needless to say, such shrunken networks won't provide the “immaterial” economy with the market it needs to thrive. It is bound to decline along with the Internet, probably specializing in niche activities.

By the way, the Dark Net won't fare any better. The Tor Network, Freenet or the parallel networks used by mafias may be secretive , they rely upon the same highly material infrastructures as the the rest of the web and will fragment as they do, which will incidentally make them far easier to monitor and control.

Then the local equivalent of Eliott will spot a mouse on the other side of the table and begin to leap around... and the rest of us will have to learn to compute with cranks, valves and pumps.