Viridian Note 00237: Digital Decay

Bruce Sterling []

Key concepts
embrace decay principle, variable media, Guggenheim Museum

Attention Conservation Notice: half-hour Papal-Imperial rant delivered in an auditorium designed by Frank Lloyd Wright.

Link:,1284,42493,00.html "Restoring the Sistine Website"

Bruce Sterling

"Preserving the Immaterial" Conference Guggenheim Museum, New York City March 30, 02001

"Digital Decay"

You may wonder why a science fiction writer like myself takes such obvious delight in hanging out with museum curators. It's a paradox, like the one we heard in the name of this conference, "Preserving the Immaterial." If it's immaterial, why would it need preserving? And if you're a futurist, then how come you're in a museum?

But there's no contradiction here; it makes perfect sense if you look at right. Futurists and antiquarians both work with the nature of time. I have a passionate allegiance to my esteemed colleagues in museums. Because the future is just a kind of past that hasn't happened yet. And obsolescence is innovation in reverse.

Curators, conservators and archivists are much closer to the future than most of us mortals. That's because they store, catalog and preserve == they physically touch == the objects of the past and present that people in the future will see.

When you hang out with conservators and archivists, then you get to see what the passage of time really does. These folks have a job of work. The processes of decay may be a little hard to spot during a single human lifetime, but in a museum, you can find out that Entropy commands a mighty legion of ruin.

We've got all the usual awe-inspiring, mythical threats: fire, flood, storm, earthquake, frenzied mobs, carpet-bombings, plagues, and swarms of locusts. Egyptian rains of frogs. Asteroid impacts.

But those are the easy ones; nothing going on there that a little reinforced concrete and a swarm of well- armed guards can't control. The stuff that really gets to objects, the stuff that takes the worst toll even of cherished museum pieces == these are the very same qualities that make human life so delightful. Not the rare catastrophes, but the everyday things, the completely persistent things.

Things like sunlight. Sunlight is a blessed, glorious thing. Sunlight is also a powerful blast of radiation from a vast nuclear inferno that happens to be in our sky. I have solar panels on my roof; when the sun is high, I can run my whole house from the power in sunlight. Light may be considered all angelic, clear, and immaterial, so light is supposedly the very opposite of dense heavy matter, but light and matter are aspects of the same thing; Einstein proved that. You can try it at home: put a newspaper in the sun. In a few days, a few weeks, a few months, all that inky, analog data in that newspaper gets blasted with these hot little packets of solar energy, and the paper curls up, and turns brown, and it flakes, and finally, it breaks up, and breaks down.

Then there's water. Water's everywhere, it's in here with us right now, as moisture in the air. Water is the very stuff of life == and also the stuff of decay. Humidity itself is rather dangerous to objects, but everyday changes in humidity are ruinous. If humidity changes, then materials swell up and shrink repeatedly, until paint falls off its canvas, ink flakes off its parchment, wooden handles fall off metal tools.

Microbes love water as much as we humans do. So anything that is moist, and remotely organic, will be slowly eaten by airborne legions of mold, bacteria, and fungi.

Without any water, though, stuff mummifies horribly; then we get stiffness, brittleness, permanent shrinking, splitting, cracking, and flaking.

Consider other major benefits to life, such as food. A terrible thing in museums, food. Airborne greasy droplets off stoves, moldy scraps and crumbs, spilled Coca Cola. This is hearty nutrition for every variety of destructive vermin: silverfish, mice, carpet beetles. Those daring little creatures are the Earth's card- carrying agents of organic recycling. They must be fought constantly.

Tell me you're not smoking around a museum piece. Smoke puts a yellow lacquer on your teeth, on your lungs, and on everything in the vicinity. And even if you stop smoking, that doesn't mean that cars stop, or that factories stop; people in cities live in a sea of smoky pollution that strips stone statues; it corrodes solid bronze.

Humanity itself is gloriously physical. That means grease from our fingerprints, damp from our exhalations, sneezing, coughing, skin flakes, bacteria, steamy body heat from big eager crowds of us, all queuing up to admire that timeless masterpiece. Because if nobody sees a museum piece, what's the point of having it or keeping it? Museums exist for a social purpose, for us humans. Trees and clouds don't need museums.

The final painful paradox lies in harming what we save, as we try to save it. Preservation is itself a source of hazard. We dropped the precious china while we were dusting it. We tripped, and split the old painting frame. We tried to fix that old book with tape and rubber cement.

Entropy requires no maintenance. Entropy has its own poetry: it's all about delamination, disintegration, deterioriation, degeneration, decomposition, and doddering decline.

But thanks to fantastic breakthroughs in modern technology, we've got a cure for all that: digitalization! Flawless computer memories! Lightning- fast chips! Fat fiber-optics! Massive storage facilities! Bits not atoms! It's immaterial, so it needs no preserving; it's escaped from the python coils of history; time harms it no more; it's up there at the pearly gates, spotless and radiant, right next to Saint Peter.

You're probably guessing by now that I'm about to wax all cynical about this deeply flawed concept, and yes, indeed I am. But before I do that, I should give the digital its proper due.

I happen to be quite the devotee of digital technology; use it all the time. Wrote this speech with it, even. Before you heard these words, they were in my computer as digital data. They're still sitting there, probably, unless my house has burned down. What have we got in that computer down there in Texas? We've got a long string of ones and zeros.

This situation has great virtues. First of all, ones and zeros are extremely clear. No blurriness, no ambiguity, no proofreading problems. There's just two possible symbols here, a charge or no charge, a pit or a bump, some light or some darkness, they're very crisp and distinct.

Next, it's universal, because binary notation conquered the computer world. We don't have rival computer systems that use ones, zeros, and twos, and also some threes. It's ones and zeros for one and all, here there and everywhere.

Then there's error checking. A tremendous boon. You add up all those ones and zeros. Is it the right number? Then you must have an accurate copy. Bookkeeping doesn't add up? Your copy is bogus!

Then there's shipping via Internet. What a joy! Stream of photons, stream of electrons, all ones and zeros. Arrives like lightning anywhere on the planet.

And better yet, copies are practically free. Want ten of 'em, you just write down ten Internet addresses! Whip 'em on out there, new copies appear all over the planet. Copies come so easily, so cheaply, that it hardly seems like work; no one has to feel responsible; everybody always cheerfully assumes that somebody else kept a copy.

And finally there's storage. Storage too cheap to meter. We're putting vast libraries on a magnetic strips or optical disks, there's plenty of room on the bottom, we just cram in that info, we're packing the phone booth with the entire University here, let's just keep on shrinking, it's the way forward.

That is a marvelous set of virtues. No wonder people were impressed. Most of these virtues are attributable to the fact that, by historical standards, digital media have very little materiality.

However. Very little materiality, is very, very far from no materiality at all. Total immateriality is a metaphysical illusion, it has nothing to do with physics or engineering. It's exhilirating to watch these heaps of data vanishing into microscopic scales, and if it's doubling every eighteen months == hey, everything in computerland wants to double every eighteen months == then

it looks like it's going to totally vaporize, just any
second now. But it never does. Never. Even vapor is a
material. Mass and energy are conserved in an

Einsteinian universe, so things just don't "immaterialize." Forget about it.

Software is very protean, so you can call it a lot of things: you can call it art, science, free expression, mathematics, a medium, data, information, code, artificial intelligence, cyberspace; frozen thought; you can call it the noosphere and the Holy Ghost. But if you don't preserve it in some material form, you are not preserving immateriality: you are preserving nothing.

Now let's compare, let's say, a top of the line Apple iMac with some Van Gogh sunflowers. A 19th century oil painting versus a 21st century computer. Side by side, on their dual, heroic journey, into the distant future. Against all the hazards of the world. What are they up against, how are they doing? First, let's check those Biblical catastrophes: fire, flood, asteroids, check, check, check, they're both totally obliterated.

Let's get down to the everyday stuff. Number one threat, sunlight. Bad for oils, quite bad for plastics too. Ten years of hot sunlight on that glossy iMac console, and our candy-colored iMac is a lot less yummy.

Next, water. Glass of water onto the oil painting. Oh my. Glass of water into the computer? A catastrophe!

How about microbes? Not much to eat in a computer, but it sure attracts dust; that's because it's electrically charged. Just run your finger across that user-friendly screen: it comes up black. If you dare, look inside at the thick fur coats of dust growing on all those highly-charged chips and capacitors. That's pollen for microbes to eat. It's also specks of corrosive smog.

It's warm inside computers; bugs like warmth. Can you eat around a computer? You really shouldn't. Smoke around it? No way. Touch it with your big greasy hands? Maybe you shouldn't, but you've got to; it won't run otherwise. You're pounding that keyboard, you're rolling that mouse. A machine has moving parts, so it's taking physical wear every time it performs. It's not much use if it's cold and dead inside a vitrine.

So a computer's got the conservation problems that a painting has, plus a bunch of new ones. Computers run on electricity: they get power surges, floods of bad voltage that just crack components and blow them up. The screens are also vulnerable: cathode-ray tubes have big electronic guns firing beams at a powder.

Heat cycling cracks chips, stressing them mechanically every time you boot up or shut down. Maybe you don't dare to turn the computer off, in which case it simply sits there, sucking energy, tethered to the wall by a dangerous leash, costing you money.

Computers have dozens of components; if one malfunctions, it takes the whole machine down. If you try replacing a specific computer component, you soon find that the computers themselves were assembled by computers. They're made of a gaggle of components from factories all over the planet, picked up on the cheap in fast, fluid, friction-free, digital markets. These are super- sophisticated buying and assembling processes that are never quite the same day by day. That's why Dell and Gateway can make a computer just for you. But if it's made just for you, how can somebody else truly duplicate it? The answer is that you don't duplicate computers. You don't restore them to some past, pristine condition. That's unheard of. You just buy a brand-new one. A bigger one, a faster one. A different one.

Maybe it seems like I'm ducking the issue here, because we're not really trying to preserve mere computers, are we? Because when you really and seriously contemplate preserving a computer: for fifty years, seventy-five years, a hundred and fifty years == oh well that's obviously hopeless! Of course the thing would burn out from its constant torrent of voltage. Of course its plastic mechanical parts would jam and crack. Of course the screen blows up, the mouse wears down and you can't get replacement parts. The processor is the heart of a computer, but a chip fabricator is a vast enterprise with super-clean rooms, costing billions of dollars. You can't re-tool these huge factories to recreate obsolete processors.

But we're not worried about all that, are we? Because who cares about the merely material computer == we just need the immaterial software, the data. We all know how easy it is to save that stuff; you just pour it from an old computer into a new one. It's a stream of ones and zeros, all pure and clean and invisible, kind of like ether, or phlogiston.

This is the part where we really have to scrunch up and stare, ladies and gentlemen. Because every time that the computer industry confuses its hardware with philosophy, we've got a serious problem. A stream of bits is not just ones and zeroes. Ones and zeroes are numbers, and even if arithmetic is immaterial, computers aren't. Bits are not different from atoms: bits are bits of atoms. Bits are not ghosts or spirits or good intentions, bits have to be measurable, observable physical objects, like a Greek vase. Bits may be too small for the naked eye to see, but just like a cold germ or a hepatitis virus, they are most definitely around, and they're a lot of trouble. Bits are moving electrons, moving photons, or they are magnetized clumps of atoms, laser burn marks in plastic, iron filings stuck together with tape. That's what bits are.

Bits have no archival medium. We haven't invented one yet. If you print something on acid-free paper with stable ink, and you put it in a dry dark closet, you can read it in two hundred years. We have no way to archive bits that we know will be readable in even fifty years. Tape demagnetizes. CDs delaminate. Networks go down.

There is a whole chain of additional failure points, inherent in the nature of contemporary computers.

The recording media we have right now are bound to fail quickly: but that's just hazard one. The computer itself becomes obsolete: that's hazard two. Then comes the operating system: the OS can be replaced, or upgraded, and made incompatible with earlier versions, or it can simply vanish from the market.

The same thing can happen to the application that created the data. That application can fail in several places. It can't run off that operating system. It can't run on that make of computer. It can't even run off that storage system; in the case of multimedia or streaming video, maybe the bandwidth isn't fast enough; you've preserved every one and zero in fine order, but you can't move them fast enough or in the proper order, so you can't make it work.

Then there's a host of other little pests: video cards, sound cards. Different makes with a host of incompatibilities, so they refuse to plug and play. Different mice, different joysticks. Different monitors. Color systems: this software runs only under 256 colors. Weblinks may have been built into it; it may be making lonely system-calls to some piece of equipment or some data that no longer exists.

Compression algorithms may have squeezed the data into some unrecognizable shape. It may have been encrypted, carefully put under digital lock and key; but the key may be gone, the whole encryptor may be gone. The system can't boot because its extensions are incompatible. The software's got viruses. It's an emulator of an emulator's emulator.

When it piece of software decays, it doesn't degrade like a painting, slowly and nostalgically. When software fails it crashes; it means the Blue Screen of Death.

These are dire problems. They are massively underappreciated. That is why conferences like this are important today, and even more important to the future.

Let me end with a futuristic speculation. In his book CLOCK OF THE LONG NOW, Stewart Brand took a long look at the scope of this problem: systems too complicated to understand, with incompatible standards, and dying, obsolete hardware. A world with giant legacy systems too expensive to replace, too big to comprehend, being run by giant, unwieldly financial markets, utilities, health care systems, national and international bureaucracies. Software written in great haste and under terrible market pressures, digits in decay that lack any safe havens, like museums, where data might be kept and safely restored. And Stewart Brand sums up that future world in this way:

"The system doesn't really work, it can't be fixed, no one understands it, no one is in charge of it, it can't be lived without, and it gets worse every year."

That doesn't sound good. Yet it sounds rather plausible. Why? Because that is a universal human experience. Anybody pushing seventy knows that situation; that is the authentic voice of human mortality there: "I don't work any more, they can't fix me, nobody understands me, nobody's in charge of me, and I'm getting worse every year."

The reason history never ends at this point is that new people come along. New people with a fresh perspective, and a new set of understandings, and a bunch of problems they inherited. They can see the legacy of history in a different context. The future has become their past.

That's the kind of activity that I hope you'll be seeing at this event. It's a big job, but it's no surprise that the digital art world should declare itself an avant-garde in this matter. There are very bright people here and I wish them the best of luck. That's all I have to say tonight; thanks for your attention.

O=c=O O=c=O O=c=O O=c=O
O=c=O O=c=O O=c=O O=c=O