← Back to blog

Decay Value

February 19, 2016

Where does information go to die? This is one of those phrases that has popped into my head from time to time without getting traction. Last week I got an email that gave the phrase traction: “PDFs are where information goes to die, rather than to be used.” This observation by Ben Wellington refers to the particular uselessness of public data released by governments—New York City’s, in this case—that comes trapped in a format difficult to parse using the tools available for grappling with data at scale.

Maybe, though, PDFs are an example of information hiding rather than dying. Let’s think of dead information as digital ruins—a societal product that has outlasted its usefulness to that society. Because information increasingly tends to live forever, we can expect more digital ruins to accumulate than built ruins, which inevitably (and helpfully) decay. Digital ruins decay too, but in a different way: by losing relevance while remaining intact, as their content, format, or context separate them from contemporary streams of meaning or retrieval.

Ruins are a signal to the present about past relevance, a message in a bottle. Digital ruins, by this definition, are not the information stored on a hard drive, or even in a PDF—which, in Wellington’s example, someone still wants and would like to access more easily—but information that nobody needs or cares about. Increasingly invisible in its raw form, information largely accumulates on servers and hard drives, while buildings linger on in full public view, thereby inducing at least the minimum degree of interest reserved for the visible. To describe the built environment in informational terms, it is searchable in the broad sense.

There are thus three criteria by which information degrades in usefulness over time, as an impossibly huge flood of new information pours down on top of it:

  • Fidelity: Analog media (including buildings) decay gradually, bleeding away their original content in that process. Digital media doesn’t. The vast majority of information that exists today, growing exponentially, is the digital kind that lasts indefinitely but can still disappear through intentional deletion or large-scale crashes that wipe everything out at once.
  • Relevance: Thanks to the fidelity and durability of digital records, outright loss or decay of information is less common than ever. The increasingly familiar version of digital ruins is the perfectly preserved relic of past significance, like the (now-iconic) derelict Space Jam website, seemingly untouched since 1996, or an abandoned LiveJournal.
  • Searchability: PDFs, paper documents, analog records of all kinds, and obsolete digital formats confound the “search don’t sort” paradigm that Google ushered in, unless converted to a more current format. Other information might be unsearchable because it occupies a position in the network or stream that makes it difficult to find, or has characteristics that make it difficult to search for.

Fidelity, relevance, and searchability: Each of these qualities is inherently desirable, but under conditions of information surplus, there’s also a desirable version of each quality’s opposite. Some information probably should decay at a certain rate; not all information needs to be relevant; and there’s certainly information that shouldn’t be easily searchable.

The theory of ruin value, infamous for its association with Albert Speer and the Nazis, posits that buildings should be designed and built to leave aesthetically pleasing, durable ruins for posterity. Irrespective of that theory’s merits, the digital requires a completely different approach: When data lasts forever by default, durability is assumed, and what’s needed is a plan for how it can decay in the most socially beneficial way. At one extreme is preserving all this data in an unindexed mass, like the Library of Babel; at the other is a memoryless society that keeps no records and burns everything after reading. Both extremes have the same outcome.

Ruins reflect a series of decisions by the society that built them and the societies that followed those—to construct something in the first place, to preserve it, or at least to not demolish it. Where information is concerned, one inherent goal of civilization is to reduce entropy as it encroaches (societies that fail to restrain entropy disappear). The best ruins accomplish this, by maintaining connections between otherwise disconnected cultures and ideas. Information is central to the goal of minimizing entropy, and a disorganized increase in data often means more entropy as well. Letting information die in the right ways is a necessary process in the larger effort to minimize civilizational entropy and create or maintain cultural meaning.

The euphoria of the early Information Age manifested itself as a gluttonous appetite for the virtually unlimited storage, reproducibility, and universal instant access that had never been possible before. Like the automobile in the 1910s, the benefits of the future had arrived in advance of that technology’s particular costs, and bad habits emerged. Email and the stress surrounding it is one strong indicator of the general need to make decisions about how to let go of information—early evidence that we need guiding principles for how to preserve that information or let it decay. The growing popularity of ephemeral chat apps as a replacement for email, and the unspoken philosophy that lies beneath that preference, is easily mistaken for an increase in societal entropy and frivolous exchange but is really the opposite: an entropy-reducing decision about what information to kill and what to save, a decision that email doesn’t facilitate as easily. A nascent Theory of Decay Value.