metadata as history


my thought today1 upon waking: will metadata, which is a waste product of machine—machine interactions, become the historical record, ie externalised or technical human ‘memory’. ie will it (functionally if not literally) replace the library/archive: this person was here, that person was there, said this to this person, did that to that person, manipulated this, re-coded that, exploited these people, screwed those ones, invented this new value-extraction-from-living-flesh app which saved us all, ergo made civilization, programmed this satellite community’s shoddy air system, modified x gene in their lab-grown meat, sent the ppl in that other satellite to their deaths in space in the name of fearless technological progress but actually for the insurance money, etc… and this info wouldn’t be in a library, in the sense of a funcionally separated/localised institution, it would just be backed up in the logs on the machines running whatever different kinds of infrastructure. maybe it would be located in a data centre, maybe it would be in an totally localized on-board system. but anyway, data destroys distance and what matters instead is functional connections and proximities: metadata would be proximate (in the digital sense) to the functions it pertained to. with this, the archive would have converged with infrastructure itself, with the entire built environment. anyone could go looking for information, and anyone could become a ‘researcher’ or ‘historian’ by ‘reading’ or interpreting the logs, which might, for example, contradict aspects of the civil war of deep fakes permanently raging up on the medial surface. you could try to intervene as a ‘truth teller’ by spilling the logs, as though the atomic log entry was the apotheosis of factual truth, and people would say no you’re peddling cowshit, this deep fake is truer than your stupid factoids. and anyone could collate stacks of this information based on any parameter they liked. they might collate all data pertaining to a particular person on heaps of different servers, creating a kind of live log feed of someone’s quantified self, and so work biographically. (who knows, at some point in time it may be trivial to convert this into a map and/or animation, ie both a replay and a live broadcast of a life.) or they might pull data based on exclusive access to particular processes in massive corporations/institutions (big man view of history), or they might pull data based on existing social profiling, do research/a feed on say everyone in russia, try to gain some insight into how russians live today (everyday life history). by this point though it is starting to sound like the ‘digital humanities’ will basically have fused with totalitarian social profiling, customerization, propaganda, and information warfare in the pursuit of total power. which in a sense sounds a lot like large swathes of… the analogue humanities. you could have ‘research centres’, ie sweatshops of data-workers, that just pull whatever mountain range of data is needed by whatever centre of political or economic power for whatever salvo is on that day, as think tanks already do with their reports.

i always thought the ‘tank’ in think tank was the same as in fish tank, but actually it makes more sense to think of it as think sherman tank, no?

metadata is both more private and more public than most archival information: logs are normally protected by basic server security, which operates on the assumption that folders and files can only be accessed by users with sufficient privileges, ie only those who need to know. and the people they pertain to typically have no privileges at all. so these records are by default totally privatized, they don’t or aren’t supposed to exist anywhere outside of a digital vault (of the becoming-value of information), and any would-be researcher would have to obtain permission from the infrastructure owners to access them for the purposes of research, which would obviously be subject to power and interest like any other instance of withholding or granting access to information. or of course they could turn hacker to obtain non-whitewashed data, provided that is, that the data was not whitewashed in the same moment it was generated. but once accessed, such data can be immediately sent anywhere on the network and duplicated infinitely. if it did end up becoming a kind of currency of the historical record, it would then (already) be subject to the manipulation needs of power (eg to a bit of good old find-and-replace, in order to shift time, or shift identities). then differing versions of different sets of metadata could end up standing for different readings of events, with one set (to take a very simple example) ‘proving’ that x person arrived at y time and so was guilty of z crime, another set ‘proving’ that no they arrived 2 seconds later on, and so could not have been present. war at the level of information units.

  1. it wasn’t really today, but i’m changing the historical record. it was today when i wrote the word ‘today’.