The Human/Nonhuman Assemblage
The Cultural Heritage as Big Data
the police. Archives are good at covering this on the level of
"analogue" materials (papers, documents, pictures, analogue
files), but when dealing with digitized or born-digital material,
the models of curating and organizing tends to adhere to
analogue models. Today computation - another word for what
codes, programs, software, protocols and algorithms do - is
involved in almost all working processes in a society. Businesses
and states, including the police, hospitals, armies etc., would
not be able to do their job without computation. Sometimes a
specific program or algorithm was inevitable and crucial for
completing a task. Should not state archives be interested in
harvesting some of these programs and algorithms? Indeed,
their job is to collect important documents in a work-process,
and if one of the key-elements of this process is algorithms, we
cannot easily circumvent this. Collecting them should essentially
not be any different from the traditional collecting of context
information (for instance, the organization of work-processes),
which is needed to understand the collected records. All key
elements of a work-process should be available.
Even if most programs contain many minor algorithms, we can
also talk about unique algorithms. Books like, John McCormick's
Nine Algorithms That Changed the Future (2011), explores this.
Tech-giants - Google, Facebook, Amazon and Microsoft - are
totally reliant on specific algorithms for their success. The Google
Search Algorithm is probably the most powerful algorithm in the
world right now. Many of these algorithms are of course business
secrets, but most of them are not, and they are usually available
and out in the open after a certain time. GitHubs are good at
tracking these declassifications of algorithms. Google Map,
TomTom and other GPS systems used by drivers all over the
world are the result of the US Department of Defense's decision
to declassify their Geographical Positioning Algorithm in 1983
after a tragedy involving a civilian airplane. The algorithm was
suddenly available for everybody. It changed driving habits
around the world.
"The robots are coming for Wall Street," reads a headline
from New York Times recently (25/02/16). It predicts that within
a decade between a third and a half of the current employees
in finance will lose their jobs to automation software.
The algorithms will take over. Will the same happen within the
archives? Yes and no. However, it is important to note that
"intelligent" machines may both liberate the archivists from
tedious routines and foster a new kind of creativity which will
demand more people and researchers. Humans are not
eradicated from the equation, but humans have to form a new
creative assemblage with machines. The new archivist is the
human/nonhuman assemblage. This makes the archives' services
more interesting - perhaps even mind-blowing.
I will here briefly discuss three areas where creativity and research
is of utmost importance: Big Data, Articulation and Curation.
Big Data is the new catch-all term. Everything has become big
data in the sense that all human activities emit and generate data
that can be stored and assessed to produce new knowledge.
This knowledge is usually characterized as predictive analytics.
Virtually all big companies run according to this model. Big Data
analytics originates in the military/space industrial complex. As
armies implemented cameras, satellites and computers in their
surveillance and intelligence efforts they gathered so much data
that human cognition could no longer assess it. Only computers
can deal with Big Data. Big Data algorithms have travelled from