Big Data and Cloud Culture - To the Cloud: Big Data in a Turbulent World - page 223

of science and technology. Clearly the former builds clouds out of iction
and is assessed for its capacity to create worlds that may or may not bear
a close relationship to the world we know, whereas the latter create clouds
of data and applications that are judged by their capacity to represent
an empirical reality. But it is all too easy to dwell on simple differences;
it is more important to consider the subtle ones that shed light on each
enterprise, particularly by providing a cultural grounding from which to
think about cloud computing.
For Mitchell, the cloud that counts is drawn from a rich pool of subjec-
tivity, including emotional intelligence, that is constantly sensitive to the
risk of reducing consciousness, character, spirit, or soul, to a few notable
data points. Cloud Atlas is not just a story about the seeming universal-
ity of people preying on others, mainly for material gain, but also for the
sheer pleasure of domination, and it is not just a tale about how people
respond, sometimes successfully but often not, through struggle and
resistance. If this were all that mattered, we would not need a cloud atlas
because all clouds would be the same. Their richness and diversity emerge
from the historical context in which each node in the network of clouds
is immersed. This is often missed in big-data analysis, which addresses
history by examining networks or even networks of networks over time,
but does so through a process of extrapolation, typically from quantitative
data. It is an approach that has dificulty with those key historical turns
or slow, crescive changes that are vitally inluential but hard to detect.
To correct this problem requires imagination and experience as well as
human or machine intelligence.
Making matters more complex are the subjective categories and inter-
pretations of those, including the novelist and the reader, who provide
descriptions and assessments. The classic description of the communica-
tion process, Shannon and Weaver's mathematical model (1949), distin-
guishes transmitter from receiver, information source from destination,
and signal from noise. When it is relatively easy to identify each of these,
primarily when each step in the process is mechanized, the model makes
some sense. But for most forms of human communication, the terms are
far more ambiguous than it might at irst appear. “Boundaries between
noise and sound are conventions,” declares Frobisher in Mitchell's novel,
and all conventions can and should be transcended. As Nate Silver (2012),
one of big data's best-known champions, understands, one cannot simply