Page 406 -

405
Q9-9 2026?

mining application for analysis. The application finds patterns, trends, and other business intel-
ligence and reports the results to the human analyst. The BI analyst examines the results and pos-
sibly iterates by finding more data and running more analyses.
But what happens when BI applications become sophisticated enough to replace the BI ana-
lyst? What happens when the unsupervised data mining application has features and functions
to find its own data sets and to evaluate those data sets based on the results of a prior BI analysis?
And then decides which BI analysis to perform next?
Machines work faster than humans, and they work 24/7. At some point, will machines know
so much about us that we are incapable of understanding the results? What happens when,
because of complexity, such BI machines can only communicate with other BI machines?
Ray Kurzweil developed a concept he calls the Singularity, which is the point at which
computer systems become sophisticated enough that they can adapt and create their own soft-
ware and hence adapt their behavior without human assistance. Apply this idea to unsupervised
18
data mining. What happens when machines can direct their own data mining activities? There
will be an accelerating positive feedback loop among the BI machines. Then what will they know
about us? Is it important that at that date we will lack the capacity to know what the machines
will know?
This line of thinking exposes a future flaw that runs through this text. We’ve defined informa-
tion as something possessed only by humans. If it’s on a piece of paper or on a screen, it’s data. If
it’s in the mind of a human, it is (or can be) information. When we’re talking about simple report-
ing operations such as grouping and filtering, and so on, that’s legitimate. But, in the day when
unsupervised data mining truly is unsupervised, machines will possess and create information for
themselves.
Do you know what your data mining application is doing tonight?