FEATURE: ARTIFICIAL INTELLIGENCE
//////////////////////////////////////////////////////////////////////
Artificial Intelligence
(AI) techniques have
experienced a resurgence
during the 21st Century
and has become an
essential part of IT. But
even though they have
helped solved many
complex problems, is
there still room for
the human brain in the
modern workplace?
Andrew Quixley, Watson
and Cloud Platform
Leader at IBM, analyses.
A
rtificial Intelligence is being
used in two broad ways; to
analyse very big data sets
much more deeply and
effectively to produce better answers.
We are able to access unstructured
information, like words, speech and images
by using a computer rather than a human,
and because machines can do it, it means it
speeds it up and you can scale it up to any
you like. With images, tremendous progress
is being made to understand what’s in an
image and being able to do something
about it.
The second main way that AI is being used
is to automate tasks that require decision
making. If you have a machine that makes
a widget, and it’s the same one every
time, then there’s no decision to be made.
But we’re talking about automation that
humans do inside their heads rather than
generally with their hands. It’s called Robotic
Process Automation but it’s a confusing
term because it generally doesn’t involve a
physical robot.
It only really happens on a computer
screen and could be used to automatically
pay an ‘easy to pay insurance claim’ or
automatically reject ‘an easy to reject’ one.
And it leaves a grey area in the middle where
it can be a lot more complicated and there’s
a lot more to it than may first appear.
This relies on modelling previous decisions
and actions that humans take and
determining what the unwritten code is. The
Machine Learning capabilities of AI will create
a model that is followed again in the future
and AI can extract the meaning of what a
human wants and automate the process.
What are the security
implications of using
Artificial Intelligence?
At IBM, a lot of our AI services are web
services in the cloud. There are implications
if you want to know where your data is
going to be and where it’s going to reside,
and your local laws may prevent you from
exporting sensitive information.
The implications for us as a vendor are
that we have to accommodate that and
we’re doing that by making private clouds
available for data.
Artificial Intelligence and Machine Learning
are also useful in breaking codes where code
making and code breaking are constantly
trying to leapfrog over each other. There will
always be that arm wrestle between the two
because it will always be necessary for the
Artificial Intelligence
revolutionising IT
34
INTELLIGENTCIO
www.intelligentcio.com