Page 31 - ASR_Forewords

Note from IDRBT Director,

Dr. A S Ramasastri

AI : A Few Philosophical Thoughts

Intelligence is one of the defining features of
being human and it comes in various forms such
as linguistic, spatial, mathematical, and emotional.
From a purely practical point of view, intelligence can
be defined as the capacity to absorb and learn from
experiences; it is the ability of deal with and address
problems and be able to adapt to new situations.
Learning is the process of acquiring new
understanding, knowledge, skills, values, attitudes,
and preferences. The ability to learn is possessed by
plants in a very limited way, animals to some extent,
and human beings on a larger scale. Human learning
starts at birth (might even start before) and continues
due to continuous interactions with other people and
environment. Some learning is immediate, induced by
a single event (like burning due to fire), but most other
learning accumulates from repeated experiences.

Brain is considered the seat of learning among
human beings. It is the most complex organ in the
human body; made up of about 86 billion neurons
that communicate in trillions of connections called synapses; with whose help it
experiences the world and learns. The brain receives inputs, sorts, and stores the
data, analyses, builds indexes and links, and when required retrieves it along with all
associated information. Despite advancements in neurosciences, several aspects of
intelligence and learning exhibited by the brain, remains in the unknown frontier.
Artificial intelligence (AI), unlike the natural intelligence displayed by brain, is
intelligence demonstrated by machines. The term AI is used to describe machines
that try to mimic cognitive functions possessed by humans, such as learning and
problem solving. Yet it is difficult to define what exactly AI is.
AI effect highlights the major difficulty in defining AI. As per the famous Tesler’s
Theorem, “AI is whatever hasn’t been done”. AI is an ever-changing goalpost. Plenty
of examples exist of the AI effect in action. For example, optical character recognition
has become routine technology and is often excluded from AI discussions.
It was once thought that a machine that could beat a grandmaster at chess embodies
an AI. Deep Blue achieved this feat in 1997 against chess grandmaster Garry Kasparov.
Then the goalpost moved and Go became the game AI needed to beat. (And it did
in 2016 when AlphaGo defeated Lee Sedol in four of five games.)
Chatbot that appeared as though it was talking to you was once considered intelligent.
But it is not considered now because it does not understand the intent behind your
messages. The point is, every time an AI completes a new feat, that feat is no longer
a benchmark.

The present AI capabilities may include areas like strategic games, autonomous
cars, and military simulations. And possibly certain critical areas in banks like fraud
detection, risk management, and customer behaviour.

Banks have been in the forefront of adoption of newer technologies for the past few
2