Page 140 - Computer - 7

DISCIPLINARY APPROACHES TO AI

The table given below describes the various disciplines of learning and study that contribute towards
AI’s developments and measure its implementation.

Discipline Contributions Questions
Studies the modelling, creation, and How can we code this behaviour?
Computer Science implementation of software that How can we improve the response
performs the “intelligent” processes. time of this program?
How can we build a device that will
Studies the construction of the be sensitive to its environment and
hardware that becomes a computer.
Computer Engineering be able to respond accordingly? How
It intersects the fields of Computer can we build a device that can be
Science and Electrical Engineering.
programmed to act “intelligently”?
How do we reach a logical conclusion
Studies knowledge, values, reason, given incomplete information?
mind, logic, and language, among
Philosophy What is knowledge? How is it
other issues. It also includes the field represented? Is a given behaviour
of ethics.
ethical?
How can we produce a given
Studies the particularities of the behaviour? What leads people to a
Psychology
mind, reasoning, and behaviour. certain behaviour? How do people
learn?
How can we create a mathematical
Studies the numerical relations of
objects and abstract representations. model to represent knowledge?
Mathematics & Statistics What can be computed? Is this
Provides numerical tools to represent model robust enough for the given
and operate models.
data?
Studies the workings of the brain, How does the brain store
Neuroscience including both unconscious and information? How does the brain
conscious processes. process learning?
How do we communicate? Can we
Studies language and communication,
Linguistics including its symbols, syntax model human communication? Is
there a relation between language
(structure), and semantics (meaning).
and thought?
How do living things reproduce and
Biology Studies living things. evolve? What are the mechanisms
of evolution?

Thus, Artificial Intelligence can be defined as a field of science in which different disciplines converge
with the purpose of creating agents that have the ability to learn, to create models from their learning,
to make predictions based on those models and new data, and make autonomous decisions based on
those predictions and any environmental data.

138