Before we can understand artificial intelligence, we must first understand what
intelligence is. We do not have an exact definition for intelligence, but in
this context, intelligence is the ability of something:
To find connections between pieces of information or to reason.
To gain knowledge or skill by study, experience, or being taught or to
learn.
To find an answer to a problem or to solve.
To gather information about its surroundings or to perceive.
To share information or to communicate.
Agents and Environments
To help us visualize how an intelligent system can perceive its surroundings,
we shall take a brief look at the concept of agents and environments.
An agent is any system that can perceive its environment using
sensors (input devices) and react to its environment using
effectors (output devices).
Humans are agents that have sensory organs (eyes, ears, nose, tongue,
and skin) as sensors and muscles as effectors.
Robots are agents that have cameras, microphones, and other similar
devices as sensors and motors, speakers, and other similar devices as
effectors.
Software robots (or softbots) may still be considered agents. They have
input fields as sensors and output fields as effectors.
An environment is the area wherein an agent operates. This
environment can either be the real world or an artificial world.
Human and robotic agents function in the real world. Hence, the real
world is their environment.
Meanwhile, softbots function in artificial worlds controlled by
programmers. Hence, this synthetic world is the environment of
softbots. This world usually consists of large databases, extensive
file systems, countless input fields, and several output screens.
The Premise
Artificial Intelligence is an artificial machine, usually a computer of some
sort, that can think like a human to some degree. The field of study which
focuses on creating AI is also called Artificial Intelligence. Many exciting
advancements occurred in this field over the past few years. However, let us go
back to the basics first and think about how machines become intelligent in the
first place. Considering most machines we see every day do not possess any sign
of intelligence, can we even make a machine that can think?
To settle this question, a mathematician named Alan Turing came up with a test
to determine how close a machine's intelligence is to a person's intelligence.
This test became known as the Turing Test. Essentially, this test compares the
performance of humans and computers by making them answer questions or perform
tasks.
From the Turing test, we derived the three types of AI:
Artificial Narrow Intelligence - Practically everywhere, from
games to the social media
Artificial General Intelligence - An AI that can pass the Turing
test
Artificial Superintelligence - An AI that is far smarter and
more capable than a human
The Advantages
Utilizing Artificial Intelligence has several advantages, such as:
AI can learn by looking at statistics and information. Meanwhile,
humans sometimes need an explanation to understand.
AI can teach itself much faster than humans ever can.
AI can respond at any time of the day, as long as it has a power
source. Meanwhile, humans need some time to rest.
AI is less prone to making mistakes compared to humans.
AI can efficiently organize data and get the most out of this data.
The Uses
AI also has a lot of uses in the following fields and more:
Data Science - AI can arrange, process, and present data
efficiently, thus saving data scientists a lot of time on their work.
Natural Language Processing - This is the mechanism behind
predictive text technology, a commonly used feature built into most
mobile phones. AI can predict the next word a user will type.
Game Development - AI can simulate the behavior of non-player
characters (NPCs). These NPCs make the game more exciting, even if you
are playing alone.
Computer Vision - This is the technology that enables
self-driving cars. AI can accurately let a computer identify objects in
an image.
In summary,
Artificial Intelligence is a machine that can think like a human to
some degree.
AI must be able to perceive, reason, solve, learn, and communicate.
We can think of AIs as agents functioning in an environment.
That environment can change to make AIs more efficient.
The Turing test tells us how human-like a machine thinks.
From the concept of the Turing test, we can derive three types of AI:
one that scores decently (ANI), one that passes the test (AGI), and one
that gets over perfect (ASI).
