What is
artificial intelligence (AI)? It is usually used to describe computers,
machines, or software that mimic “cognitive” functions that humans associate
with the human mind. Among them are “learning” and “problem solving.” The
definition of AI has changed dramatically over the years.
As
computers have become faster and more capable of simulating these cognitive
functions, many of these functions have been removed from the definition of AI.
A prime example of this is optical character recognition, which just a few
years ago was considered an intelligent function for computers, mainly because
the task was difficult.
Optical
character recognition is able to take a picture of a page of text and turn it
into text that can be edited in a word processing program or analyzed for
content. Dr. Larry Tesler created a theorem of AI that is commonly quoted as
“Artificial Intelligence is whatever hasn’t been done yet.” What he said was:
“intelligence is whatever machines haven’t done yet.” In Tesler’s definition AI
will constantly change because if it ever reaches a point that it can do anything
we can do, it would then just be intelligence with nothing artificial about it.
The
current technologies referred to as AI include human speech recognition,
strategic game systems, autonomous vehicles, intelligent content delivery
networks, and military training simulations. All of these will eventually
become such mainstream technology that something new will define AI.
There
are currently three main fields of study in AI, data analytics, human-inspired
intelligence, and humanized artificial intelligence. The most advanced of these
three fields in my opinion is the data analytics field. Analytics AI learns
based on past experiences to inform future decisions using cognitive
intelligence to represent the world, much like a toddler learns to talk. Among
tasks that fall into this category are image recognition, “Is this a picture of
a cat or a dog?”; text content classification, “Is this article for or against
a flat tax?”; intelligent routing, “Which route should I take to get around the
accident ahead?” and product recommendation, “Since you bought the new iPhone,
you might be interested in a case?” We have all experienced AI in our everyday
lives, probably without even noticing it.
The area
of human inspired intelligence makes an assumption that human intelligence “can
be so precisely described that a machine can be made to simulate it.” This
particular field brings about all kinds of ethical issues. Should we be
attempting to create artificial beings endowed with human-like intelligence?
There are several science fiction books and movies based on AI becoming a
danger to humanity. This is the area of AI that is addressed in most of these
stories. I believe the much larger risk of AI is the creation of mass
unemployment, as machines are more and more capable of performing everyday
human tasks.
The
third area is humanized artificial intelligence. This is using AI to replace
human characters in real life and in stories. Prime examples are the soldiers
in the popular video game series “Call of Duty.” These soldiers make
“intelligent” decisions based on the environment, including the interaction
with human players. These humanized beings are also used in telephone support
to remove the first level of contact. If you call a help desk, bank, or credit
card company and a computer answers with a human-like voice asking, “What can I
help you with today?” and gives the ability to answer in complete sentences, it
is using humanized AI.
Is AI
something we should be worried about, or something we should embrace? I leave
that up to the reader.
No comments:
Post a Comment