Artificial intelligence (AI) experts apply their skills in engineering
and computer science to create computers and software programs that can think
for themselves. Scientists must be able to recreate every tiny part of the
thinking process just to get a computer to walk, understand a question or
write music.
While sci-fi imagination can achieve that in the flicker of an eye, experts
predict it will take centuries before we really have a thinking, walking,
talking robot like Lieutenant Commander Data of the television series Star
Trek.
No matter how clever computers are, they're never smarter than the people
who create them. This is the challenge of artificial intelligence.
"It took humans 50 years to design the paper clip," says Tim Finin. He
is a professor and researcher in artificial intelligence at the University
of Maryland.
"How long is it going to take to create true intelligence? What's worse
is that the public think we're ready to have a Data in 10 years' time, when
I think it'll be more like 1,000 or even 10,000."
Work in artificial intelligence began in earnest in the 1950s. Yet it has
only been in the last 15 years that software has allowed computers to communicate
with one another through networks and the Internet.
At the root of artificial intelligence is one of the biggest philosophical
questions ever: what is knowledge?
Fred Popowich has been in the AI field for more than 15 years. His work
included a project to program computers to translate from Spanish into English.
"You have to use the ideas of the past -- stand on other people's shoulders
-- to get to today," he says.
Researchers have developed artificial intelligence systems that can be
compared with the abilities of a small child. According to Sue Abu-Hakima,
a researcher in the field, AI is still only trying to get computers to perform
the most basic tasks like providing an answer to a simple question. So, artificial
intelligence has a long way to go.
Still, the promise of this technology is very interesting to the business
world. Increasingly, private industry is supporting research in this field
because it offers opportunities for considerable profit.
Within the field of telecommunications alone, companies are scrambling
to get artificial intelligence into devices so that they can communicate with
one another. That will take the computer beyond the realm of a regular tool
and make it an irreplaceable fixture that will help people organize their
lives.
"Work that's being done in laboratories today will take 10 years before
it's out as a product," Abu-Hakima says.
Expert systems are among the new tools being designed by scientists to
help humans in their daily lives. It's a small part of the world of artificial
intelligence.
Some expert knowledge systems can help colleges screen student applications.
They also help foresters find out what kinds of bugs are damaging Douglas
fir cones, and help companies run their factories after their most experienced
workers have retired.
Intelligent appliances and the intelligent home are big items on the horizon.
A lot of companies, such as Whirlpool and General Electric (GE), are working
in this area.
Computer networking is another area of interest to AI specialists. Experts
in this area work to make it easier to find information from the world's biggest
and most complex database -- the Internet. These so-called smart agents would
be more "intelligent" than ordinary search engines.
Computer scientists normally work in offices, laboratories or universities.
Given current technology, telecommuting -- working from home -- is possible.
"AI people work in private industry and in universities," says Popowich.
There are few physical requirements in this field, according to Popowich.
Anyone who can work with computers can work in AI. Since artificial intelligence
specialists work with computers, however, they risk suffering from carpal
tunnel syndrome, eye and neck strain and back problems.