High-Tech
Times Article 038
Blame
It All On the Robots!
Welcome
back to the High-Tech Times. The Internet has quietly taken over the jobs
and routine tasks of what used to be the province of clerks, librarians,
and information desks. And you can thank robots for these extra billions
of freed-up man-hours.
But
just what is a robot? The word is derived from the Czech word “robota,”
which humorously enough means “drudgery.” Robots have populated science
fiction novels for over 90 years. In the real world, robots have made their
way into hospital operating rooms, auto factories, and even the surface
of Mars. What differentiates the mechanical robot from programmable machines
is that the robot uses sensory input to make independent decisions without
operator actions. When I place my aircraft on automatic pilot, for example,
it’s a robot that takes over, interpreting flight data and adjusting flaps,
ailerons, and rudder to maintain the flight path.
The
software robot is an entirely different animal. Webopedia.com defines the
term robot as “a program that runs automatically without human intervention.
Typically, a robot is endowed with some artificial intelligence so that
it can react to different situations it may encounter. Two common types
of
robots
are agents and spiders.” And it is these little autonomous beasties that
make the Web such a useful place.
With
somewhere between 320 million and 2.2 billion Web pages that can be indexed,
finding anything except by accident takes some sophisticated help from
our robotic helpers. When you access your favorite search engine to find
your favorite song, game, or stock value, the search engine is not doing
a real-time search of the Web to get your data (except in a very few cases).
Instead, it is doing a search of files that have been found in the past
by robots, spiders, or other crawlers.
You
can think of a software robot as a curious animal that traverses the Web’s
hypertext jungle, retrieving one document after another, and also retrieving
any document or image that is referenced by the first document. Don’t be
mislead into thinking that these robots actually move along the Web like
a virus; a robot simply visits sites by requesting documents from them.
An
agent is a more advanced type of Web-crawler that does travel between sites,
deciding for themselves when to move and what to do. Agents are not as
common as robots, but have a lot more flexibility. An intelligent agent
is a program that helps users with things, such as choosing a product,
or guiding a user through form filling, or even helping users find things.
Internet browsers are also called “user-agents,” as they are programs that
perform networking tasks for a user.
A
more advanced version of search technology involves the use of natural
language processing. The “Ask Jeeves” robot <www.aj.com> is the most
famous of these, with over seven million questions and answers written
by people. If the answer it provides isn’t useful, Ask Jeeves also queries
and provides links to Web-sites that may provide the requested information.
Intelligent
service agents act as substitutes for information desks. The most familiar
form of this technology is interactive voice-recognition (IVR) offered
via the telephone, where the caller selects from a menu of items and gets
a voice recording of the information, and may also be transferred to an
appropriate department, if needed. A step up from a recorded-voice system
is a telephone-based system that searches a database for an answer, and
uses a synthesized voice to provide the information.
The
newest intelligent service agents combine a search engine and a natural-language
robot. These agents answer common questions in a conversational way. If
the agent doesn’t understand the query, it asks for clarification. For
example, if I were to ask “How do I write my State Senator,” the robot
would ask where I live to determine the answer. If it cannot give an acceptable
answer, it may transfer the caller to a human agent, or tell the user where
to call. These types of robots are also used to answer e-mail. They can
scan thousands of incoming messages, decide which ones are spam (and delete
them), which one are routine queries (and answer them), and forward the
remainder to an employee for handling.
There
are several additional robots to handle specialized tasks. Update bots
periodically search specified Web-sites or files and report back on any
changes. Shop bots are primarily used for business-to-consumer transactions,
but can also be adapted to obtain the best prices for goods and services.
Data-management bots simplify the process of filling out forms; other varieties
of these robots mine data from the Web and produce summaries.
Many
software products incorporate bots which are concealed as other functions.
Programs such as Microsoft FrontPage and Adobe GoLive allow the user to
create spectacular Web-pages without knowing how to write a single line
of HTML code. You work in a WYSIWYG GUI, while the program creates the
code in the background.
There
are thousands of Web-sites giving information on robots and artificial
intelligence, as well as hundreds of free bots that can be downloaded.
Two places to start are BotSpot <www.botspot.com> and WebCrawler’s Web
Robots Database <www.webcrawler.com/mak/projects/robots/active.html>.
There
used to be a big concern that robots would replace people and put them
out of work. But as technology develops, people are instead freed from
performing many repetitive tasks. And bots can often assist with ideas
and projects that were never done because of time considerations.
Just
because Deep Blue was able to defeat chess Grand Master Garry Kasparov
in 1997 doesn’t mean that you shouldn’t be taking advantages of the many
times when bots can help you. Select them carefully and train them well,
and robots will make your life a whole lot easier!
See
you next month.
|