曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
in recognizing complex patterns.
Complex pattern recognition
is critical in two quite different
kinds of tasks – optical recognition
and physical movement
(security guards, etc.), and tasks
involving higher order cognitive
skills. We can usefully divide these
higher order tasks into two broad groups.
The fi rst are tasks that involve solving new
problems – problems that cannot be solved
by applying well-understood rules. The second
consist of tasks that require explanation,
negotiation, persuasion and other forms of
intense human interaction. We will call these
two sets of tasks, respectively, tasks requiring
expert thinking and tasks requiring complex
communication.
In joint work with David Autor of MIT, we
have argued that tasks requiring expert thinking
and complex communication are two of
4-year college graduate high school graduate high school dropout
30
25
20
15
10
5
1973 1978 1983 1988 1993 1998 2001
2001 DOLLARS
source: Data from Current Population Surveys as reported on the Economic Policy Institute Web site.
REAL HOURLY WAGE TRENDS FOR FEMALES 1973 – 2001,
BY EDUCATION
Fourth Quarter 2004 73
fi ve broad kinds of tasks carried out by the
U.S. labor force:
Expert thinking. Solving problems for
which there are no rules-based solutions.
Examples include diagnosing the illness of a
patient whose symptoms seem strange, creating
a good-tasting dish from ingredients that
are fresh in the market that morning, and
repairing an auto that does not run well but
that the computer diagnostics indicate has no
problem. While computers cannot substitute
for humans in these tasks, they can complement
humans in performing them by making
information more readily available.
Complex communication. Interacting with
humans to acquire information, to explain
it, or to persuade others of its implications.
Examples include a manager motivating
workers, a biology teacher explaining how
cells divide, and an engineer describing why
a new design for a DVD player is an advance
over previous designs.
Routine cognitive tasks. Mental tasks that
are well described by logical rules. Examples
include maintaining expense reports, fi ling
new information provided by insurance
customers, and evaluating applications for
mortgages. Because these tasks can be accomplished
by following a set of rules, they are
prime candidates for computerization.
Routine manual tasks. Physical tasks that
can be well described using rules. Examples
include installing windshields on new automobiles,
and packaging pills for pharmaceutical
fi rms. Since these tasks can be defi ned in
terms of a set of movements to be carried out
over and over in exactly the same way, they
are also candidates for computerization.
Non-routine manual tasks. Physical tasks
that cannot be well described as following a
set of if-then-do rules because they require
optical recognition and fi ne muscle control
have proven extremely diffi cult for computers.
Examples include driving a truck, cleaning
a building, and setting gems in engagement
rings. Computers do not complement
human effort in carrying out most such tasks.
As a result, computerization should have little
effect on the percentage of the work force
engaged in these tasks.
Earlier, we saw how the nation’s occupational
distribution has changed in the last
three decades. In the same way, we can look
for changes in the nature of the tasks that
comprise this work. While the Census does
not ask about the content of work, a second
survey does – the U.S. Department of Labor’s
Dictionary of Occupational Titles (DOT).
The DOT is a compilation of 12,000 detailed
occupational descriptions, each containing
professional observers’ ratings of the training
time required for the occupation, the occupation’s
physical and cognitive requirements,
©lester lefkowitz/corbis
74 The Milken Institute Review
and other characteristics. Because an occupation’s
details can vary across work sites,
each occupation is rated by workers in multiple
sites and the DOT provides an average
of the ratings.
The DOT data is far from perfect. Updates
are infrequent, making it diffi cult to track
changes in tasks that occur within occupations.
Nonetheless, the DOT allows an
approximate translation of changes in the
distribution of occupations into changes in
the kinds of tasks that people perform in
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
航空资料22(8)
