大至急、英訳お願いします!!
Researchers in the field of artificial intelligence (AI) may get a useful lesson from the history of artificial flight. The development of aircraft succeeded only when people stopped trying to imitate birds and instead approached the problem in the ways, thinking about airflow and pressure, for example. Watching hovering birds inspired the Wright brothers to use wing warping – turning an aircraft by twisting its wings – but they did not set out to imitate the bird’s wing. Starting with a box kite, they first worked on achieving sufficient lift, then on vertical and horizontal stability., then on steering and finally on power and engine design, carefully solving each problem in turn. After that, no airplane could be confused with a bird either in its overall shape or in its flying abilities. In some ways, aircraft may never match the elegant precision of birds, but in other ways, their performance is clearly superior to that of birds. Aircraft do not land in trees, catch fish from the ocean or use the natural breeze to hover motionless above the countryside. However, no bird can fly at 45,000 feet or faster than sound.
Rather than limiting the scope of AI to the study of how to imitate human behavior, we can more usefully think of it as the study how computing systems must be organized in order to behave intelligently. AI programs are often components of larger systems that are not themselves labeled “intelligent”. There are hundreds of such applications in use today, including those that make investment recommendations, analyze medical symptoms plan military troop and supply movements, schedule the repair of the space shuttle and detect illegal use of credit cards. These systems make expert decision, find meaningful patterns in complex data and improve their performances by learning. All these actions, if done by a human, would be taken to display good judgment, expert knowledge or responsibility. Many of these tasks, however, could not be done by humans, who are too slow, too easily distracted or not sufficiently reliable. Our intelligent machines already surpass us in many ways. The most useful computer applications, including AI applications, are valuable exactly by virtue of their lack of humanity. A truly humanlike program would be just as useless as a truly pigeon like aircraft.
