The Computer Revolution/Artificial Intelligence/Diagnosis

A form of artificial intelligence is being used more and more frequently in diagnosis. There are all types of areas where A.I. can determine whether the behavior of different systems is correct. Initially, the easiest systems for computer intelligence to calculate whether systems are being run to specifications is other computers. Automobiles are a prime example of this. There are increasing numbers of sensors for every new model year in automobiles. A computer can be plugged in and, after copying the information recorded by the sensors, make a determination as to any problems that fall outside the normal range of use. It is important for computer system algorithms to be able to determine which part of a system is failing, where that part is located, the type of fault in the part and many systems now provide the steps to fix the problems. All diagnosis is based on observations which are taken by different sensors.

A major determinant to the effectiveness of A.I. to accurately diagnose problems is the base of knowledge it works with. The number of variables involved in even simple diagnoses like an automobile are large. Then for each variable there is a different weight applied to each. Then each weight can change over time. These calculations become more and more complex the more variable the subject. The initial knowledge usually comes from people who are experts in a particular field. However one of the major tasks of the creation of A.I. is the development of intuition which mankind is able to utilize.

In time it is likely that artificial intelligence will be able to diagnoses with a high degree of accuracy, not only technical systems with the necessary number of sensors, but organic matter as well. Walk into a doctors office and let the wall know how you are feeling. Next thing you know the room speaks to you and tells you both your diagnosis, treatment, and prognosis.