Artificial Intelligence

See all articles from Sinc-Link v4 n5
Publication: Sinc-Link
Pub Details: Sinc-Link v4 n5 (Volume 4, Issue 5)
Date: September/October 1986
Pages: 7-8

It seems that I have always been interested in “the study of how computers can be made to act intelligently,” or in other words, Artificial Intelligence. But I suppose I really became interested in Artificial Intelligence, or AI for short, when I purchased my first computer, a Sinclair ZX81. I was fascinated by the way a computer could solve problems and follow instructions so precisely. However, I find myself becoming increasingly dissatisfied with all the new “high-tech” computer equipment I find myself surrounded by each day. What I am talking about is the problem of computer-programmers-headache. This is similar to “tennis elbow”, or a runner’s “athletes foot”. It’s not a new dilemma, but it’s an annoying one. No matter how advanced the technology is that we work with, there are always some tasks that become tedious with time. There are many ways in which smart AI systems can be implemented to make life more enjoyable for everyone. This technology should not be feared, computers will always be under the control of man. Even though man will leave many tasks to machines to think about and perform, he will still have many tasks to perform that can’t be carried out by any computer or machine.

One of the major problems with computers today is that they are stupid. They have to be reprogrammed to perform even the simplest new task. It would be ideal if someone could write a very smart program that could program other computers. The development and application of an AI system that could program computers is one of the most difficult, and fascinating projects facing computer scientists in this decade.

The development of an AI system would revolutionize automation.This type of automation would not be physical, but mental. In the near future, processes that require a logical examination of many facts, and require split second decisions, will probably be automated by an AI system. We might find this type of AI on assembly lines of the future. Consider this scenario: A GM car assembly line experiences a short circuit in one of its’ robot workers. As a result plastic sheilding on a high voltage wire nearby is accidentally ripped and left bare by the malfunctioning robotic arm. To make matters worse, the resulting sparks have started a fire in the spray painting section, setting off the fire-warning-system. Instead of allowing thousands of water sprinklers to flood the fire and the live wires, the intelligent assembly-line-monitor, directs an alternate course of action, far less dangerous than the former. It sounds the evacuation alarm. When all human workers have exited, it locks all human accessible doorways, sprays the electrical fire started by the live wires with CO2, and uses highly controlled water jets to extinguish the solvent fire nearby. After circuit and atmosphere tests show that no danger to humans remains, the doors are unlocked. Later, a complete emergency-diagnostic report is printed and filed, in unvolatile memory, for a permanent record of the accident.

Fantasy? No, most of the technology described above is available now. In fact, systems similar to this may even be in existence today.

I hope that this look at the advantages, and potential of Artificial Intelligence has given a greater insight into the current and future state of technology. We will benefit greatly from AI technology. We should not be afraid to leave a few, or many, decisions to machines if we have faith in our own intelligence and ability.

* Author:Unknown
* Downloaded from The Ascii Exchange