Conclusion

Next: Appendix: Programming with POP - 11: Development to Artificial Intellion Previous: Beyond the Symbol Processor

Artificial intelligence is the center of new business building intelligent computing models. The main assumption is that intelligence (people and others) can be represented by symbolic structures and symbolic manipulations that can be programmed with digital computers. There is much debate as to whether computers that are properly programmed are ideas or just computers, but AI researchers do not have to wait for the conclusion of that discussion. Aspects of intellectual behavior such as problem solving, reasoning, language learning, understanding are encoded as computer programs, and in very limited fields such as soybean disease identification, the AI ​​program is a human expert It is better than that. A major challenge for today's artificial intelligence is finding ways to express commonsense knowledge and experiences and conduct daily activities such as people conducting large conversations and finding their way in the downtown area It makes it possible to do. Traditional digital computers may be able to run such programs, or we may need to develop new machines that support the complexity of human thought.

The conclusion is clear that the brain functions like a computer rather than a continuous system like the weather. I am certainly not the first person who came up with this controversial conclusion. In fact, Stephen Wolfram makes extensive conclusions that all physical phenomena are caused by discrete calculations. In his book "New Science", Wolfram says, "What will happen to science if computers are found before Newton's calculus?" Computation of simple components The excessively luxurious Byzantine mathematical theory is not necessary, the fundamental cause of complexity comes from simplicity. Wolfram has not developed anything that proves this in an airtight manner, but I wonder why substances (and energy) are discrete at the quantum level.

You may not agree. You might say, "Wilka, some things are simple and you can easily draw conclusions." I agree. People's confidence in the conclusion should increase in proportion to the complexity of the situation. In other words, people should have high confidence in the conclusion of a simple situation and lack of trust in the conclusion of complex situations. For discussion, we will focus on when to construct a causal structure to explain a series of events. The structure of causality is just a proverb, you describe something through a series of causes and consequences. In literature, this is called inference - when you see the cause of some cause and cause, you infer reasons. I will use two examples.