People and different books and papers define different meaning for Artificial Intelligence; depend on what the application is. As its name would suggest, the very first defining of artificial intelligence would be the development of a system, usually in computer, that perform tasks that requires human intelligence. In computer science, artificial intelligence refers to the systems can define the environment, by data recognition, voice recognition, and any kinds of sensors; and the system is able to apply the best action on the environment in order to maximize the profit. The common thing in all definitions is the nature of data, which is big.
To shed a light on that, imagine the process of walking in human’s brain. While you are walking, your brain works as the main processor of the system, which is collecting data from your eyes with a high frequency and sending signals to your muscles to keep your stability. I this example, your brain is the main processor, your eyes are sensors (getting data from environment), and your muscles are actuators. The final goal of this process is to keep your stability; from now on we refer to final goal as a profit.
Let’s do the math for this simple example. Eyes capture approximately 255 frames per second. If you walk a mile in 15 min, your brain receives 229500 (15*60*255=229500) frames (which is input data) and it is sending the corresponding right signals to all your muscles to keep your stability. In other words, your brain is predicting the next moment and it’s applying the best decision. Do you think if it is possible to formulize our system or drive a formula for that? Now imagine that you are walking in a windy day. In this case, you brain not only get data from your eyes, but also from tactile, and maybe other sensation. How big will be data for 15 min of walk, considering all sensation? Is it still possible to formalize the system?
In my own definition, artificial Intelligence is continues learning process throw finding patterns from big samples data to predict the future and applying the best action ,corresponding to your prediction, to get the highest profit. Just like a small child that learns how to walk by trial and error, and by continues learning in his/her adulthood, we want to improve methods of learning and decision making.
Let’s say your company’s short term strategy is highly dependent on the industry stock price. Is it possible to predict stock market with artificial intelligence methods? To answer this question we should look for samples data, and a method of finding the pattern out of samples data. Then we can predict the future of your industry’s stock market. Artificial intelligence method is a tool to find the pattern and predict the stock market by analysis of stock market history for like 10 years. Do you think broker reviews can improve our predicting approach? What else can improve our intelligence method?