Five Basic Information About Artificial Intelligence (AI)

Five Basic Information About Artificial Intelligence (AI)

Five Basic Information About Artificial Intelligence (AI)

1. A software package, not an electrical device

Artificial intelligence is designed to design a machine capable of achieving a specific goal in a way that is similar to humans (or beyond human capacity). Often they are algorithms (specific software) running on a computer or a group of computers (such as in the image). It is not an alternative electrical device to the human brain (mistaken belief in some).

2. It consists of a set of different goals and abilities

Including Machine Learning, the ability to organize and understand science, NLP, speech recognition, computer vision, problem solving, creativity , Emotional and societal interaction, motivating robots (not an essential part of artificial intelligence, although some people think it is artificial intelligence), general intelligence (AGI), including doing all of the above, in addition to super intelligence. There is no agreement on the supernatural intelligence among scientists, but the way to reach it is that once intelligence is reached in human form it can be copied and the goal of these algorithms is to design an AI agent smarter than to return the ball to design supernatural intelligence. In theory, all you need to make supernatural intelligence is a set of powerful computers and plenty of electricity - if you get intelligence at a human level.

3 - Search old and complex

The dilemma of artificial intelligence is old and complex and has many details and there is a difference between scientists on what comes within the artificial intelligence and a difference in the ways to be followed in order to reach the goal. There was great optimism 70 years ago that within 20 years the dilemma of artificial intelligence would be solved !! So far we are very far from solving it completely. There is a belief in a large group of scientists in the field known as Connectionism that all artificial intelligence can be solved only by connecting a large network of artificial neural networks and training them on a very large amount of data. This trend is embraced by companies like Google.

4 - Abundant applications, of which we have not seen very little yet

His research greatly contributed to the development of useful tools such as machine learning (ML), which in turn contributed to applications in medicine and economics (applications for tools rather than artificial intelligence itself). The applications of artificial intelligence are great and will change the lives of people very much, we have seen few of them in our daily dealings (Google applications mostly use one kind of artificial intelligence, the new iPhone is used for face recognition), Amazon uses to sell suitable products (Recommender System ), And share price prediction programs, relatively speaking, we have not seen much of the artificial intelligence yet, but all of this will change radically Auto-driving and early detection of cancer are among the most important current applications being developed. Our daily lives will be driven by it In order to increase productivity or cost reduction or to replace the human side with Automation.

5. Transformation point: data and computing power

Although science related to artificial intelligence has not developed radically in the last decade, the field has returned to the surface strongly. What happened recently? The vast abundance of data, and the significant development of computing power and specifically parallel computing, found in graphics cards for game consoles (GPUs). All this contributed to the ability of the old algorithms (from the eighties and nineties) to work very well now and beyond the known hand-crafted features in many human tasks. It is now easy to train these models to perform tasks that were considered difficult or almost impossible on the machine, such as identifying cancer cells or self-driving cars.

Post a Comment