Basics and Fundamental concepts of Artificial Intelligence
Artificial Intelligence is software which has distinctive behaviors and specific features that emulate human mental abilities and methods of action. One of the most important features is the ability to learn, infer and react to situations that have not been programmed into the machine.
Artificial intelligence (AI) is a branch of computer science that insists on the creation of intelligent machines that act and behave like humans.
Many of the books define artificial intelligence as "Artificial Intelligence is the study and design of smart customers." Smart customers or clients are considered as a system that accommodates their environment and takes positions that increase their chances of success in fulfilling their mission. Artificial intelligence is also defined as "the system's ability to learn from external data, interpret this data correctly, and use the basic knowledge acquired from this data to achieve specific goals and particular tasks through flexible adaptation".
Five Basic Points About Artificial Intelligence
1. Artificial Intelligence is a software package, not an electrical device
Artificial intelligence is designed to operate a machine capable of achieving a specific goal in a way that is similar to humans or resembles human behavior or beyond human capacity and strength. Artificial intelligence is often a computer algorithm (specific software) running on a computer or a group of computers. Artificial Intelligence is not an alternative electrical device to work with the human brain, it is software.
2. Research in artificial intelligence is old and complex
The artificial intelligence dilemma is old and complex and has many details. There is a difference between scientists on what comes within artificial intelligence and a difference in the ways to be followed in order to reach the goal.
70 years ago, there was great optimism in artificial intelligence that within 20 years, the artificial intelligence dilemma would be solved!! So far we are very far from solving it completely. There is a belief among many scientists in the field of machine intelligence known as connectionism that all artificial intelligence issues can be solved effectively only by connecting a large network of artificial neural networks and training artificial neural networks on a very large amount of data to perform properly. This trend is embraced by companies like Google.
3. Point of transformation: 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 enormous abundance of data, and the significant development of computing power and specifically parallel computing, found in graphics cards for gaming devices- Graphics processing units (GPUs). All this contributed to the ability of the old algorithms (from the eighties and nineties) to work fine now and overcome the known hand-crafted features in many human functions. Now it has become very easy to train these types of models to perform tasks that were considered difficult in the past, or almost impossible on the machine, such as identifying cancer cells or auto-driving cars.
4. Massive applications, of which we have not seen very little yet
Research in artificial intelligence greatly contributed to the development of useful tools such as machine learning (ML), which in turn contributed to applications of artificial intelligence in medicine and economics (applications for useful tools rather than artificial intelligence itself).
The applications of artificial intelligence are great and will change people's lives dramatically. We have seen a few applications of artificial intelligence in our daily dealings (Google apps mostly use one type or another type of artificial intelligence, the new iPhone uses it for face recognition), Amazon uses it to sell suitable products (Recommender System) And stock price prediction programs.
Relatively speaking, we have not seen much of artificial intelligence until now, but all of this will change radically! Auto-driving or self-driving cars and early detection of cancer are among the most important current applications that have been developed in a few years. Almost every area of ??our daily lives will be artificially inserted by one artificial intelligence to increase productivity, cost reduction or to replace the human side with automation.
5. Artificial Intelligence consists of a set of different objectives and capabilities
Artificial intelligence includes the ability to learn (Machine Learning), the ability to organize and understand science (Knowledge Representation), the ability to analyze language (Natural language processing - NLP), the ability to understand speech (Speech Recognition) Artificial Intelligence has many other capabilities such as computer vision, problem-solving, creativity, emotional and societal interaction. Motivating robots are not necessary component of artificial intelligence, although some people think they are a part of artificial intelligence). Artificial general intelligence (AGI) includes all the above, in addition to supernatural intelligence or simply superintelligence.
There is agreement on the superintelligence among scientists, but the way to reach it is that once the intelligence is reached human level can be copied and make the goal of these algorithms for strategic agents (AI agent) smarter and return the ball to design a supernatural intelligence. In theory, all you need to make supernatural intelligence is a powerful set of computers and plenty of electric system- if you find intelligence at a human level.Disclaimer- This information is entirely by a computer program and has not been created or edited by Just Learning.