AI – Artificial Intelligence

AI or Artificial Intelligence, is the intelligence of machines. We humans and many animals have natural intelligence. We learn from our experiences and improve with time. We also have intelligence to decide what to do in a situation which we do not have previous experience. These capabilities come from our brain and natural instincts.

Basics

Machines or Computer on the other hand usually has the intelligence that is fed or programmed to it. It acts according to the instructions in the program and acts accordingly. Sometimes we may think that computers show intelligence, but normally that is not true. They do not know what to do in a situation which is not pre-programmed or will not take a wise decision. Artificial Intelligence or the Intelligence of Machines was formulated to make the computers learn and decide based on the data fed to them and learn from analyzing the huge amount of data it can process. An AI machine must maximize it’s chance to successfully solve a problem and imitate the human mind in learning and problem solving. Systems that involve AI are speech recognition systems (natural language processing), machines playing chess and other strategic games, self driving cars, medical diagnosis, genetic engineering, weather forecast, simulators, image recognition, targeted advertisements, predictions based on data etc.

The term Artificial intelligence has been used for long, but actual intelligent systems were developed only in the last two decade or so. These researches require large funding and the success rate cannot be guaranteed. So the development in the field was in a very slow pace. To make a machine intelligent, it must have any or many of the features – reasoning, learning, natural language processing, perception, statistical analyzing etc. In recent years the development of AI has taken a fast pace with advanced computing power, vast amount of data available for analysis and other resources. An AI system analyzes the environment to which it’s programmed to, using the maximum data available to minimize the the possible chance of errors. It must analyze the data and learn from it. The more data it receives, the more accurate will be it’s predictions. Here usually the system will perform the task it has been assigned to, like weather forecast, playing chess, automated driving etc. Scientists are trying to develop AI system that are capable of performing based on whatever data is fed, not a programmed domain. When that becomes a reality, it will become an true AI system.

Development

Earlier AI systems used step by step analysis of data and process accordingly. This takes much time for the machine to respond and won’t be classified truly intelligent because we humans usually don’t think that way. To implement the way human think and analyze data is the most difficult part of any AI system. For example we can almost predict what a fellow human being is going to do by viewing his/her facial expressions, body movements etc. For machine, it is not easy to analyze the same. So an AI system must have a component which mimics the human mind and another which correlates the data with computer hardware.

An intelligent system must have the ability to collect data and analyze without having a human to classify them. The system itself must classify and analyze the input it receives. It must implement methods such as regression to study the input pattern and decide what the output should be. Changing the input must change the output also. Scenarios where natural language processing is involved must have these feature to process requests. For example, one can ask a system – Will it rain tomorrow? or What is the climate for a date (say tomorrow’s date)? or Is it okay for outdoor games tomorrow? All these questions pertain to the weather for tomorrow. So an Intelligent system must analyze the question and answer accordingly. There can’t be a preset question in natural language processing.

Machines are becoming more advanced and many system has features like deep learning built-in. Deep Learning systems draw information from various input sources it has and process data. For example in a parking, lot a deep learning camera system can analyze the incoming vehicle and allot parking space according to the vehicle size. This will make parking more efficient. In robotics the same algorithm can be used to pick and stack items without worrying about the size of the package. An intelligent system will adjust the arm size and pick up the item. It can also decide where to place it so that the storage space is used efficiently. Systems like neural networks which resembles the human brain are being implemented to make the AI more human like.

Future

Robotic Process Automation (RPA), Machine Learning, Block Chain Technology, Big data, all these relies upon AI for their development. Natural language processing, Machines playing chess and other strategic games, Self driving cars, Medical diagnosis, Genetic engineering, Weather forecast, Simulators, Image recognition, Targeted advertisements, Prediction of election results by analyzing social media etc. are the contributions of AI. Soon systems will take over many jobs once we thought only humans can do. Artificial Intelligence was part of fiction, especially movies, long before it became reality. Several movies had themes with AI systems and even had stories where the machines controlling the human in a future world.

Credits: Wikipedia