Artificial Intelligence: Applications in Everyday Life
What Is Artificial Intelligence?
Artificial Intelligence refers to the branch of computer science which deals with the intelligence exhibited by software and machines.It refers to the design and study of an intelligent agent that could perceive the environment and act accordingly.It is a highly technical and specialized field involving some of the greatest thinkers and scientists working hard to taste success in this relatively new field of research.
The central problems (or goals) of AI research include reasoning, knowledge, planning, learning, communication, perception and the ability to move and manipulate objects.General intelligence (or "strong AI") is still among the field's long term goals. Currently popular approaches include statistical methods, computational intelligence and traditional symbolic AI. There are an enormous number of tools used in AI, including versions of search and mathematical optimization, logic, methods based on probability and economics, and many others.
The field started as an attempt to replicate the brain's unlimited capabilities to train the 'brainless bots'.
Who Wants a Smart Robot ?
Since the machine age was introduced, newer inventions have reduced human effort, but on the other hand, made people lazy too. People prefer having a mechanical slave do things for you rather than engaging their own muscles. This motivation led the modern day scientists to invent newer devices which could 'ease' human life. With an aim to automate the robots, Artificial Intelligence emerged out to be sole answer to all the questions.
In this article, I will briefly talk about what exactly Artificial Intelligence (AI) is, why it is important and how it's slowly merging into each and every sector of our life.
Would you trust a smart bot with your household chores or rather do everything yourself?
What Can a Smart Bot do ?
The capabilities of an autonomous robot are immeasurable. Imagine a bot capable of doing everything you can, repeatedly without any error, all on its own. This may seem like an unreachable goal, but science has progressed a lot to go ahead of just concepts. The introduction of these robots has brought in the idea of an error free world.This will certainly prove to be a huge boon in the medical and industry sector as well. Human error will be completely out of the scene.The whole world will now be seen from a different view. The next few examples will show the true extent to which we have marched ahead in this field.
Banks and Financial System
Banks use artificial intelligence systems to attain assistance for numerous activities based on observed patterns. Some of which include organizing financial operations, investing in stocks, and managing properties. Robots have even beaten human beings in trading challenges.
Financial institutions have long used artificial neural network systems to detect charges or claims outside of the norm, flagging these for human investigation. AI is more S/W related so the difficulty in the situation doesn't matter.AI takes care of everything.
Hospitals and Medicine
Artificial neural networks, which are considered to be highly reliable for environment perception, are used to assist lifesaving equipment, which has proven to be a crucial breakthrough in case of medical negligence. Also, Clinical support systems for medical diagnosis are based essentially on observing and interpreting patterns based on AI. Detection of tumor, which involves Computer Aided Interpretation of data, is one of the most common applications.
The use of autonomous vehicles for transporting goods is not a very old concept. Marine transportation has ensured the usage of standard sized shipping containers on a large scale. Automatic navigation of shipping containers on docks using rails is not uncommon. Automobile industries have adapted conveyor belts also known as 'Assembly Line Production’, these are switching to autonomous ones too.
One of the most systematic transport services, the air transport could barely survive in the absence of Artificial Intelligence. The controllers in the planes which calculate and apply control forces simultaneously have almost eliminated the need for pilots in modern day aircraft. PS aided navigation allows pilots to leave their seats as the 'autopilot' takes care of everything.
During highly sensitive (military) situations, pilots receive assistance from their computers regarding strategies and performing maneuvers, which would be impossible with just the 'human brain'.
In commercial aviation, the ATCs continuously monitor the position of each plane and how the course of each plane is to be altered. Some of the busiest airports of the world have to cater to nearly 2500 flights per day. This makes it an almost impossible task to monitor by the humans. A huge network of neural network along coupled with voice recognition is used to accomplish this mammoth task. In 2003, NASA's Dryden Flight Research Center, and many other companies, created software that could enable a damaged aircraft to continue flight until a safe landing zone can be reached. The software compensates for all the damaged components by relying on the undamaged components. The neural network used in the software proved to be effective and marked a triumph for artificial intelligence.
Speech and Face Recognition
With the advanced uses of image editing software and voice recognition software like Siri, we've been able to interact with these devices as well. Such progress has opened various new aspects of research and input methodology. Analogous to these are face and hand writing recognition software included in laptops.
Computer Games and Gaming Bots
Computer and PC games have developed a lot. There was a time once when Super Mario was considered the most promising gift of technology to the gaming world. Now, we have extremely games like Crisis, Diablo and many more. Playing with computer bots has become one of the most common features in any game. Thanks to Artificial Intelligence, we don’t always need anyone to play against. Computer’s bot is here for play.
Online and Telephone Customer Service
Artificial intelligence is implemented in automated online assistants that can be seen as avatars on web pages. It can avail for enterprises to reduce their operation and training cost. A major underlying technology to such systems is natural language processing.
Similar techniques have been used in answering machines of call centres, such as speech recognition software to allow computers to handle first level of customer support, text mining and natural language processing to allow better customer handling, agent training by automatic mining of best practices from past interactions, support automation and many other technologies to improve agent productivity and customer satisfaction.
Tools for Artificial Intelligence
Artificial Intelligence incorporates some specific tools developed over the years. We’ll talk briefly about some of the tools which have enabled us to achieve the impossible. These methods are based on hardcore mathematics, so I’ll put the concepts in a relatively simple manner for those readers who might be interested in AI and not the mathematics involved in it.
The study of neural networks started with an aim to replicate the thought process of a human brain into a few microchips. It refers to a vast network of data sets which are interconnected and continuously sending data to each other.
According to an experiment conducted in The University of Stanford , object recognition through traditional Computer Based Vision used to be around 87% accurate, whereas those based on neural network were 97% accurate. Though Neural Network is a new field, it has shown great promise in the future.
Classifiers and Statistical Learning Methods
This is the most attractive part where AI has captured the minds of great scientists. Classifier refers to the ability of the robot to classify objects from their inherent properties. For example, you could program a bot to pick a ball if it is black, kick it if it is red and so on. This ability brings in decision making to some extent. Apart from this, robots use statistical algorithms which can match the current event to any of the stored data and take a decision accordingly.
These methods are used primarily to train the robot for a large set of events and decisions to be made are memorized by the bot for future references.
Probabilistic Methods for Uncertain Reasoning
The method of statistical learning will work only if the current event is a replica of any event in the bot’s memory. Many situations in AI need the bot to deal with uncertain, incomplete or completely new information. Algorithms have been devised to take care of such instances with the help of probability and economics.Reasoning is dealt with the help of Bayesian interference algorithm. Bayesian networks are a very general tool that can be used for a large number of problems: learning (using the expectation-maximization algorithm),planning (using decision networks)and perception (using dynamic Bayesian networks)probabilistic algorithms can also be used for filtering, prediction, smoothing and finding explanations for streams of data, helping perception systems to analyze processes that occur over time (e.g., hidden Markov models or Kalman filters).
A key concept from the science of economics is "utility": a measure of how valuable something is to an intelligent agent. Precise mathematical tools have been developed that analyze how an agent can make choices and plan, using decision theory, decision analysis,information value theory. These tools include models such as Markov decision processes, dynamic decision networks, game theory and mechanism design.
It is used mainly for knowledge representation and using this for decision making. Such logic programming has been used in the past for path planning as well. Logic can be of many types:
logic of statements which can be true or false
First Order Logic
allows the use of quantifiers and predicates, and can express facts about objects, their properties, and their relations with each other
version of first-order logic which allows the truth of a statement to be represented as a value between 0 and 1
models uncertainty in a different and more explicit manner than fuzzy-logic: By this method, ignorance can be distinguished from probabilistic statements that an agent makes with high confidence.
Search and Optimization
Search and optimization has been a huge problem for a long time. Search Engines like Google, Bing etc have been improving their search algorithms, which would give better and faster results. The solution, for many problems, is to use "heuristics" or "rules of thumb" that eliminate choices that are unlikely to lead to the goal.
A very different and new solution came into the big picture in the 1990s. This involves starting with an initial guess and refining the guess repeatedly till we reach a point where no further refinements are needed. These algorithms can be visualized as blind hill climbing: we begin the search at a random point on the landscape, and then, by jumps or steps, we keep moving our guess uphill, until we reach the top.
Augmented Reality and Google Glass
One of the most complex machines which have extensively incorporated Artificial Intelligence is 'Google Glass'. You can read about how the concept of 'Augmented Reality' was implemented and how things will seem pretty different for a Google Glass user on my hub - 'All You Need to Know About Google Glass'.