The world of artificial intelligence (AI) includes many areas in computing, which makes it a complex field.
ACCA introduced a new course called CERTDIF to help finance professionals take a bridge and embrace newer technologies. Here's an article on what is Artificial Intelligence and their fields of application.
An important part of learning about artificial intelligence is understanding what the term AI means. At face value, the term artificial intelligence, or AI, may seem pretty straightforward. But, as you drill down into its meaning, its nuances, and perhaps most importantly what it means to different people, you will realize that AI is a broad term and an even broader discipline with many different meanings.
Learn the nuances of new techonlogies with ACCA CERTDIF. We register students for this course.
Kinds of AI
To begin with, there are two primary kinds of AI, artificial general intelligence, or AGI, and artificial narrow intelligence, or ANI. And this is where the important distinction lies because the average person, when faced with the question of what do you think AI is, will invariably respond with an answer that invokes HAL 9000 or Commander Data. And that's an extreme that while arguably the endgame for AI research is really not what we mean when we discuss the field of AI.
AGI, general intelligence, does embrace the spirit of famous science fiction machines, like HAL and Data – machines that can reason and think like humans. But it's probably better to say that AGI references machines that are as intelligent or more intelligent than humans. We'll get into reasoning and thinking later, but let's generalize the term intelligence for the moment as having the ability to make decisions when faced with problems, when faced with problems, regardless of their complexity.
ANI, narrow intelligence, on the other hand, refers to machines that are designed to perform specific tasks, like recognize a language or drive a car.
So general AI is usually what people think of when talking about AI for a couple of reasons.
First, intelligence in machines, that is machines that can think like humans, is a goal unto itself. Narrow AI can be used for many purposes, like parsing a sentence in English or recognizing a face. But the culmination of all the narrow aspects of AI leads to an end goal of creating machines that act and think like humans.
Second, general AI tends to garner more interest in the public domain anyway. It's the stuff found in sci-fi novels, movies, TV shows, that type of thing. It's interesting. It's controversial. It spurs heated debates and inspires a broad range of disciplines, computer science, philosophy, mathematics, linguistics, engineering, neuroscience, even economics.
General AI involves the direct influence of and the development of several key abilities. Reasoning, the thing that makes us consider factors and develop conclusions. Representing knowledge, drawing upon previous conclusions and using that experience to form new ones. Planning, making strategic choices based on what we know and what we think. Learning from experiences, storing our successes and failures and using those to form responses for future challenges. And processing natural languages, hearing and understanding language in a way that lets us formulate ideas and responses.
Fields and Applications
AI is a popular field to be in. It impacts virtually every industry and, ultimately, every person that interacts with artificial intelligence, which is most of the world today – if you call your bank and interact with an automated phone system that uses voice commands, if you use a fingerprint reader or camera to log into your PC or phone, even a Google search with auto-filled predictions based on your typing – use some pretty sophisticated AI.
But you might be surprised at the number of disciplines that have a stake in AI research. Philosophy is constantly challenged with questions like what constitutes knowledge. And how can thought processes be defined in a series of steps or rules that can be programmed? Psychology is keenly interested in determining how creatures, humans, and other animals form thought processes and actions from thought processes.
Economics concerns itself with determining the best way to make decisions in order to maximize benefits and how to act independently and in concert with, or in spite of, others.
Mathematics focuses on turning decision processes into rules and computations that will result inaccurate results. It also deals with uncertainty, a tricky issue that presents a significant problem in the creation of AI.
Computer engineering, of course, has a great focus on AI from building the actual machines. But more than that, making efficient machines that can leverage the code that goes in the AI. And to implement it with mechanisms like visual systems and robotics, for example.
Neuroscience needs to understand how a human brain accepts, processes, and reacts to stimuli on a neurological level. It even plays a part in visual and audio systems, face and voice recognition, tactile response and touch, other senses like smell and taste. And it strives to build a bridge between the way a brain's neurons work and how that can be translated into digital computing.
And linguistics is keenly interested in how language is compiled, processed, and understood. This is, of course, complicated not just by many human languages. But also idiomatic and vernacular language, different dialects and accents, and perhaps most difficult, nuance.
There are plenty of well-known applications for AI in use today. Probably the most obvious and earliest application of AI was game playing. In part because this was a good way to demonstrate computing abilities and engage users. Modern game play, of course, is an entirely different stratosphere than its early counterparts. But games like Tic Tac Toe and Checkers were good choices for computer games. Because they're relatively simple to program, graphically uncomplicated, and didn't require sophisticated processing on the part of computers.
But chess has been a popular choice. Because it's a very strategic game with a massive amount of possible outcomes. And it was well known. It was quite tempting for programmers to demonstrate what computers could do with chess. And the earliest chess match between a human and a computer occurred in 1956 at Los Alamos Labs. The computer won in 23 moves. But it's important to note that it was playing an amateur. Since then, there have been a number of notable human-computer chess matches.
Although in truth, AI had been a serious science and a looming reality long before that. Voice interfaces like Alexa, Siri, Cortana, and Google's voice-activated AI are popular and quickly growing applications of artificial intelligence, allowing users to ask questions and engage with computer intelligence. Self-driving or autonomous cars seemed like science fiction 10 or 15 years ago. But, in fact, the idea has been around and researched since the 1920s, if you can believe that. But it's only been the past few years that they've become reality, with the first successful test of a self-driving car in public traffic in 2013. Today, major automotive companies are creating autonomous cars. And they're already on the roads in some places. Robotics is another popular application of AI, with robots being used either to automate processes, like in manufacturing facilities. To work in hazardous situations such as collapsed buildings or bomb scares.
Other common applications of AI include logistics planning, determining the best use of resources and routes. For example, utilizing mapping systems in conjunction with GPS to determine the best routes for delivery trucks, identifying construction zones or traffic jams, and rerouting.
Machine learning, the study and development of machines that learn and improve themselves, became a field unto itself, evolving out of AI research. And health care is another field where AI has taken deep roots. Using expert systems for diagnoses and robots that can assist in surgeries, providing precision and greater control.