Hollywood movies and science fiction novels have caused many to fear the rise of machines resulting from the overdevelopment of artificial intelligence.
Elon Musk says, that AI is our biggest existential threat. Stephen Hawking fears that machines can become Earth’s dominant life form and replace humans. Mark Zuckerberg defends artificial intelligence and believes that within the next 5-10 years it is going to bring major improvements in our lives.
It is difficult to predict how humans will co-exist with the artificial intelligence when it reaches a critical level of development. But nowadays AI is serving medicine, industry, education, business and entertainment, and has great prospects of becoming the major technology of the century.
Consider this article as an introduction. What is artificial intelligence? What is it capable of? Why are machine learning and natural language processing so important nowadays? What are the real use cases of AI? Is the AI apocalypse possible? Patience, my friend — all these questions will be answered in the next few minutes.
Artificial intelligence — what is it actually?
Modern cinematography is full of stereotypes. Any story that involves artificial intelligence involved turns into a real disaster with the rise of the machines. So, most of us see AI as an army of aggressive robots that aim to conquer Earth. This is similar to Stephen Hawking’s prophecy, but even if this scenario becomes a reality, it’s still decades away.
However, the reality is somewhat different. Artificial intelligence is a subfield of computer science, which has the primary aim of enabling computers to do the work that is normally done by humans. To make this a reality, computers must be able to perceive and analyze information, to think and to make decisions, and to learn from their own experience. So, AI is seen not as an enemy, but as an ally and helper of mankind.
Artificial intelligence classification
There are a lot of criteria used to classify AI, however, let’s take a look at the general classifications first.
Strong AI vs. Weak AI
This classification considers how closely a machine imitates human intelligence. A program is called strong AI if it exactly replicates human reasoning and the way humans think. For weak AI, on the contrary, the result is important, not the process. The program should be able to make the same decision a human would in the same situation, but how it makes this decision isn’t that important.
Narrow AI vs. General AI
Programs can also belong to the group of narrow AI if they are designed to complete specific tasks, and aren’t capable of some decisions beyond the tasks required. General AI are programs that possess general intelligence and “thinking” capability, and are able to perform tasks from different domains of human activity.
Currently, strong general AI remains hypothetical, first of all because scientists still don’t understand how the human mind works, therefore, teaching the computer to think like a human is a bit problematic. And, to calm the alarmists, an AI-caused apocalypse isn’t possible until we invent strong general AI.
Artificial intelligence and its derivatives implemented in modern software
Scientists have found that artificial intelligence can be implemented in many ways, and a lot of directions in AI field have appeared. The most well-known of these are machine learning and natural language processing (NLP), and their subsets.
Since the mid of 1900th when Alan Turing first spoke about self-learning computers, the terms artificial intelligence and machine learning have been inseparable. Machine learning is the science of algorithms that enable a program to learn from the data it has to work with and make decisions based on this data. The program improves itself through its own experience without the intervention of developers.
This sounds like magic, but it is already a reality! Tens of applications are now based on machine learning to simulate email database and predict the open and response rate for every email before it is sent or to analyze millions of symptoms to improve disease diagnosis, which is critical in the medical domain.
Being one of the subsets of machine learning, pattern recognition (PR) refers to the ability of a program to recognize regularities in data. It can be supervised, wherein the program learns using labeled “training” data. During unsupervised learning the system can use other algorithms to discover new patterns in data.
Pattern recognition can be used in multiple ways. Many applications nowadays use text, sound or image recognition, which is nothing else as a result of artificial intelligence development. Let’s take a closer look at them as well.
When speaking about computer vision (CV) we mean the ability of a program to recognize and understand digital images and videos, aiming to automate tasks that the human visual ability can cope with. Computer vision has already been implemented in a variety of applications. The most successful use cases are surveillance systems, navigation for autonomous vehicles, species identification, modelling objects or environments, and many others.
Computer vision has already been successfully implemented in sports and gaming.
Hawk-Eye is a system developed to visually track the trajectory of a ball, which is very helpful in sports such as cricket, volleyball, badminton, tennis. Hawk-eye is used in official games to provide a second opinion when the trajectory of a ball cannot be defined in other way.
Kinect is a motion-sensing input device that was developed for Xbox and Windows PC. It enables interaction with a computer or console without using any controllers — only the movements of your own body.
Computer audition (CA) is a branch of pattern recognition and refers to the ability of a program to identify and understand audio, such as natural sounds, music, etc. Speech recognition isn’t the focus of CA. Computer audition has already been implemented in a range of applications. It successfully contributes to sound and music recognition, tool and machinery monitoring and maintenance, traffic cameras controlling, etc.
Recently, enthusiasts at the University of Surrey have even taught a machine to recognize certain sounds or a human voice filtering them from the background noise. The prospects of this technology are very promising!
Natural language processing
Another field of artificial intelligence is natural language processing (NLP). The main subject of study is the interaction between human language and computers. NLP aims to teach a computer to perceive and understand the actual spoken or written language of humans, to derive meaning from it, and react in an intelligent way. NLP can be combined with machine learning, and the program then educates itself by analyzing huge corpuses of texts. The more data that is analyzed, the smarter the system will be.
Are there real life applications for NLP? I bet you have already interacted with it. Apple’s Siri, Amazon’s Alexa, Google’s Home, Microsoft’s Cortana — these are personal assistants that use natural language processing to understand user queries and to act in an appropriate way. In addition, all these assistants improve with their experience of interaction with the user, which means they also use machine learning.
Industries that can benefit from implementing artificial intelligence
Artificial intelligence is a revolutionizing technology that can be implemented in any modern industry to improve the way work is done. Let’s take a look at areas where AI can bring (and in some cases is already bringing) radical improvements.
AI in Healthcare
Healthcare is an area where millions of decisions are made every day. Doctors make decisions regarding diagnosis, efficient treatment, and predicting possible side effects. Their decisions are based on their knowledge and decision-making improves with their experience. Does this sound familiar?
This exactly is the principle machine learning relies on; making decisions based on the experience extracted from data. And healthcare industry can offer an incredible amount of data, which has been gathered over hundreds of years of conscious observation and research. A program that is able to process all these data can potentially become wiser and more efficient than any doctor with a degree. Moreover, the program’s decisions are exact — the human factor is excluded.
One of the first programs that used AI to identify the cause of a disease and recommend an effective treatment was MYCIN, developed in the early 1970s at Stanford University. The program was never used in practice, but researchers proved that it offered the correct therapy in 69% of cases. That was over 40 years ago; can you imagine the power of such a software nowadays?
Artificial intelligence is also highly efficient in monitoring the spread of infections. The system can monitor media files, news reports and social media posts and based on this information make predictions to prevent a pandemic. No, this is not another scene from a Hollywood movie, but a real situation. In 2014 an algorithm spotted the Ebola virus spreading nine days before it was announced by the World Health Organization. It’s hard to estimate how many lives can such an algorithm save in a long run. And what if it constantly improves itself? Fantastic prospects await the healthcare industry with the implementation of artificial intelligence.
AI in education
The expansion of online courses caused a revolution in modern education. Most recognized courses from world’s best universities became accessible to everyone anytime for free. The invaluable knowledge began to spread all over the world, and artificial intelligence is now making online education even more powerful.
Third Space Learning, an online math tutoring platform, improved the quality of its services by launching an AI project based on recognizing learning patterns, and positive and negative teaching techniques.
The system collects data from online lessons, analyzes it, and provides real-time feedback to tutors to help them to improve. The program can, for example, recognize the reaction of the student to a new concept introduced by a tutor and match it with patterns of misunderstanding. The tutor can instantly react to this, offering an alternative explanation of the concept and thereby improve the teaching program.
Such a personal approach is also applicable to study materials. The program is able to recognize the topics the student has problems with, and can offer customized explanations and exercises to fill the gaps more efficiently and without any delay.
AI in engineering
Artificial intelligence is also bringing major improvements in engineering and other areas of industry. A few years ago, a self-driving car seemed to be magic, but now Google, Tesla, and Uber are already testing the first cars that will be able to move around without any human intervention. This is possible due to computer audition and computer vision — two technologies that take into account prevailing sounds and pictures to make responsible decisions about where and how the car should move. Self-driving cars can also reduce the number of accidents on the road, as the program makes only responsible decisions, never exceeds the speed limit, and avoids risky maneuvers. The statistics makes it clear; in the USA 4.2 accidents occur per million miles. Self-driving cars are showing much better results — only 3.2 accidents per million miles. When we take into account that the capabilities of AI increase every single day, we can already predict that self-driving cars will be the safest means of transport.
AI in business management
Artificial intelligence shows impressive results when applied to business management.
For example, AI can take over most of the routine tasks of an HR manager and reduce the costs of the recruitment process. Restless Bandit is already able to screen applicants, find those that match requirements, and even schedule interviews. The algorithm behind the program keeps improving itself with experience, which promises better matches and more efficient recruitment process with time.
With artificial intelligence customer service has also reached new levels. AI enabled chatbots to be programmed to interact in a way that is very similar to humans, but with the advantage of processing huge amounts of information in milliseconds. As a result, customers receive a reliable around-the-clock support. The best thing about this is that the chatbot analyzes all conversations with customers and learns from them.
AI in retail
Retail, especially e-commerce, has already successfully implemented artificial intelligence as well.
Most online stores, for example, recommend products based on the items you have previously purchased, or the those that you have recently viewed.
Virtual shopping assistants is another niche AI has recently filled. Niki.ai, for example, is already helping it’s users to make purchases right in the chat window. The possibilities of Niki.ai are greater than you may imagine.Not only are you able to purchase certain products, but AI also assists you to choose the best option based on your criteria. Fascinating!
Artificial intelligence may seem like a distant future for us, but actually, AI programs are already a part of our reality. When buying stuff or watching movies online, when talking to personal assistants, when taking an online course or chatting with your bank support — every day you interact with artificial intelligence and don’t even notice it. This is a clear sign of the potential that AI based applications have.
Using the power of machines to replace human work will result in a significant improvement in different spheres of our lives. Better healthcare, more efficient education, innovative engineering, predictive sales processes, improved business management — these are just a few use cases for artificial intelligence. And considering the current progress of the AI industry, we are confident in saying that the development of AI will be even more rapid, drastically changing the world we are used to.
Being on the edge of technology is important for every type of business. So, don’t miss your opportunity to become the innovator in your field!
Check out our research on the most popular frameworks for AI development or contact us directly to discuss the prospects of implementing artificial intelligence in your business!