What is machine learning and why it can deprive you of work

Until recently, programmers had to write complex and very precise instructions, even for computers to perform the simplest tasks.

Languages programming always developed, but the most significant achievement was to simplify the work with the code in this area. Now, computers can not be programmed as before, and adjust so that they trained themselves.

This process, called machine learning, promises to be a true technological breakthrough, and can affect anyone, regardless of the scope of its activities. Therefore, to understand the topic would be useful for each of us.

What is machine learning

Machine learning frees the programmer from having to explain in detail the computer, how to solve the problem. Instead, the computer is taught to find a solution on their own. In fact, the machine learning - this is a very comprehensive application of statistics to find patterns in the data, and based on them the necessary forecasts.

The history of machine learning began in the 1950s, when informatics managed to teach a computer to play checkers. Since then, along with computing power was growing complexity of laws and predictions that the computer is able to recognize and reach, as well as the problems that it can solve.

First the algorithm obtains a set of training data, and then uses them to process requests. For example, you can load in the car a few photos with a description of their contents, such as "in this photo depicts a cat" and "in this photo there is no cat." If you then add the new image computer, it will begin to identify pictures with cats already own.

The algorithm continues to improve. True and false recognition results enter into the database, and treated with every photo software becomes smarter and better and better cope with the task. In essence, this is the training.

Why machine learning - this is important

Now the machine can be safely used in areas which have previously been available only to the person. Although the technology is still far from ideal, the fact that computers are constantly improving. Theoretically, they can grow indefinitely. This is the main idea of ​​machine learning.

Machine learning to see the images and classify them, as in the above example with the photo. They can recognize text and numbers in these images, as well as people and places. And computers are not just written words reveal, but also take into account the context of their use, including positive and negative shades of emotions.

Among other things, the machine can listen to us and respond. Virtual assistants in our smartphones - whether Siri, Cortana and Google Now - embody the breakthroughs in machine processing of natural language, and continue to develop.

In addition, computers are learning to write. machine learning algorithms have generated news articles. They can write about finance and even sports.

Such functions can change all kinds of activities, based on the introduction and classification of data that were previously only by man. If the computer can recognize an image, document, file, or other object, and accurately describe it, it opens up opportunities for automation.

As machine learning is used today

Machine learning algorithms are already able to impress.

Medecision company uses them to calculate the risk factors for various diseases in large settlements. For example, the algorithm has identified eight variables that can be used to conclude a diabetic needs to be hospitalized or not.

After searching for the right product in the online store, you might notice that you can see on the internet advertising of this product for a long time. Such marketing personalization is only the tip of the iceberg. Companies can automatically send e-mails, coupons, offers and display recommendations tailored for each client individually. All this is more accurately pushes the consumer to buy.

Natural language processing are different applications in many areas. For example, it can help replace employees in support services to quickly issue the necessary information to users. In addition, these algorithms to help lawyers in deciphering the complex documentation.

IBM recently interviewedAutomotive 2025: Industry without borders. automotive CEOs. 74% of them expect the emergence of intelligent machines on the road as early as 2025.

Such vehicles will get information about the owner and his surroundings with the help of Internet of things. On the basis of these data, they will be able to change the temperature settings, audio, seat position, and other settings automatically. Intelligent machines will also solve problems themselves, their own drive and make recommendations, taking into account traffic and road conditions.

What to expect in the future from the machine learning

The opportunities offered to us in the future, machine learning, are almost limitless. Here are a few impressive examples.

• Personalized health care system that provides health care to patients individually, taking into account their genetic code and way of life.
• Protective programs that calculate precisely the hackers and malicious software.
• Computerized security systems for airports, stadiums, and other such places, to identify potential threats.
• Self-governing cars, which are oriented in space, minimize the number of traffic jams and accidents on the roads.
• Advanced protection from scams that can protect the money in our accounts.
• Universal translator that will allow us to obtain an accurate and fast translation via smartphones and other smart devices.

Why you should watch out for machine learning

While many of these features will be felt with the advent of new technologies, the majority does not want to understand how it works from the inside. But all of us better stay alert. After all, with all the benefits of further progress will bring tangible consequences for the labor market.

Machine learning based on the ever-growing amount of data, which generates almost every person on Earth, will change the profession. Of course, these innovations will simplify the work of many people, but there are also those whom they lose their jobs. After the algorithms have already answered the letter, interpret medical images, help in trials, analyze the data, and so on.

Machine learning from their own experience, so programmers no longer need to write code for each abnormal situation. This ability to learn, along with the development of robotics and mobile technologies allow computers to handle complex tasks better than ever before.

But what happens to people when they exceed the machine?

According toThe Future of Jobs. World Economic Forum, for the next five years, computers and robots will take five million jobs, which now belong to the people.

Thus, we need to keep track of how machine learning alters the workflow. It does not matter who you are: lawyers, doctors, support staff, a truck driver or someone else. Changes may affect anyone.

The best way to avoid unpleasant surprises when computers begin to select jobs - think preventively and prepare.