In recent years, there has been a real boom in the IT market due to artificial intelligence. And this is not surprising: modern computing and neural network technologies have reached a level that allows AI systems to solve practical problems that are very difficult for humans, and developers to create innovative services.

In this article, we will talk about how AI is already being applied in various fields and about the most promising technologies of the future.


Optimized handling of patient data

The number of patients around the world is growing daily. Artificial intelligence can help process this amount of data and automate patient information around the world. One example is the OLIVE service created by a neural network – a platform for automating healthcare tasks (especially important in the coronavirus pandemic).

Creation of medicines

Creating new drugs is a time-consuming process because doctors need to run many tests to find the right formula. AI helps them in this. Atomwise is one example of a technology that enables the discovery of new molecules. It is being used to develop new drugs for 27 diseases in collaboration with Harvard and Stanford Universities and pharmaceutical companies.

Cancer diagnostics

Pathologists are using AI to diagnose cancer more accurately. Data on different types of cancer are used to create a predictive model. For example, PathAI technology is used for this.


Money goes digital. The total amount of electronic payments to date is $ 4 trillion and is projected to be over $ 8 trillion by 2024.

All data on such transactions will be processed by neural networks, which will improve the financial industry by the end of 2021. For example, the Dataminr service already collects information from various text sources and presents the user with a graph of important events that may affect his investment.


Digital payments have certain risks. For example, in 2020, Russians lost 2.5 billion rubles due to scammers. Machine learning is perfect for dealing with them. For example, the British company AimBrain uses machine learning to prevent account theft and detect fraudulent accounts. We are waiting for a similar service from Russian developers. 😉 


The algorithms created by the neural network can be used to automate trading. If you provide data on prices, sales volumes, dates, and people’s moods (or “weather” in society), then in 2021 it is possible to create a system that can create forecasts for the market. The algorithm will be able to learn and adapt to changes in real-time to create the most accurate predictions. This project is handled by Kayrros, a company that analyzes data for successful investments. 


A generative adversarial network (GAN for short) is a machine learning model that can simulate a given distribution of data. GANs consist of two neural networks, one of which is trained to generate data, and the other is trained to distinguish simulated data from real data (hence the “adversarial” nature of the model). If you are interested in this topic, you can read our previous articles about GAN, for example, ” Generating Anime Using the StyleGAN Neural Network ” or ” Realistic Landscapes from Drawings .”

Using GAN, you can get datasets of images, faces, cartoon characters, translate images into text and vice versa, create 3D objects, and so on. There are many applications for GANs, but they can bring more than just benefits. One of the most recent uses for GAN applications is deepfakes . 

Deepfakes use artificial intelligence technologies to synthesize images, as a result of which one character seems to be superimposed on another and a “combo” is obtained (yes, you can even make a terminator out of Brad Pitt ).

Deepfakes 2020 differ from their predecessors in higher quality – the technology does not stand still. Here are a few areas in which GANs are used:


NVIDIA has created the GauGAN neural network that turns sketches into real images. The program helps architects to collect building designs from drawings, and game designers – to quickly create locations for games. Try it yourself and see! 😉

Fashion industry

For example, the Russian modeling agency Areola Models has digitized 10 models – today virtual copies of girls take part in filming all over the world and work 24 hours, 7 days a week.


Thanks to GANs and deep fakes, investigative journalists change the appearance of heroes in their reports in order to preserve people’s anonymity. This technique, for example, was used by the HBO channel when creating a documentary film ” Welcome to Chechnya “.


The GAN algorithm is proposed to be used in 2021 in astrophysics in order to get rid of interference and noise when photographing space objects and to obtain high-quality images.

Today, GAN technology remains the subject of heated debate: on the one hand, you can create useful services, and on the other hand, deep fakes can destroy someone’s reputation. For example, in the fall of 2020 in Italy, thousands of women suffered : a bot appeared that could connect the face of any person and a fake naked body in the picture. Also, deepfake videos have been repeatedly used for political or fraudulent purposes. For example, scammers were able to lure 220 thousand euros from the CEO of a British energy company using a deep fake imitation of the voice of his CEO.

It is important to remember that GAN technology itself cannot be regarded as good or bad – the main question is whether a person will use it as a weapon or as a tool.

Reinforcement learning

RL is a field of AI machine learning, where it is studied how the system under test (agent) interacts with the environment (environment) to obtain maximum reward (the response of this environment, that is, reinforcement).

RL can be compared to dog training. Imagine you have adopted a puppy. To teach him something, you need to use a reward system. If the dog listens to you, you give him a treat. This also happens with machine learning.

For example, watch how programmers at OpenAI , a company founded by Elon Musk, showed how agents play hide and seek.

They were not given explicit instructions on how to play. After millions of simulations, agents have learned to interact with the environment on their own:

  • the one who hides has learned to build small forts and barricades;
  • the seeker began to use the ramps to climb the walls and find the hiding.

What is the plus from this technology? It is with the help of it that you can train robots that can radically change our lives.

Everyday life

If you were asked to answer without thinking about the areas in which robots are used, you would probably first imagine the futuristic landscapes against which the androids conquer space. But no, Reinforcement Learning is closer than you might think. Surely everyone has already heard of a robot vacuum cleaner or a robotic lawn mower. 


Already today, devices with special sensors are used, which promptly detect fire hazardous situations and successfully prevent them.


View an example of an academic project of a machine learning course at the University of Rome La Sapienza. It uses a neural network that uses RL to learn how to drive a car using only three sensors on the front of the car. Perhaps in a couple of years we won’t even have to drive the car – who knows!


What do you think about it? Will artificial intelligence technologies be useful, or maybe you see a danger in them? Share your thoughts in the comments.

And if you want to change the world with your project, don’t forget to choose the right infrastructure. Change the future with us!