Education, medicine, marketing, business as a whole are being reorganized to work in the digital space. The continuous flow of information in the modern world pushes a person to the development of technology. This helps to interest and retain customers.
Technology develops along with the needs of the audience, sometimes even ahead of them. Fantastic ideas appeared in books and films because the authors found them useful and convenient. Many of these ideas are now reality.
What can already be applied
In the twentieth century there was a boom in futuristic creativity. Fantastic novels, films about space travel and superhumans from the future excited the minds of ordinary people and motivated scientists for new research. Many fictional objects subsequently appeared in real life and were useful, for example:
The bionic arm of Luke Skywalker from the Star Wars saga became the prototype of modern prostheses.
The flip switch from Star Trek is reminiscent of the first flip mobile phones.
Watches with the display of Elroy Jetson from the animated series about the future of the Jetsons are similar to the already familiar smart watches.
Information technologies are designed to make people’s lives more convenient, entertain, expand the boundaries of possibilities. They help to quickly absorb and process huge amounts of information.
Already today, digital tools are available for widespread use, which a few decades ago seemed an unattainable future. The Russian IT segment is actively developing and producing products that can compete with foreign services.
Virtual Reality (VR)
Parallel worlds do not just exist – scientists have learned to create them on their own. With the help of computer technology, a person can immerse himself in a virtual space, every detail of which will look like a real one.
Modern VR glasses and helmets are connected to a computer, gadgets or game consoles. Inside, the devices are equipped with LCD displays and stereoscopic lenses with a viewing angle of 100 degrees, which create the effect of complete presence. Smart sensors monitor the movement of the head and eyes, according to which the picture on the screen changes.
The helmet can be supplemented with “smart” gloves, treadmills, devices for voice control. Full-fledged VR suits are being developed.
All this in the near future will become a way to immerse yourself in virtual worlds – metauniverses with their own laws, hierarchy, structure. Metaverse is already developing, you can even buy a piece of territory in them and develop a business.
Augmented Reality (AR)
AR unites the real and world. The hologram of Leia from Star Wars no longer surprises anyone and does not look like a picture from the future. Virtual images are used at exhibitions, concerts, in the circus. Any user of social networks and various applications can age or rejuvenate themselves with the help of special filters, try on an animal mask or a new hairstyle.
Some large stores provide a service in which the product is superimposed on the space or objects of the user, for example, the sofa from the store on the smartphone screen is inside the room. The Sber ecosystem includes the MagicLook SDK library for integrating the display of 3D augmented reality effects into iOS and Android applications.
The tool allows you to create virtual masks, services for trying on clothes, hairstyles, and accessories. The solution recognizes faces and poses at a distance of up to three meters.
The technology includes several trained neural network models and fast rendering systems.
A few years ago, gamers from all over the world were looking for virtual Pokemon on the streets of their cities. The game works thanks to neural network training. It remembers marks and space parameters, evaluates geotags and markers with digital codes.
You can apply this technology in any application using the naviar SDK tool from Sber. They can bring real benefits, for example, virtual arrows will help a person find the right place in an unfamiliar city or inside a large shopping center.
Modern technologies allow you to communicate with a computer in natural language. If suddenly someone nearby, looking at the smartphone screen, says “Salute, tell me about the weather”, this will not surprise anyone. A pre-scripted phone call can be answered by a voice bot — recording a welcome message on an answering machine may soon become a thing of the past.
Machines are trained to understand spoken language and give answers to questions, and in the future they will be able to conduct a full-fledged dialogue.
Several decades passed from the first VODER synthesizer and spectrographic devices to full-fledged computer synthesis. The computer is able to recreate the voices of any key, timbre, male and female. In synthesized speech, the phonetic norms of the language, the placement of stresses, the pronunciation of the letter Y, complex words with several roots, geographical and professional terms are strictly observed.
Sber’s SaluteSpeech speech service platform contains rich ML models for implementing speech synthesis into any service. Additionally, you can use unique AI models for machine learning, for example:
Golos with a set of speech data;
Brain – to determine the meaning of the user’s phrase.
Voice synthesis can be used for voice acting of content and interfaces, voice control of functions, in telephony and for other purposes.
Narrow AI, or weak artificial intelligence, is already widespread. The computer checks the solvency of applicants at the bank, helps doctors decipher tests, sorts the results of evaluation work in education, and monitors the quality of scientific developments.
The possibilities of AI are much wider than those of humans. It does not forget anything and is able to quickly process large amounts of data. Creating ML models for training neural networks is one of the leading directions of the future in IT.
The developers of the Sber team are actively working on tools that allow them to independently train AI and use already configured neural networks.
Ecosystem products include:
RuGPT-3 is a service that can write texts from scratch according to the given parameters. Just a few inputs are enough for the system to generate ready-made stories, stories, posts for social networks, reviews, descriptions and much more. The neural network is trained on texts from social networks, encyclopedic data, fiction, scientific and business literature.
MLSpace is a platform for developing and implementing ML models. The service allows team development at every stage. This reduces operating time and enhances collaborative management.