Code Scientist: Only a third of developers in Russia use AI to create applications

In Russia, only a third of IT specialists use neural networks to create applications, while in the world they are used by about 85% of developers. AI has already become the standard in the context of a growing shortage of qualified personnel in the field of information technology. However, despite the global trend, domestic developers are still cautious about AI tools, considering them rather auxiliary. The reasons are the desire of specialists to dive deeper into the product, as well as doubts about the quality and stability of the code that the neural network writes, experts say.

Creating applications using AI

Over 90% of machine learning specialists already use neural networks in their professional activities. However, only a third of developers are ready to entrust artificial intelligence with writing code and creating applications. This is evidenced by data from a joint study by Avito and the portal for IT specialists Habr, which was reviewed by Izvestia.

Meanwhile, such technologies are being introduced much more actively abroad. According to open source data, more than 75% of IT team leaders worldwide report that their employees are already using AI tools in the process of creating applications. This indicates that neural networks are moving from the category of experimental solutions to the category of working tools, providing faster processes and increased efficiency.

— We are entering an era when each new generation of models is being developed faster than the previous one, opening up opportunities for breakthroughs that previously would have required decades of research. In the future, this may lead to a technological revolution in which artificial intelligence will accelerate its own development," said Konstantin Myagki, Director of Data Science at Avito.

He noted that a developer is turning from a lone programmer into an operator of an entire ecosystem of intelligent solutions, which increases efficiency and partially solves the issue of staff shortages in the industry.

— But security is extremely important here — at Avito we use only our own AI solutions in an isolated loop so that all data remains under our control. We have created an entire ecosystem of internal AI tools because we understand that the future belongs to those who can safely integrate AI into every stage of development," he said.

Do it, Sam: Oracle and OpenAI collaboration will launch a new wave of AI race

Altman and Ellison companies have agreed to build a digital ecosystem of unprecedented scale

Modern trends speak about the imminent transformation of the entire industry. AI assistants not only speed up work, but also open up access to the creation of complex algorithms for those who were previously far from this field, the study notes. In the future, this may lead to the emergence of such technical solutions and architectures that it would be impossible for a person to come up with on their own.

As the MWS AI press service told Izvestia, more than 20% of the code at MTS is already written with the help of AI assistants. Neural networks help not only to write it, but also to supplement it, to look for errors and vulnerabilities.

— At the same time, some developers using AI services admit that they have to fix the generated code in 80% of cases. Others make edits in half of the requests to AI. We expect that in the next couple of years the quality of such tools will grow significantly, and they will become an essential tool for the programmer. Therefore, developers should learn new skills now, the company believes.

Yandex has confirmed that its developers are actively using artificial intelligence technologies in the process of creating software solutions. AI tools help speed up processes, especially for those who already know how to use them effectively. Experts use neural networks for various solutions: web services, APIs, and mobile applications. AI is especially effective in routine tasks such as generating sample code, writing tests and documentation.

"We consider AI to be a powerful tool for increasing productivity, but the ultimate responsibility for quality, safety, and architectural solutions always rests with the developer," said Dmitry Ivanov, head of the Yandex SourceCraft platform.

Why is the Russian IT industry lagging behind the Western ones

The Western market is more focused on solutions that are launched in the startup format, when you need to quickly do something, launch it, get a response from some audience, and then edit, change, improve it in some form, or simply throw it out as a test of some hypothesis and take the next one, the executive believes Vasily Grebennikov, Director of WPP.DIGITAL (which develops digital products for businesses).

— This approach is less common in our country. Today, domestic specialists are trying to delve deeper into the product and work it out better. There is an opinion among them that AI tools are more fiction," the expert said.

According to him, artificial intelligence in their understanding increases the speed of development, but in the long term, the IT community believes that such products cannot be used. In his opinion, more than 50% of developers in Russia use AI.

The lower availability of technologically advanced AI tools in Russia also plays a role, said Margarita Gavrilenko, head of product development at NGR Softlab (a Russian developer of security management systems).

— I think that a more responsible approach to software creation is also important. I don't think that mass use of AI tools and models is the norm. So far, it's more of a desire to try new things, which is not always backed up by a proper critical eye," she said.

That's how AI will be: why Russia doesn't cut IT specialists

Due to the introduction of artificial intelligence in the Russian Federation, "hidden unemployment" is being formed with its contracts and underemployment

At the same time, Anton Nemkin, a member of the State Duma Committee on Information policy and federal coordinator of the Digital Russia party project, added that this opens up an important perspective from a business point of view. Instead of trying to replace specialists, companies can focus on increasing their productivity.

— One developer with an AI assistant is able to complete the amount of work that previously required several employees. This speeds up not only the writing of code, but also the process of testing, documenting, and integrating solutions. This increases competitiveness and allows us to bring products to the market faster," he said.

According to the deputy, IT companies are now facing a shortage of qualified personnel and neural networks are able to smooth out this shortage, namely, generate code fragments, offer optimal solutions, and in some cases accelerate individual processes at times, he added.

In the long term, we can expect that some of the routine tasks will be spent on algorithms, and people will focus on strategic, architectural and creative aspects, emphasized Anton Nemkin. The main question is no longer "whether AI will cope with the tasks," but "how quickly the industry will rebuild the training system for new realities," he concluded.