The latest Artificial Intelligence technologies for optimization and digitization - Dmitrii Rykunov on modern tools in the hands of enterprises and institutions.

According to IDC's forecast, global investments in the field of generative artificial intelligence will reach $143 billion by 2027. The use of AI and machine learning technologies is becoming a globally recognized trend. Dmitrii Rykunov, an internationally recognized expert in the development and implementation of AI solutions in business, discusses the possibilities of integrating AI into the operations of enterprises across various industries.

The new generation of AI models has introduced an entirely new level of creative capabilities in various aspects - text, image, audio. Although video currently lags behind, it is approaching a common level. This means that computers can now generate images, texts, and audio across a wide range of settings that are indistinguishable from real or human-created ones.

Generative AI and big business.

One of the most promising ways to quickly optimize a wide range of workflows is through the use of generative AI (GenAI). Fundamental changes in the field began with the release of two remarkable products - Midjourney (July 2022) and ChatGPT (November 2022). They quickly gained immense popularity and widespread attention among people, followed by interest from businesses. Working on strategies and implementation projects of generative artificial intelligence for large companies in the USA, I can note that with the emergence and popularization of large language models in late 2022, it became very easy to use AI to solve a wide range of simple tasks (such as text classification, data extraction, etc.), enabling the use of AI for tasks that could previously only be performed by humans (such as reviewing insurance claims, content generation, providing knowledge-based advice). This change sparked significant interest in the field and related technologies among many companies. However, with the increasing adoption of AI technologies driven by GenAI, more traditional techniques and consistency in implementation still remain as relevant as ever. Thus, part of the specialist's work includes informing leaders about AI capabilities, conducting diagnostics, developing AI implementation strategies, and implementing AI solutions according to their specific needs.

AI and marketing.

The advertising market has always been one of the most receptive industries to new technologies, and the advent of AI is no exception. In this sphere, it can confidently be said that a professional engaged in ad targeting (basically, the identification of consumer desires) has gained a truly powerful tool in the form of AI. For example, early in my career while working at the well-known IT giant company "Yandex," I developed an approach to build user profiles that would assess the relevance of materials for customers who had not seen them yet. Early experiments showed that the system, when used by a team of specialists, greatly increases the efficiency of marketing communications…

Progress at the Service of Retail Trade

Of course, the application of AI in commerce is not limited to marketing. In retail, entrepreneurs are responding to changes in the world of technology as well. Today, systems are being developed to track and proactively manage the day-to-day operations of retail stores in real-time. For example, while working at McKinsey & Company on one of my first projects, we developed a system to track the remaining stock of goods on store shelves and proactively assign tasks for in-store product replenishment. In developing one of the first such systems, I not only had to plan and develop the technical part of the solution but also significantly adapt logistical processes within the store for its implementation. These systems will significantly optimize the workflow in both traditional small retail points and large shopping centers.

New Technologies and Energy

Not only traditional commercial projects can be a successful field for experimenting with AI technologies. The application of machine learning in the energy industry looks promising. For instance, there are systems that predict heat consumption in district heating systems and Hydronics, facilitating optimal planning and management of energy networks. While such modifications are not frequent, they can be considered a part of the future development of the entire industry.

Forecasting and Finance

The first thing that comes to mind when discussing data analysis in the financial sector is the AI-driven execution of routine tasks, such as document systematization, analysis of credit histories, and overall customer banking transactions. However, the spectrum of AI possibilities in fintech and a broader range of financial institutions goes beyond this. For example, banks can use advanced analytics solutions for forecasting balance sheet movements even on a granular level of groups of customers. Such capabilities can be effectively used in testing and optimizing the overall business strategy of a bank. In the course of my career, I led the development of such a system at one of the largest retail banks in Russia - Raiffeisen Bank.

Neural Networks - the Key to Technical Progress

We see that neural network technologies and machine learning are an absolute trend in a vast number of capital-intensive industries, providing humanity with many new possibilities. For example, thanks to the breakthrough in natural language processing, it was discovered that a sufficiently complex language model can be applied to a wide range of tasks following naturally sounding text instructions without any changes to the model itself (think of the "buzzworthy" ChatGPT). This discovery allowed many more people to utilize AI models simply by using text instructions, without any special technical skills, making their practical value accessible and ready for use for those who eagerly embraced the technology.

In commercial applications, this means that in many cases, you no longer need to create your own solutions; you can use readily available large language models with appropriate prompts. This not only eased development but also allowed the use of AI in cases where it was not possible before due to the lack of enough available examples of solving the given problem that would have been required to create a specialized AI model from scratch, thereby making the technology applicable to a broader range of tasks.

Changes in the world of neural networks have sparked increased interest from many enterprises in the possibilities of AI. Among professionals, there is a significant fear of missing out on opportunities, but the shift is real: the boundaries of the possible have radically changed, and people are eager to seize new valuable opportunities.

(Devdiscourse's journalists were not involved in the production of this article. The facts and opinions appearing in the article do not reflect the views of Devdiscourse and Devdiscourse does not claim any responsibility for the same.)