Are the galleys bigger but with fewer rowers? How AI will transform outsourcing
Companies are getting bigger, with fewer people and higher productivity. How AI will change outsourcing.
Companies are getting bigger, with fewer people and higher productivity. How AI will change outsourcing.
Companies are getting bigger, with fewer people and higher productivity. How AI will change outsourcing.
AI is changing business tools, products, and our work habits. Already, 90% of Belarusian IT professionals use AI tools, although most don’t believe artificial intelligence will replace them.
How will AI transform the outsourcing business in the foreseeable future? Will there be fewer developers, and will they become more versatile? Will the trend toward AI solutions increase the number of orders for outsourcing companies, should we expect an outsourcing renaissance, or conversely, will development become so simple that companies will abandon the services of service providers? Imagine hiring a couple of IT specialists who can do everything with AI. Will there be more or fewer outsourcing companies?
What should the average IT professional learn today to fit into the next decade?
We discussed this with two executives from large service companies.
Alexander Khomich, CEO of Andersen, previously thought AI would lead to tectonic shifts in the market, causing many IT professionals to lose their jobs. Now he has changed his mind: the transformation of outsourcing is happening, but without revolutions.
«Clients are interested in AI technologies, but they aren’t rushing to implement them,» he says. «Sometimes transitioning to AI costs more than solving problems the old-fashioned way—for example, by using specialists from India.»
«They won’t. Try, for example, to design a bridge across a major river. You need to consider thousands of details. It’s the same with IT projects. Outsourcing companies solve not only technical issues but also organizational and logical ones.»
«I don’t expect the market to grow—rather, to shrink. This, in my opinion, is caused not so much by artificial intelligence as by the division into geopolitical blocs, market uncertainties (such as 100% US tariffs), and the fact that a large amount of software has already been developed and works well for its intended purpose.
Similarly, after laying railway tracks and lighting streets, those industries reached a plateau—something similar is happening now in IT.»
Outsourcing will consolidate and grow larger. There will be several giants like Accenture—this is already evident from the frenzy with which large companies are buying smaller ones.
And the total number of IT professionals will decrease, probably by about 30%. The market is already saturated, and the trend has faded.
AI will make IT services more expensive and at the same time more comprehensive. A specialist who masters AI tools will command higher rates. But the requirements will also be higher, and they will do more than IT professionals did in the 2010s.
The barrier to entry into the profession will increase. People don’t become doctors or lawyers—even junior ones—after six-month courses. Similarly, IT engineers will only become professionals after 4-5 years of education.
«Learn to look at tasks more broadly: how they interact with other blocks, what risks exist during integration, etc. In other words, a developer should be not only a coder but also somewhat of a business analyst, PM, and DevOps specialist.»
Another executive from a large service company recalls a futuristic lecture that introduced the specialty to first-year students back in 1984. Even then, the professor warned: soon programmers would be unnecessary.
«Describing future courses, the lecturer noted that 'programming' would be among them but advised not to pay special attention to it. The reason was simple: within a few years, they would develop a universal programming language and a universal human-computer interface (the term 'artificial intelligence' wasn’t very popular then, but the essence of expectations was described correctly). Engineers would only need to properly formulate a task and describe it in natural language. A small number of 'advanced programmers' would develop such a language and interface, while we should focus on our core subjects. It wasn’t worth spending valuable time learning programming.»
These predictions proved incorrect in terms of implementation speed, but after ~40 years, they’re starting to materialize. So far, AI handles mainly routine tasks, but there are no inherent limitations to its development.
Our speaker notes that IT companies are already experiencing steady demand for AI solutions and modifying old solutions for the new AI reality. Another source of orders is venture business and startups stimulated by artificial intelligence.
«AI allows venture entrepreneurs to try various recurring ideas with fewer initial investments. We see startups that took the first step using AI, then turn to service companies for qualified services and resources that can develop the idea into a product.»
As long as AI creates demand for product companies by requiring various AI functions, this maintains demand for outsourcing companies. Many companies with stable old solutions have realized they’re not ready for the new AI world and are actively modernizing them, which also supports demand for outsourcing companies.
«AI imposes additional requirements on developers' skills that affect the quality and speed of work. At the same time, writing code is no longer the bottleneck, while product design remains the key task. We’re certainly facing a serious transformation of IT services, but it’s not about complete replacement by artificial intelligence but rather about qualitative development based on the reasonable use of AI.»
With the development of various AI assistants (copilots), engineers should focus more on the «why» and «what for» of their solutions rather than the «how.» AI takes over some of the repetitive, understandable engineering tasks, leaving tasks on the edge of creativity to the engineer.
«I think the proportions of specialists in companies will change dynamically. Some roles will be less in demand, others more. Testing AI solutions is already a huge challenge—an engineer’s brain has to work at 120% throughout the entire workday. A new type of developer is emerging—Vibe Coding engineers are creating software products exclusively with AI assistants, but this trend is just beginning. Complex products and solutions are still made by experienced engineers, while AI can already handle simple tasks.»
The main change is the shift toward Product Mindset. Now clients come to get help making a product, not just implementing some functionality. This requires a proactive approach, understanding the client’s needs and interests. A Product Manager—an engineer who knows how to make products that end users love, who notices the best solutions on the market, and can synthesize new ones based on them—will always be in demand.
«Developers produce IT solutions that become more complex every year. AI increases the complexity of IT solutions even faster than before. Will fewer people armed with AI be able to manage such complex solutions? I don’t think so, because human capabilities are limited.»
«We have good mathematics and engineering education, which is important for mastering AI tools—this helps us remain on the list of promising regions.»
However, our competitors from «warm countries» are also actively mastering AI tools, and this somewhat negates the advantages of education and experience that engineers from Eastern Europe have. As the American saying goes, «God created all people different, but Colonel Colt made them equal.»
«The world is changing rapidly. AI is changing education and methods of obtaining information, allowing comprehensive research to be conducted in minutes or even seconds. At the same time, fundamental knowledge (mathematics, computer science), critical thinking, and the ability to find cause-and-effect relationships remain very important.»
The importance of interdisciplinary knowledge and experience is increasing: the world is becoming more complex, accumulating more knowledge and experience, and it’s important not to fall behind. Practical knowledge and the ability to do and create new things are needed.
As for AI tools, the ability to use them will be necessary everywhere. As our clients tell us: an «AI-first» approach.
Релоцировались? Теперь вы можете комментировать без верификации аккаунта.