What Remains Scarce When Code Becomes Cheap?

Sergey Berezhnoy reflects on his path from frontend engineering to Developer Relations and CTO of Yandex Practicum. Ya.ru and Yandex's question-and-answer service provide examples of why large companies struggle to retire products: invested years, blurred goals, and accountability for consequences.

BEM shows how a local solution can become an ecosystem of methodology, tooling, and a shared component library. Yandex Subbotnik meetups, public talks, and outside criticism helped explain the approach, spread it, and strengthen adoption inside the company.

In the product-technology matrix, products provide context for outcomes while the technical line provides mentoring and growth. One-to-ones and rotation preserve continuity; Staff+ handles cross-cutting concerns by aligning with teams rather than relying on formal authority.

AI is framed as a meta-tool for task framing, decomposition, and redesigning the work cycle, not merely generating code. In education it promises a personal tutor without removing the need for effort: as content and code get cheaper, focus, practice, engineering judgment, critical verification, and accountability remain scarce.