ITMO Expert Ilya Blagorodov: AI Does Not Remove Developer Responsibility for Code

Ilya Blagorodov, who has spent 30 years in development and serves as an expert for the online master's program "Frontend, Backend Development and AI Solutions" at ITMO in partnership with Yandex Practicum, published a column on Habr about why using AI in work does not free a developer from responsibility for code — and why he himself, despite active use of neural networks, has started programming more rather than less.

Why AI Does Not Replace a Developer

The central thesis of the author: AI assistants — from code autocompletion to full-fledged agents writing entire functions — speed up code writing but do not relieve the developer of the obligation to understand what exactly this code does. Responsibility for a bug, vulnerability, or suboptimal architecture in production falls on the person who signed the commit, not on the model that generated it. This distinguishes the use of AI from simple automation of routine: the higher the share of generated code in a project, the more important it becomes to quickly and accurately review it.

How an IDE Differs from an AI Tool

The author draws a line between the familiar development environment and an AI assistant: an IDE is a tool that suggests syntax, finds compilation errors, and speeds up code navigation, but the developer makes the decisions. An AI tool in this sense might look like a more powerful version of the same thing, but in practice it offers not suggestions, but ready-made solutions — and it's easy to lose the habit of independently thinking through the logic if one doesn't maintain conscious control over it.

• The author — Ilya Blagorodov, 30 years of development experience
• Expert for the online master's program "Frontend, Backend Development and AI Solutions" at ITMO in partnership with Yandex Practicum
• Despite active use of AI in his work, the author claims he has started programming more rather than less

What's Worth Studying to Remain in Demand

The author formulates a paradox familiar to many developers: the faster AI writes code, the more time must be spent checking that code, understanding the boundaries of its applicability, and fitting it into the existing architecture. The column comes out in the context of a master's program that trains specialists at the intersection of frontend, backend, and AI solutions — that is, where the question "what exactly should a developer be able to do in the age of AI assistants" is particularly acute for educational programs.

The author's position echoes what is increasingly heard in the industry: junior developers who rely on AI tools without understanding what's happening "under the hood" risk getting stuck at the level of copying ready-made solutions, while experienced engineers use the same tools as an accelerator — but check, test, and rewrite the result by hand where needed. This is precisely why, according to the author, the use of neural networks in his own work has not reduced but increased the volume of hand-written code: the more generated code passes through a project, the more effort is required to review and refine it.

What This Means

The column fits into a broader industry discussion about how the profession of developer is changing against the backdrop of massive adoption of AI assistants: tools speed up code writing, but the value of a specialist increasingly shifts toward the ability to understand, verify, and take responsibility for what these tools produce.