The Developer's Work Stack in 2026: AI Instead of Apps

At the bank, we have started holding internal sessions where developers and not only developers show how they work with AI, which apps and technologies make up their work stack, and share advice in a livestream format.

It reminded me that 14 years ago I gave a short interview to Andrey Olesko's Worq project, where he explored how people in different professions worked and which tools they used.

For developers, that really mattered back then, and someone's toolkit could tell you quite a lot about how they worked.

Today, though, I caught myself thinking that most of my work is gradually converging on a single interface: a chat.

Code → chat
Reading → chat
Research → chat
Documents → chat
Analytics → chat
Actions → also chat

Of course, IDEs, Jira, Notion, Figma, and all the other apps have not disappeared. But now you do not necessarily have to open them to find something out.

More than that, interfaces often limit the scale of what you can do with a tool.

An interface gives you a predefined set of workflows. With AI, you are more likely to describe the result you want.

We used to choose apps. Now we are more likely to choose what an AI can access and which models, data, and tools it can use.

That makes a person's ability to work with AI increasingly important: defining the task, providing the right context, and judging whether the result is any good.

I have already written that soft skills are becoming increasingly important, and AI makes this especially clear: sometimes a model cannot understand the task simply because the person has not properly formulated it yet.

If that interview were conducted today, I would be much less interested in which editor, browser, or note-taking app someone uses.

I would much rather ask, “How do you work with AI every day?”