RUUZEN
AI

AI accelerates development, but programmers fear losing skills to solve complex tasks

AI tools speed up code writing and testing, but some developers miss the old way of working: solving complex problems and manually debugging errors.

20 Nov, 11:19Read in Russian

Programmers are using AI for writing, debugging, and testing code. In a 2026 Stack Overflow survey, 83% of developers reported using at least one such tool.

Gemma Rogers, a software developer from the UK, shared that previously, finding an error could take days or weeks and brought a sense of accomplishment. Now, some tasks are resolved with just a few AI queries. According to her, this may lead to the loss of skills for independently identifying causes and solving problems—skills that will be necessary if AI services or the internet stop working.

The scale of work is also changing: James Broadhead, CTO of Taktile, noted that instead of creating one prototype in three weeks, an engineer can now produce ten in a day. The key skill has become selecting truly successful options and verifying AI-generated code. Joshua Wöhle, CEO of Mindstone, believes the technology widens the gap between those who can use it effectively and the rest.

However, not all developers see only downsides in these changes. Chris Sees, CEO of Hoxton Mix, compares AI to a junior engineer: it can be delegated routine tasks, but the results require supervision. Developers also worry that newcomers will lose debugging practice—it is on such tasks that foundational skills are formed.