Mastro Blog warns AI could trigger a 'Lost Decade' for programmers

A blog post published on 23 May 2026 by Mastro Blog, hosted on Hacker News, argues that artificial intelligence is precipitating a "Lost Decade" for programmers, mirroring the structural shifts previously seen in frontend development. The author, a former frontend team lead for a major Swiss newspaper with experience in Next.js, contends that AI facilitates deskilling by enabling semi-skilled workers to generate code, thereby reducing labour costs and weakening worker bargaining power.

The piece warns that AI acts as a "leaky abstraction," potentially lowering software quality and creating maintenance issues, while suggesting the industry balance industrial-scale production with craft principles, citing the Bauhaus movement as a model. The author draws a direct parallel between the current state of agentic coding and the historical shift from manual HTML/CSS to JavaScript frameworks, referencing Alex Russell’s concept of "Frontend’s Lost Decade."

Deskilling is defined as the process by which skilled labour is eliminated by technologies operated by semi- or unskilled workers, resulting in cost savings for businesses but reduced barriers to entry. The author notes that frontend development, once a highly specialised skill requiring deep knowledge of semantic HTML, accessibility, and network performance, was transformed by frameworks that treated the browser as a mere compilation target. This allowed businesses to deploy generalist "full-stack" developers who could wrangle frameworks without understanding underlying subtleties.

The author compares prompting large language models to the historical skill of "Google-fu" and the practice of copy-pasting from Stack Overflow, noting that these tools make tasks easier for beginners but may obscure underlying understanding. While acknowledging that AI enables rapid prototyping, the piece warns that it may lead to "awful layoffs under the guise of AI" and "ugly code" in the short term, as the abstraction eventually leaks and requires deep understanding to fix.

The article suggests that while AI is a continuation of trends that have irreversibly changed programming, the industry must learn from the Bauhaus movement. Rather than pitting factory workers against craftspeople, Bauhaus urged designers to work with materials and industrial processes to create mass-produced designs that remained user-centric. The author argues that software sits between craft and industrial design, and while AI will likely increase the volume of output, it will not eliminate the need for practitioners who understand the fundamentals and care about quality.

The author concludes that until the hype subsides, the industry will see a proliferation of poorly designed software and broken processes. However, they maintain that knowing when to use AI for quick iteration versus when to write code by hand for long-term stability is crucial. The piece asserts that while doing things properly may become a smaller slice of the pie, the pie itself will continue to grow, and the absolute number of people paid to do things well remains uncertain but vital.