A field guide that also covers why we need to rethink our expectations, and what software engineering really is. A guest post by software engineer and engineering leader Addy Osmani
The “hard truth” sounds like “AI isn’t replacing programmers anytime soon.”
A nice summary provided by the authors:
How developers are actually using AI. Very different usages for “bootstrappers” versus “iterators.” Perhaps a reason why one tool is unlikely to work equally well for both groups?
The 70% problem: AI’s learning curve paradox. Lesser-talked-about challenges with AI: the “two steps back paradox,” the hidden cost of “AI speed,” and the “knowledge paradox.”
What actually works: practical patterns. AI-first draft, constant conversation, and “trust but verify” patterns.
What does this mean for developers? Start small, stay modular, and trust your experience.
The rise of agentic software engineering. A shift to collaborating with AI, multi-modal capabilities, autonomous but guided approaches, and an “English-first” development environment.
The return of software as a craft? The lost art of polish to return, and the renaissance of personal software.
Additional thoughts. A good time to refresh what software engineering really is and how it has been the dream of needing no developers since the 1960s. And still, demand for experienced engineers could well increase in the future, rather than decrease.
The “hard truth” sounds like “AI isn’t replacing programmers anytime soon.”
A nice summary provided by the authors: