AI Coding Tools Blur Line Between Vibe Coding and Professional Development
Original: Vibe coding and agentic engineering are getting closer than I'd like
Why This Matters
Highlights evolving standards for AI-assisted development in professional software engineering
Developer Simon Willison discusses how AI coding assistants like Claude are making the distinction between casual 'vibe coding' and professional 'agentic engineering' increasingly unclear, as developers rely more on AI-generated code without full review.
In a recent podcast interview, developer Simon Willison explored how AI coding tools are changing software development practices. He previously distinguished between 'vibe coding' - where non-programmers generate code without understanding it - and 'agentic engineering' - where professional developers use AI while maintaining quality standards. However, Willison now admits this distinction is blurring in his own work. As AI tools like Claude Code become more reliable at generating standard code like JSON APIs and SQL queries, he finds himself not reviewing every line of generated code, even for production systems. This creates a moral dilemma about responsibility. Willison compares it to using code from other teams without reviewing every line, suggesting trust in proven systems may justify the approach. The challenge is maintaining professional standards while leveraging increasingly capable AI assistants.