'Vibe Coding' Doesn't Work
Recent study: "When AI tools are allowed, developers primarily use Cursor Pro, a popular code editor, and Claude 3.5/3.7 Sonnet. Before starting tasks, developers forecast that allowing AI will reduce completion time by 24%. After completing the study, developers estimate that allowing AI reduced completion time by 20%. Surprisingly, we find that allowing AI actually increases completion time by 19%--AI tooling slowed developers down."
Putting aside serious and unsolved legal issues, slop 'coding' or so-called 'Vibe Coding' doesn't work. Well, if it works, then it fares no better - only worse - than properly coding one's project that one comprehends and builds upon existing, predictable libraries (which also get properly updated and tested over time), not slop that can 'hallucinate' (full of defects).
The supposed advantages and promises of so-called 'Vibe Coding' never made sense to most coders because they, unlike their nontechnical managers, know how coding and code generally work.
In a lot of ways, so-called 'Vibe Coding' is already considered vapourware or a passing fad promoted in the media by managers who try to justify mass layoffs, especially ridding companies of "very expensive" software engineers. Those layoffs are not driven by any real potential of slop. Those layoffs are caused by commercial failures and financial problems in those companies. Eventually, no money will be saved. █
