Two publishers and three authors fail to understand what "vibe coding" means

Meaning and Evolution of “Vibe Coding”

• Several commenters say the phrase has already drifted to mean “building software entirely with AI/LLMs,” regardless of the original tweet definition.
• Others argue meaning comes from community consensus, not a single originator; counter-arguments claim the speaker’s definition (or the audience’s interpretation) is what “really” matters.
• Long side-thread on how language changes: overlapping communities, code-switching, no ultimate authority, and analogies to “hacker/cracker,” “algorithm,” and “devops.”
• Some find the whole debate pedantic; others see it as central to avoiding confusion, given AI/LLM hype.

Critiques of the Term and of the Books

• Many find “vibe coding” too vague or meme-like to be a good technical term, suspecting it’s mostly marketing designed to stand out on a book cover.
• Suggested replacements include “chat-oriented programming,” “agentic AI coding,” “apping,” and joking acronyms like AOCS.
• Some mock the idea of teaching people to be “programmers without code,” calling it “vibe writing” and doubting editorial rigor.

What Vibe Coding Looks Like in Practice

• Linked demos show LLM-assisted pair programming and non-programmers using tools like Cursor and Swagger to glue APIs together.
• Observers note users learn real software-development skills (iteration, API usage, CLI habits) while skipping syntax.
• Others report LLMs are still “spectacularly useless” in domains like scientific computing, producing subtle but dangerous errors.

Impact on Programming and Code Quality

• One camp: “vibe coding” is fine for quick prototypes where robustness and maintainability don’t matter; for production, it’s just AI-assisted programming and developers remain responsible.
• Another camp predicts code becoming an invisible backend artifact, with natural language and tests as the main interface, and traditional high-level languages fading in importance.
• Skeptics emphasize nondeterminism, difficulty debugging from prompts, and the enduring need for readable, deterministic code.
• There is strong concern about AI-generated “slop” entering critical systems and about professionals lacking fundamentals under the banner of avoiding “gatekeeping.”

Broader Social/Process Themes

• Comments highlight trend-chasing founders who pivot after every blog post, and managers parroting buzzwords.
• Some propose rigorous text specs and docs as a foundation so future models can regenerate entire AI-written systems.