From Vibe Coding To Vibe Engineering
summary
Argues that LLMs shift front-end work from typing to directing agents and coins vibe engineering as supervising agents rather than hand-coding. Cites Composer One and Cursor as game changers that make agents interactive and faster, warns against giving LLMs to juniors, and shows concrete migrations: moved Benji to Next 16 with TRPC, monorepo and TurboRepo and revived Glink and Zero to Ship while untangling Sizzy’s Electron spaghetti. Claims Composer One delivers in weeks what used to take years and reports productivity gains of completing in two weeks more than a prior year.
AI-written summary · about
chapters (14)
- 0:00Introduction and welcome
- 1:59What to expect from this talk
- 3:01Front-end development state of play
- 6:08Defining vibe coding vs vibe engineering
- 8:39Why vibe engineering matters
- 10:02Tips for successful vibe engineering
- 12:30Who should use vibe coding
- 13:53Overcoming common vibe coding challenges
- 16:40Evolution of AI coding tools
- 18:25Diagnosing pain in the ass developers
- 20:10Should you study computer science now
- 21:33Future of AI in development workflows
- 23:17The role of legacy system maintainers
- 24:39Final thoughts and call to action
related talks
38mNobody Pages the LLM: Engineering Rigour for Vibe Coding
30mVibe Coding That Actually Works
12mNavigating the World of AI Development Without Going Insane
47mPanel about vibe coding
49mVibe Coding for Production
33mWhen Vibe Coding Meets Agent Engineering: The Next Evolution of Software Architecture
30mVibe Coding Full-Stack Dart – Build with Joy Beyond the UI
06h 03mVibe Coder Conference 2026 - Day 2