Why 2,500 Vibe Coding Entries Show Ordinary People Now Build Their Own Tools
dotey · x · 2026-09-10
Serving as a judge for Weibo's VibeLab AI contest (2,500+ original entries, 240M topic reads), the author reflects on two key shifts in vibe coding:
- Coding's barrier shifted from skill to expression: environment setup and frameworks used to block newcomers; now anyone who can type or speak can direct an AI Agent to build an app. The author himself went from skeptic to daily heavy user.
- Long-tail needs finally get served: traditional software only serves head demands — user ideas get filtered through PMs, designers, developers and QA. Now individuals can build tools for their own "wish it did this" needs directly.
Notable entries like SiaoCut by @机器旁白 illustrate the trend.
Related event: VibeLab Contest Draws 2,500 Works as Vibe Coding Goes Mainstream(2 posts)→
More from coding & agent
- Parallel ML Orchestration Is Hard in Production: Lessons From Scaling a Voice AI Runtime — Slight_Republic_4242 · 2026-09-10
- Open-Source ecommerce-agent-starter Feeds AI Agents Live Product Data via MCP — Loud-Run6206 · 2026-09-10
- User burns through 20x Codex quota in two days with heavy subagent use — Aryvyo · 2026-09-10
- Matt Pocock: The Scariest Part of AI Coding Is Not Feeling Your Strategic Mistakes — mattpocockuk · 2026-09-10
- ScreenContext: free open-source Mac recorder that feeds screen videos to AI agents as context — MarcusSchiesser · 2026-09-10
- Dev: Plan 9's 30-year-old 9P protocol fits the external shape of agent systems — remilouf · 2026-09-10