Ask HN: Have you observed this trend of building "internal vibe-coding apps"?

Commenters discussed a rising practice of non-engineers using AI to build lightweight internal apps - what some called "vibe-coding" - with examples ranging from a powerline company automating land‑register mailings to an industrial firm creating an internal social network. Several people likened the phenomenon to the Excel era: domain experts can quickly scratch an itch and produce useful, if rough, tools. One commenter said this is inevitable as AI gets cheaper and more available and mentioned an open‑source project, Pando, intended to give these apps a safer home. Others framed it as an arbitrage opportunity between buying software and building in‑house solutions.

Opinion split over whether vibe‑coding is progress or a new source of fragile, risky systems. Supporters argued it raises the floor for many businesses and will deliver practical gains despite rough edges. Critics warned about design deficiencies, poor deployment practices, and governance: management backing non‑engineers to ship code, uncertainty about who maintains broken apps, and the risk that flawed tools drive business‑critical decisions. Several commenters predicted many low‑quality and insecure apps alongside useful ones, with disagreement over whether this is net improvement or merely an expansion of the messy spreadsheet problem.