Google Play opens a private lane for vibe-coded apps
Google Play now hosts vibe-coded apps that only you can use. Google announced Tuesday that you can build a personal Android app in Google AI Studio, then install it on your own phone straight through Play — no developer account, no public listing. The app never appears in Play's search or listings and is visible only to its creator. That closes a real gap: until now, putting your AI-built app on an Android device meant publishing it to Play's roughly 3 billion users, which is why most people who vibe-code have stuck to web apps instead of native ones. The feature rolls out today in the U.S. for users 18 and up; sharing with a small circle of friends and family is promised "in the coming months," and if an app outgrows its personal niche there's a path to the public catalog — after the same trust and safety reviews everyone else faces. Google framed the shift as a move from "there's an app for that" to "there's an app for me," and the surrounding announcements lean the same way: an AI game-recommendation engine called Game Finder, comparison answers in the Ask Play assistant, and deeper Gemini integration for store listings. Our take: this is distribution infrastructure for AI-generated software more than a feature drop — Google is quietly building the tier where a million one-person apps can exist without becoming a million public store listings, and it keeps those apps inside Play policy instead of pushing them to sideloading. Personal apps still have to comply with Play's existing rules, which is exactly the point.
Facebook's feed algorithm is now something you can talk to. Meta is bringing its Instagram "Your Algorithm" feature to Facebook, letting users add or drop topics from the three-dot menu on any post — and saying recommendations "respond right away," with the system asking follow-up questions in natural language when it needs more detail. Alongside it, Meta says LLMs now play a larger role in ranking, claiming they can carry "the interests, context, and intent behind natural language, unlocking an algorithm you can actually talk to." It follows Threads' "Dear Algo" experiment, so this is now the pattern across Meta's apps: the recommender stops being a black box you suffer and becomes a settings surface you negotiate with. The open question is whether real-time conversational tuning changes what people actually see, or mostly changes how they feel about it.
What to watch: whether personal, unpublished apps draw regulator attention as a moderation blind spot — apps no one but the author can audit.
Would you trust an app store that lets anyone ship software no one else can see? Tell us in the comments.
