🤖 AI Summary
A year-long spot-check of AI “app builders” across company showcases and communities finds surprisingly few bona fide native Android or iOS apps landing on the Play/App Stores. While demos and web-app prototypes abound, real store-listed apps are scarce — and when they do exist they often come from Expo/JS pipelines rather than true platform-native toolchains. The investigator tracked SDK usage across AppGoblin’s 100k+ app/SKD database and sampled vendors: fastshot.ai (YC-backed, sparse showcases), a0.dev (lots of web examples; a handful of July store uploads), replit (many web apps, no clear mobile store releases), and platforms like could.ai, dashwave.io and gobuildmy.app that publish pricing but no public native examples.
Why this matters: it suggests AI coding tools are already useful for rapid prototyping but struggle with the “last mile” — production hardening, app-store packaging, native performance and developer willingness to disclose that an app was “vibe coded.” The ambiguity around “native” (Expo/JS wrappers vs. true native SDKs) complicates claims and SDK-analysis. For the AI/ML community this highlights where tooling and developer workflows need to improve — robust native build pipelines, clearer provenance, and production-grade QA — and points to opportunities for better tracking (AppGoblin offers free SDK scans) and real-world case studies from developers who have pushed prototypes into stores.
Loading comments...
login to comment
loading comments...
no comments yet