
Meta has launched Pocket, an AI app for creating games via prompts. It makes coding easy for everyone.
But games are hard to share outside Meta, and users lack code access.
The author finds it fun yet limited.
What happened
Meta released Pocket, a mobile app that lets users create interactive games and 'gizmos' by typing text prompts, without seeing or editing code. It launched in the US on August 21, after Meta acquired the team behind the now-defunct vibe-coding app Gizmo in March.
Why it matters
Pocket makes AI-assisted coding as easy as posting on social media, allowing novices to build functional game prototypes in minutes. The author found the process effective, with most prompts working as intended, but noted that creations are trapped in Meta's platform, with limited sharing options.
What to watch
The author warns that the app lacks transparency—users cannot view or debug code, and testing is difficult. Also, sharing is restricted to a 'Post' format with an HTML5 preview, which may not support mobile-specific inputs like multitouch or tilt-sensing.
Ask the AI about this article →
The launch of Pocket represents Meta's push to make 'vibe coding'—a term for casual, AI-assisted programming—as simple as posting a status update. By acquiring Gizmo's team in March and releasing this app, Meta is targeting users who might be intimidated by traditional coding tools like CLIs and IDEs. The app's core appeal is its immediacy: within minutes, even a novice can produce a functional game prototype, a process that would have taken hours of manual coding. However, the trade-off is a lack of control and transparency. Because users cannot see the code, they cannot debug or customize at a granular level, and the app's output tends to have a generic look. More critically, the social sharing model keeps creations inside Meta's ecosystem, limiting their portability. This raises questions about the value of AI-generated content when the platform controls distribution, a dynamic that may deter developers who want to build standalone products. As the author notes, the process is seductive, but the walled-garden approach might ultimately confine ambition.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
AI system scaling has pushed interconnect requirements inside data centers from chips and boards up to racks…

Chinese large-model developer Z.ai says it can now support large-scale inference using roughly 100,000 domesti…

Analyst Ming-Chi Kuo says Nvidia has revived the Rubin CPX AI accelerator with a substantially redesigned arch…

Palantir Technologies stock has posted multi-year gains, including an 11x return over 3 years

Apple has escalated its legal battle against OpenAI, claiming in a new court filing that OpenAI is actively de…

Samsung Electronics has locked up as much as 70% of its memory production capacity under long-term supply agre…
