Wired tests Google’s AI Playground by making intentionally bad games

Wired reports today that its writer created a batch of intentionally bad video games using Google’s AI Playground to test the tool’s game-making chops. The hands-on indicates the system can spin up basic, playable prototypes but repeatedly falls short on polish and coherence, according to Wired. The critique lands as Google touts consumer AI for creators and as game studios debate how far to lean on generative tools.

Across a long day of prompting, the writer describes frequent glitches, repetitive patterns, and the need for constant human nudging to keep projects moving, Wired says. The piece frames the results as more novelty than utility for now, underscoring questions about how quickly AI can deliver reliable end-to-end game development. Watch for Google to clarify AI Playground’s intended use, ship improvements, or respond to the review, and for developers to share their own results. [Developing. This article will be updated.]