Hacker News developer simulates AI permission fatigue with new web game

A new interactive web application has been published on Hacker News, designed to illustrate the growing issue of "permission fatigue" in artificial intelligence workflows. Titled "Continue? Y/N," the project invites users to simulate the experience of approving or denying commands from an AI coding assistant, specifically referencing the tool Claude Code.

The game presents a rapid-fire series of prompts where participants must decide whether to execute commands by selecting "Y" for yes or "N" for no. The core mechanic aims to demonstrate how users may become desensitised to repeated requests for oversight, potentially leading to careless acceptance of instructions without reading the underlying code or intent.

Discrepancies exist regarding the exact duration of the experience. While the original headline on Hacker News describes the project as a "60-second game," the accompanying summary text characterises it as a "30-second game." Regardless of the precise timing, the tool is intended to be a brief, intense exercise in decision-making under pressure.

Permission fatigue is a recognised phenomenon in the integration of large language models into professional environments. As AI agents gain more autonomy in coding and operational tasks, the volume of approval requests increases. The concern is that human operators, overwhelmed by the frequency of these prompts, may begin to click "approve" reflexively, bypassing critical safety checks.

The project was shared under the "Show HN" category, a section of the platform where developers showcase new projects. It is hosted at https://llmgame.scalex.dev and serves as a narrative device to highlight the need for better design in human-AI interaction, rather than a peer-reviewed study or official report from the creators at Scalex or Anthropic.