Twin Pics

Twin Pics is a daily web game where players write a prompt to recreate a target AI image and receive a similarity score out of 100. It includes a classroom mode for teachers and aims to build ethical AI literacy in kids.

Twin Pics is a browser game built around a single daily challenge. Every player sees the same target picture, which an AI generated from a hidden prompt. The player writes their own prompt, the service turns it into an image, and an AI judge compares the two pictures and returns a similarity score out of 100. Each player gets one try per day, so the game rewards careful wording over repeated guessing. The site positions itself as a way to teach children ethical AI literacy, and that framing shapes the product. Rather than offering an open-ended image generator, it uses a constrained puzzle: study the picture, work out what the secret prompt might have said, and describe it precisely. The score gives immediate feedback on how clearly a description translates into an image. A short animated sequence (reading the words, framing, developing, scoring) walks through the generation and comparison, and a mascot character called Ditto acts as the player's twin. The homepage also shows a live count of players for the day and a gallery of other players' results, all based on the same challenge. A separate classroom mode targets teachers. A teacher creates a classroom and invites students to join through a dedicated join page. According to the site, students do not need accounts or personal information, which lowers the barrier for school use. A "For teachers" page explains how educators can use the game with a class. Twin Pics sits between casual games and educational tools. It is not a replacement for professional image generators such as Midjourney or DALL-E, and it offers no editing or export workflow of that kind. Its closest comparisons are prompt-writing games and classroom AI literacy activities, where the value lies in practising how to describe things clearly and in understanding how image models interpret language. Families and teachers who want a short, repeatable daily exercise are the natural audience, while users who need production-ready artwork should look elsewhere.