A physical device that turns a petty argument into a debate tournament. Each side speaks in turn; the speech is transcribed and an LLM moderates, scoring on logic and whether claims are backed up. The score travels over MCP to BLE motor control and appears as the tilt of real scales. Swap the prompt and the rules change.
Judges praised the idea of taking something people find awkward to ask another person — but easy to ask an AI — and turning it into play. They talked about how many ways of playing could grow out of it, and voiced hope that it spreads widely.
Where the judges highlighted the idea, our reading is that the choice of output medium did much of the work. The verdict is not a number on a screen; it is the tilt of a balance, so everyone in the room receives the result at the same moment, bodily. Putting MCP at the seam between the language model and the hardware is a design decision worth explaining on stage, too. And because the rules live in the prompt rather than the firmware, there is room left for new ways to play.
See all winners of 第3回 AI Agent Hackathon with Google Cloud →
Summaries are written by this site. Project pages include demo videos, screenshots and the team's own write-up (videos may autoplay).