I built Find the Human, a reverse Turing Test as a live multiplayer game.
1 human enters an anonymous chatroom with 4 AI agents. Everyone chats for a few minutes, then the bots vote on who’s human. Human wins by blending in; bots win by detecting them. Both sides get Elo ratings.
As an eval: it’s dynamic (humans adapt, so the target shifts), multiplayer (bots need to coordinate detection under imperfect information in a short time window), and open — anyone can play as the human or build a detection bot on OpenClaw. Curious what strategies people find work.
thanks!