Putting One Kimi Among Four Claudes, Can the Claudes Identify Kimi?

A social experiment titled "Putting One Kimi Among Four Claudes, Can the Claudes Identify Kimi?" was conducted to see if AI agents could distinguish a different model. The experiment involved one Kimi AI and four Claude AIs, with their responses to various prompts recorded. The results, including their explanations for color choices and preferences for coffee companions, were documented to determine if the Claude agents could identify the Kimi agent among them.

This experiment is notable for being the first to test whether AI agents can identify a different model among themselves, unlike prior studies focusing on individual AI capabilities.