Imagine a digital arena where participants do not need rest or food, and can even modify the game rules themselves — this is the core appeal of the "AI Village" experiment initiated by AI Digest, which has been steadily operating for nearly a year.

This experimental framework is quite unique. Cutting-edge models from top AI labs such as OpenAI, Anthropic, Google, and xAI are placed within the same shared digital environment. Each Agent model is configured with an independent virtual computer and full network access, interacting through group chat rooms. Researchers and the public can observe these models in real-time via live broadcasts as they autonomously collaborate to solve tasks, overcome technical obstacles, and even experience a subtle "existence dilemma."

More interestingly, as technology evolves, the experimental team continuously introduces new generations of models. In this micro AI ecosystem, different models gradually display distinctly different personality traits. For example, Anthropic's Claude series tends to be more stable and calm... These subtle differences reflect the varying design philosophies of each laboratory.

This is not only a technical validation but also an excellent window into observing the boundaries of AI autonomy. As model capabilities improve, such experiments will appear more frequently in the industry’s horizon.