Today, Beijing People announced that after its launch in October, its “Wisdom App Store” ushered in a key update: a closed-loop capability for fixed-point performance scenes has been implemented. The robot can “download an entire performance” through the app store and execute it with one click on the spot. According to Beijing People, this ability combines the host words, greetings, actions, soundtrack, and group photos required for the performance into a complete script in a pre-set order. The script can be saved, copied, and published as a template, and then distributed to different users through app stores. Distribution supports both public and private methods. Operationally, after entering the performance text, the big model understands the intention, matches and combines actions from the action library, and outputs an executable action sequence. The voice, music, and actions are aligned according to the time stamp; the arrangement is completed through visual drag and drop without writing code. At the same time, Beijing People said that “Wisdom Discovery” uses a “one brain with multiple machines, one brain with multiple abilities” architecture. A set of general-purpose brains can be used with multiple configurations and multiple scenarios. The company called this architecture the first in the world.