Neo Cognition

Neo Cognition is described in 139. 【Agent的综述】和苏煜聊Agent技术史、OpenClaw Moment、边界的消弭和社会的辐射 as Su Yu / 苏煜’s agent research lab. Its short- and medium-term focus is Specialized Intelligence: methods that let agents quickly become expert workers inside a particular domain, profession, organization, or software environment.

The lab’s thesis is that once general model intelligence becomes cheaper and more available, differentiation moves to fast learning inside “small worlds.” That turns Continual Learning, World Models, Persistent Agent Memory, and real production deployment into core research infrastructure rather than afterthoughts.

Connections

• Su Yu / 苏煜 — founder and source of the lab’s thesis in the episode.
• Specialized Intelligence — main technical and product direction described.
• Language Agent, Computer Use Agent, and Universal Digital Agent — broader agent setting for the lab’s work.
• Continual Learning and World Models — learning bottlenecks the lab is organized around.