**[Kimi's Talent Strategy: Not Just Money, But Fame and Recognition for Researchers]** According to monitoring by Dongcha Beating, Chinese AI companies are fiercely competing for research talent, with well-funded tech giants continuously poaching from startups. The *Financial Times* reports that Dark Moon, however, has managed to retain a strong and stable domestic research team. Many founding members became acquainted with Yang Zhilin at Tsinghua University or Carnegie Mellon University. Years of studying and researching together have made the team more stable compared to star-studded lineups recruited on short notice. After the release of DeepSeek R1, Yang Zhilin believed the company had commercialized too early and decided to refocus resources on training models, making the core models open source. This allows researchers to publish their findings, gain personal recognition, and build industry reputation. Dark Moon attracts talent with cutting-edge research and ensures researchers leave their mark through open sourcing. For researchers who value publications, technical influence, and personal reputation, this approach is more appealing than simply joining a tech giant to work on commercial projects. [Original Article Link]