According to PANews, OpenAI has developed a five-level classification system to track its progress in building artificial intelligence software that surpasses human capabilities. OpenAI CEO Sam Altman expects that general artificial intelligence AGI may be achieved within this decade. Currently, OpenAI is at the first level, AI that can have conversations, and is approaching the second level "Reasoners", which has basic problem-solving capabilities at the human doctoral level.

The third level is called "Agents," which refers to AI systems that can perform multi-day tasks under the guidance of users. The fourth level is AI that can come up with new innovations, and the highest level, the fifth level, is "Organizations," which are AI systems that can perform organizational work. OpenAI plans to share this classification system with investors and the outside world to help people better understand its thinking about AI safety and the future. The classification system was developed by OpenAI executives and the senior leadership team, and will collect feedback from employees, investors, and the board of directors, and may be adjusted over time.