OpenAI discussed a new classification system for improving Artificial General Intelligence (AGI) at an internal all-hands meeting. Bloomberg announced the details the company shared with its employees. Accordingly, OpenAI proposed the following five levels to track progress towards building AGI:
- Level 1: Chatbots: Artificial intelligence with conversational language capabilities
- Level 2: Reasoners: Artificial intelligence with human-level problem-solving abilities
- Level 3: Intermediaries: Systems that can act on behalf of users
- Level 4: Innovators: Artificial intelligence that can help invention and discovery
- Level 5: Organizations: Artificial intelligence that can perform the work of an entire organization
Currently OpenAI has reached Level 1 and is on the verge of reaching Level 2. The upcoming Reasoners-level artificial intelligence system will be able to perform basic problem-solving tasks with a human power trained at the doctoral level.
During the internal meeting, OpenAI also showed off a research project that uses the GPT-4 model and demonstrates new capabilities involving human-like reasoning. OpenAI will share more information about AGI levels with investors and the public in the near future.
OpenAI is not the first organization to find these levels of AI development. Google Research’s DeepMind team had recently developed a framework for classifying the capabilities and behavior of Artificial General Intelligence (AGI) models.
Follow our site to read more technology news!
