OpenAI has developed a classification system showcasing five progressive levels of AI advancement toward achieving superintelligent AI, which it has recently shared with its employees. The system categorizes AI development from basic generative models to advanced artificial general intelligence (AGI) capable of managing organizational tasks.
According to a report, this classification starts at Level 1 with current conversational chatbots and culminates at Level 5 with AGI. OpenAI’s GPT-4 model exhibits human-like reasoning and represents a move toward Level 2 in this framework. Currently, at Level 1, the company may soon progress to higher levels as these capabilities evolve.
The five levels defined are:
- Chatbots: Basic AI that handles conversational tasks.
- Reasoners: AI that can solve problems at a human level.
- Agents: Systems capable of autonomous actions.
- Innovators: AI that contributes to creating new technologies.
- Organisations: AGI that can perform all functions of an organization.
An OpenAI spokesperson revealed these details during an all-hands meeting and indicated plans to share this classification with investors and external parties, though no specific timeline was provided. The move highlights OpenAI’s strides in evolving AI technologies from basic applications to potentially revolutionary AGI capabilities.
This tiered framework reflects significant ambitions, positioning AI as a tool and a potential innovation and operational management leader. As AI advances, it can autonomously execute actions and spark innovations, ultimately achieving the superintelligent capabilities envisioned in Level 5.
