OpenAI says there are 5 'levels' for AI to reach human intelligence — it's already almost at level 2

OpenAI is making impressive strides towards achieving human-level artificial intelligence (AGI) and has developed a five-level system to track its progress. This system ranges from the current conversational AI capabilities to AI that can function at the level of an organization. This detailed framework was shared with OpenAI employees and is set to be disclosed to investors and external stakeholders as well.

Currently, OpenAI is positioned at the first level of this system, which includes the conversational AI that we are familiar with today. However, the company believes it is nearing the second level, termed "Reasoners." At this stage, AI would have the ability to perform basic problem-solving tasks and operate at the intellectual level of a human with a doctorate degree, though it would not have access to external tools.

The third level, called "Agents," describes AI that can undertake various actions on behalf of its user over several days. This means the AI would have the capability to manage multiple complex tasks without constant human intervention. The fourth level, "Innovators," envisions AI that can contribute to the creation of new inventions, showcasing advanced problem-solving and creative abilities. The ultimate goal, the fifth level, is true AGI, where AI can perform at the efficiency and complexity of an entire organization.

Recently, OpenAI demonstrated a research project using GPT-4, indicating that the AI possesses human-like reasoning skills. This demonstration highlights the significant progress OpenAI is making in enhancing the cognitive capabilities of its AI models. The company is continuously refining this five-level system based on feedback from employees, investors, and the company’s board, ensuring that it remains adaptable and aligned with technological advancements.

In a strategic move in May, OpenAI disbanded its “Superalignment” team, which was initially tasked with addressing the existential risks associated with AI. The responsibilities of this team have been distributed across other research efforts within the company, integrating the focus on AI safety and ethical considerations into broader research initiatives.

This systematic approach to developing AGI and the ongoing adjustments based on internal and external feedback demonstrate OpenAI’s commitment to advancing AI technology responsibly and transparently.

For more details, visit this website.