OpenAI develops new AI scale

OpenAI has introduced a new internal system to assess the intelligence levels of its AI tools, aiming to track the progress of its models toward achieving artificial general intelligence (AGI). This initiative comes amid the rapid growth and adoption of artificial intelligence across various sectors.

As AI systems become more sophisticated, the need for a reliable metric to measure their capabilities has become increasingly important. OpenAI’s newly established scale classifies AI sophistication into five distinct levels. At present, ChatGPT is assessed at Level 1, with indications that it is nearing Level 2.

Understanding the Levels

• Level 1: Reflects basic AI capabilities, currently where ChatGPT stands.
• Level 2: Comparable to the problem-solving abilities of a PhD holder.
• Level 3: Capable of autonomously performing tasks on behalf of users.
• Level 4: Demonstrates creative capabilities and generates novel innovations.
• Level 5: Represents the pinnacle of AI achievement, capable of replicating the efforts of an entire organisation and achieving AGI.

OpenAI has defined AGI as “a highly autonomous system surpassing humans in most economically valuable tasks.” This structured approach reflects the company’s commitment to developing tools that aspire to reach AGI.

CEO Sam Altman has previously stated his unwavering dedication to building AGI “no matter the cost.” Recent reports suggest that the forthcoming GPT-5 model will mark a significant leap forward in AI capabilities, with new skills that exhibit human-like reasoning.

The announcement comes on the heels of a growing interest in AI technologies, with many sectors now prioritising AI integration as a standard practice. OpenAI’s efforts to track and advance AI highlight the ongoing evolution of the field and the company’s pivotal role in shaping its future.