Content
summary Summary

OpenAI has created a five-stage scale to track progress towards AGI. The company is already on the brink of the second stage, called the "reasoner", a source familiar with the matter told Bloomberg.

Ad

OpenAI announced the new classification system at an employee meeting on Tuesday and plans to share it more widely soon, according to Bloomberg's source.

The five levels are:

  • Level 1: Chatbots, natural language
  • Level 2: Reasoners, can apply logic and solve problems at a human level
  • Level 3: Agents, can perform additional actions
  • Level 4: Innovators, can make new inventions
  • Level 5: Can do the work of an entire organization

There is no universally agreed definition of Artificial General Intelligence (AGI), but OpenAI's charter describes it as AI that can surpass humans in most economically valuable tasks, which corresponds to level 5.

Ad
Ad

At the meeting, managers told employees that OpenAI is on the cusp of level 2. These systems could solve problems as well as a human with a Ph.D., without additional tools.

OpenAI reportedly showed a GPT-4-based AI model that demonstrates some new capabilities approaching human-like thinking, possibly enhanced with Q* algorithms.

Microsoft CTO Kevin Scott recently said that next-generation AI could pass PhD qualification tests. OpenAI CTO Mira Murati predicts PhD-level AI will happen in 1.5 years.

According to Microsoft's AI CEO Mustafa Suleyman, it will take until GPT-6, two AI generations from now, before we have reliably acting AI agents, corresponding to level 3.

An Artificial Superintelligence (ASI), distinct from AGI, would have significantly enhanced capabilities. Whether and how AGI and ASI can be achieved remains uncertain and controversial among scientists.

Ad
Ad
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.
Recommendation
Support our independent, free-access reporting. Any contribution helps and secures our future. Support now:
Bank transfer
Summary
  • OpenAI has developed a five-level scale to measure progress toward AGI. The levels range from chatbots, reasoners, and agents to innovators and systems that can do the work of entire organizations.
  • According to one source, OpenAI is on the cusp of level 2, or "reasoners." These systems should be able to solve problems as well as a human with a Ph.D., without the need for additional tools.
  • OpenAI CTO Mira Murati recently predicted PhD-level AI in 1.5 years, while Microsoft AI CEO Mustafa Suleyman believes it will take another two generations, or GPT-6/Level 3, for AI agents to act reliably.
Sources
Online journalist Matthias is the co-founder and publisher of THE DECODER. He believes that artificial intelligence will fundamentally change the relationship between humans and computers.
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.