In a recent internal meeting, OpenAI outlined a five-stage progression for artificial intelligence (AI), from basic human communication to managing entire organizations. These levels reflect AI’s growing capabilities, with the highest level encompassing the ability to handle comprehensive organizational tasks.
The five levels proposed by OpenAI are as follows:
Chatbots: AI systems like ChatGPT that can engage in human communication.
Reasoners: AI that collaborates with humans to solve problems.
Agents: AI capable of taking independent action.
Innovators: AI that assists with inventing new solutions and ideas.
Organizations: Advanced AI that can manage the operations of an entire organization.
While OpenAI has not specified a timeline for achieving these levels or if further levels might be introduced, Bloomberg reports that OpenAI currently considers itself at Level 1 but anticipates reaching Level 2 soon.
In parallel, Google DeepMind researchers have developed a similar ranking for artificial general intelligence (AGI), ranging from level 0 to level 5:
Level 0: Non-AI computing software.
Level 1: Systems performing on par with or better than untrained humans, indicating the emerging development stage.
Level 2 (Competent): AI systems performing at the median level of professionals.
Level 3 (Expert): AI outperforming 90% of professionals.
Level 4 (Virtuoso): AI surpassing 99% of professionals.
Level 5 (Superhuman): AI that exceeds the capabilities of all humans.
According to a report published in June, the current state of AGI, including OpenAI’s ChatGPT, Google Bard, Meta and Microsoft’s Llama 2, and Google Gemini, is at Level 1. None of these products or models have reached Level 2 yet, remaining in the Emerging stage.
This framework illustrates the ambitious goals of AI development, with significant advancements expected as the technology continues to evolve.
Related topics:
Google Data and Machine Learning: Transforming Insights into Innovations