According to a new report from The Information, Microsoft and OpenAI have a very specific internal definition of artificial general intelligence (AGI), which is based on the startup's profits. By this definition, OpenAI is still many years away from achieving this goal.
The two companies reportedly signed an agreement last year that stipulates that OpenAI will achieve AGI only if the artificial intelligence system it develops can generate at least $100 billion in profits. This is a far cry from the strict technical and philosophical definition of AGI that many have come to expect.
OpenAI will reportedly lose billions of dollars this year, and the startup has told investors it won't be profitable until 2029.
This is an important detail because when the startup enters the AGI stage, Microsoft will lose access to OpenAI technology. AGI is a vague term that means something different to everyone. Some speculated that OpenAI would announce AGI early to squeeze out Microsoft, but this agreement means that Microsoft can use OpenAI's model for a decade or more.
Last week, some debated whether OpenAI's o3 model was an important step toward AGI. While o3 may outperform other AI models, it also comes with significant computational costs, which bodes ill for OpenAI and Microsoft's profit-focused definition of AGI.