According to a new report in The Information, Microsoft and OpenAI are internally defining a very specific form of artificial general intelligence (AGI) based on the startup's interests. And by this definition, it will take years for OpenAI to get there.
The two companies reportedly signed a deal last year that states OpenAI will only achieve AGI if it develops an AI system that generates at least $100 billion in profits. This is far from the strict technical and philosophical definition of AGI that many would expect.
OpenAI will reportedly lose billions of dollars this year, and the startup has told investors it won't make a profit until 2029.
This is an important detail because once the startup reaches AGI, an ambiguous term that means different things to different people, Microsoft will no longer have access to OpenAI's technology. Some speculate that OpenAI will declare AGI sooner or later to exclude Microsoft, but this agreement means Microsoft could potentially have access to OpenAI's models for more than a decade.
Last week, some people debated whether OpenAI's o3 model was a meaningful step toward AGI. While o3 may perform better than other AI models, it also comes with significant computing costs, which is detrimental to OpenAI and Microsoft's profit-driven definition of AGI.