SAN FRANCISCO—Vahid Kazemi, a senior researcher at OpenAI, claimed on X that the company has already achieved artificial general intelligence (AGI). Kazemi’s statement has sparked significant discussion within the AI community.
“In my opinion, we have already achieved AGI, and it’s even more clear with O1. We have not achieved ‘better than any human at any task’ but what we have is ‘better than most humans at most tasks,’” Kazemi posted. He emphasized that the capabilities of large language models (LLMs) go beyond simple pattern recognition, likening the scientific method itself to following a recipe.
This assertion met with mixed reactions. A user, @castillobuiles, countered by saying, “Current language models, including the best ones, represent remarkable achievements in narrow AI, but characterizing them as AGI reflects a significant misunderstanding.”
Meanwhile, OpenAI is in discussions to remove a clause from its corporate structure that would exclude Microsoft from accessing its most advanced models upon achieving AGI. This move aims to unlock billions in future investments. Under current terms, defined as “a highly autonomous system that outperforms humans at most economically valuable work,” Microsoft’s access would be void.
OpenAI’s potential restructuring aims to allow continuous investment from Microsoft. The clause was initially set to protect technology from commercial misuse, preserving ownership with OpenAI’s non-profit board.
Neither Microsoft nor OpenAI immediately responded to requests for comment. The board is still considering options, with no final decision made.
This development follows OpenAI’s substantial $6.6 billion funding round in October, which valued the company at $157 billion. As OpenAI navigates these strategic moves, the debate over AGI’s true nature and implications continues to intensify.
