“We’ve achieved AGI” – Jensen Huang declares the main goal of artificial intelligence developers to be meaningless

Creating so-called artificial general intelligence (AGI) has become one of the main goals of some AI developers. For example, OpenAI has made implementing AGI a clear priority. Instead, Nvidia founder Jensen Huang urged not to take the milestone too seriously, saying that in many areas it has actually been surpassed.

Image source: NVIDIA

At the last quarterly report meeting, NVIDIA CEO really stated the following: “For many missions, we can already say we have achieved AGI. I think all of these milestones are meaningless at this point.”. Let us remember that powerful artificial intelligence means that the development level of related software technology can not only be compared with humans, but also surpass humans.

In this context, it sounds interesting to hear an excerpt from a Time interview with OpenAI CEO Sam Altman, in which he said the startup will have AGI for internal use by the end of the year. According to Ultraman’s definition, the AGI in this case would be “A highly autonomous system that outperforms humans at most economically valuable tasks”. Huang did not give his own definition of AGI, but emphasized that artificial intelligence agents have reached a level of development that allows them to work autonomously and “recursively” improve by rerunning tasks over and over again.

“I think three facts are most important to the industry: first, artificial intelligence does productive and useful work; second, artificial intelligence generates profitable tokens; and third, if we have more computing power, we will be able to generate more profitable tokens, which will bring more profits to all services. This is exactly the stage we are in now.” – Nvidia head summary. In an interview with Lex Fridman earlier this year, he also stated that he considered AGI to be a milestone for the industry. According to him, even now, artificial intelligence agents can create programs that are popular and attract billions of users, but then waste all the potential. “The chance of 100,000 agents creating Nvidia now is zero,” ——Huang Renxun guarantees.