Nvidia's AGI Claims Exposed as 'Senseless'

Nvidia's CEO Jensen Huang has made conflicting statements about Artificial General Intelligence (AGI), claiming it is already achieved, but then dismissing the milestone as 'senseless'.

Huang made these comments during Nvidia's earnings call on Wednesday, where he stated that for many tasks, the company had 'already achieved AGI'. However, he offered no definition or benchmark to support this claim.

This is not the first time Huang has made such statements. In March, he told Lex Fridman on his podcast that Nvidia had achieved AGI, but retreated when Fridman asked for a more specific definition, suggesting an AI capable of launching and operating a successful tech firm valued above $1 billion.

The lack of clarity around AGI is not unique to Huang or Nvidia. The term has been used loosely by various tech executives, with different definitions and benchmarks, often depending on the company's interests. For example, OpenAI's charter describes AGI as 'highly autonomous systems that outperform humans at most economically valuable work', while Microsoft reportedly negotiated a definition with OpenAI that is based solely on monetary value - systems capable of producing $100 billion in profits.

Huang's focus during the earnings call was not on AGI itself, but on AI shifting beyond basic prompt-and-response into autonomous agents that can pick up new skills and refine themselves recursively. He emphasized that what counts is AI doing productive and useful work, generating profitable tokens, and compute begetting profit.