Nvidia CEO Jensen Huang declared on the company's second-quarter earnings call that Nvidia has reached artificial general intelligence (AGI), a long-sought goal in the field of artificial intelligence. However, Huang quickly qualified his statement, describing the achievement of AGI milestones as "senseless at this point." This apparent contradiction underscores the persistent ambiguity surrounding the definition and measurement of AGI, a concept that has driven billions of dollars in investment across the technology sector.

Huang explained that for Nvidia, "for many tasks, we could say that we have already achieved AGI." He did not provide a precise definition of AGI or specific criteria for this assessment. Instead, Huang emphasized that what matters most is AI's ability to perform "productive and useful work" and generate "profitable tokens." He further elaborated that increased computational power leads to more profitable tokens, which in turn results in greater profit for services, stating, "This is the exact phase where we're at."

This is not the first time Huang has claimed AGI has been achieved. He made a similar statement in March during an appearance on the Lex Fridman podcast, where he said, "I think we've achieved AGI." However, when presented with a definition of AGI as an AI system capable of performing tasks like building a company worth over a billion dollars, Huang retracted the claim, suggesting that the odds of 100,000 AI agents achieving such a feat were "zero percent."

The lack of a consensus on what constitutes AGI is a widely acknowledged issue within the industry. Artificial general intelligence is generally understood as a hypothetical form of AI that possesses human-level cognitive abilities across a wide range of tasks, capable of learning, reasoning, and applying knowledge in novel situations without task-specific programming. Current AI systems, often referred to as narrow AI, are designed for specific functions, such as image recognition or language translation.

In contrast to Huang's pragmatic focus on business outcomes, other prominent figures in the AI field hold different views on AGI. OpenAI CEO Sam Altman, for instance, has stated that his company aims to develop AGI internally by the end of the year, defining it as "highly autonomous systems that outperform humans at most economically valuable work."

Huang's comments were made as Nvidia reported robust financial results for its second quarter, with revenue reaching $96.2 billion, a 106% increase year-over-year. Data center revenue alone contributed $89 billion, showing a 117% increase from the previous year. The company provided guidance for the third quarter anticipating revenue between $105.84 billion and $110.16 billion. Nvidia executives emphasized that the company's growth is currently limited by supply constraints rather than demand.

Huang also highlighted the evolving nature of AI, noting the rise of AI agents that can operate autonomously and improve recursively. He suggested that the compute power required for an AI agent significantly exceeds that of a human user, estimating it to be 15 to 100 times greater depending on the task. He projected that the number of AI agents could soon outnumber human employees by orders of magnitude, envisioning a future with millions of agents working alongside a much smaller human workforce.