The short version
- Nvidia CEO Jensen Huang stated that his company has achieved artificial general intelligence during an earnings call, only to immediately characterize the concept as senseless and arbitrary.
- The tech industry lacks a unified definition for AGI, with major players using vague or financially driven metrics rather than standardized technical benchmarks to gauge progress.
- Executives are increasingly pivoting away from the term AGI toward proprietary labels like personal superintelligence or useful general intelligence while emphasizing profit generation over capability milestones.
Nvidia CEO Jensen Huang announced during a recent earnings conference that his company has effectively achieved artificial general intelligence, a milestone long considered the ultimate objective of the technology sector. However, he almost immediately undercut this declaration by describing the concept as senseless and arbitrary. This contradictory stance underscores a broader confusion within the industry regarding what constitutes true machine intelligence and how such a state should be measured or recognized.
Huang did not provide a specific technical definition or benchmark to support his claim that Nvidia has reached this threshold. Instead, he suggested that for many tasks, the company’s systems already operate at a level that could be classified as AGI. He emphasized that the current phase of development is characterized by AI moving beyond simple prompt responses to autonomous agents capable of recursive learning and self-improvement. The primary metric of success, according to Huang, is no longer theoretical capability but the generation of profitable tokens and productive work.
This is not the first instance where Huang has claimed that AGI has been realized. In a March appearance on the Lex Fridman podcast, he similarly stated that artificial general intelligence had already been achieved. When pressed for clarification during that interview, he did not offer a precise definition. The host proposed a specific benchmark involving an AI system’s ability to start and run a billion-dollar tech company, to which Huang responded that the likelihood of such agents building Nvidia was zero percent.
The ambiguity surrounding AGI is not unique to Nvidia. OpenAI, which was founded with the explicit mission of developing safe artificial general intelligence, defines it in its charter as highly autonomous systems that outperform humans at most economically valuable work. Sam Altman, OpenAI’s chief executive, has previously acknowledged that the term is not particularly useful for measurement purposes. Despite this, internal estimates suggest the company is significantly close to its own undefined target, with leadership predicting a breakthrough by the end of the year.
Complicating the landscape further is a separate, financially driven definition reportedly established between OpenAI and Microsoft. Under this framework, AGI is characterized by systems capable of generating at least one hundred billion dollars in profits. This shift from technical capability to financial output reflects a growing trend among tech leaders to define success through economic metrics rather than cognitive benchmarks. The lack of a universally agreed-upon definition for intelligence itself makes any standard for general intelligence inherently subjective.
Other industry leaders have similarly struggled with the terminology, often resorting to alternative phrases that blur the lines between different levels of AI capability. Anthropic CEO Dario Amodei has described AGI as an imprecise marketing term, preferring to discuss powerful AI instead. Meta refers to its goals as personal superintelligence, while Microsoft uses the term humanist superintelligence. Amazon has introduced the concept of useful general intelligence, and Google DeepMind’s Demis Hassabis speaks of arriving at the foothills of the singularity.
The proliferation of these varied terms highlights the difficulty in establishing a concrete standard for advanced AI. Even former OpenAI cofounder Ilya Sutskever, who reportedly led employees in chants celebrating AGI, now runs a company focused on safe superintelligence. The rebranding efforts have not clarified the underlying technology or its capabilities. Instead, they reflect a strategic move to maintain hype and investment interest while avoiding the pitfalls of unmeasurable promises.
As long as the definition of artificial general intelligence remains poorly defined and inconsistently applied, claims of achieving it will remain largely symbolic. Tech executives continue to use the term to signal progress and attract capital, even as they admit its lack of utility. The focus is shifting toward practical applications and revenue generation, with less emphasis on theoretical milestones. Whether a true AGI will eventually emerge remains uncertain, but the current discourse suggests that the industry is more interested in the economic potential of AI than in defining its intellectual limits.
The implications of this ambiguity extend beyond marketing. Without clear benchmarks, it becomes difficult for regulators, researchers, and the public to assess the risks and benefits of increasingly autonomous systems. The reliance on financial metrics as proxies for intelligence raises questions about the alignment of AI development with broader societal goals. As companies continue to lean into the current phase of rapid deployment, the lack of consensus on what constitutes general intelligence may hinder meaningful oversight and evaluation.
Looking ahead, the industry is likely to continue using AGI as a flexible concept that can be adapted to fit various narratives. Huang’s dismissal of the term as senseless does not necessarily indicate a retreat from ambitious goals but rather a pragmatic acknowledgment of its limitations. The race for AI dominance will likely proceed without a clear finish line, driven instead by competitive pressures and the pursuit of profitable applications. The true measure of progress may ultimately be determined by the tangible impact of these systems on industries and daily life, rather than by abstract definitions.
Sources behind this briefing
Go to the original reporting
- The Verge↗Jensen Huang says Nvidia achieved AGI, again — not that it matters
