Final week, Nvidia CEO Jensen Huang made headlines when he instructed podcaster Lex Fridman that AGI—synthetic basic intelligence—had already been achieved.
AGI has lengthy been the final word objective of many synthetic intelligence researchers. That’s been the case despite the fact that there isn’t a universally accepted definition of the time period. It usually means AI that’s as clever as people, however there’s a fierce debate over precisely outline and measure “intelligence.”
On this case, Fridman had provided Huang a really uncommon metric for AGI: May AI begin and develop a know-how enterprise to the purpose the place it was value $1 billion? Fridman requested if Huang thought AGI by this definition may very well be achieved inside the subsequent 5 to twenty years. Huang mentioned he didn’t suppose that period of time was essential. “I think it’s now. I think we’ve achieved AGI,” he mentioned. He then hedged, noting the corporate didn’t essentially need to stay that invaluable. “You said a billion,” Huang instructed Fridman, “and you didn’t say forever.”
Few AI researchers agree with the definition of AGI that Fridman provided Huang, which was each extra particular (an organization value $1 billion), but in addition extra slim than most AGI definitions (which are likely to check with matching an unlimited vary of human cognitive abilities, not all of which could be wanted to construct a profitable enterprise.) However AI researchers additionally disagree with each other over what a greater definition ought to be. The time period stays stubbornly amorphous even though a number of main AI firms, with collective market valuations of greater than $1 trillion, say that AGI is what they’re racing in the direction of. Some pc scientists keep away from utilizing the time period in any respect exactly as a result of they are saying it’s perpetually undefined and unmeasurable. Others say tech firms like utilizing the time period for utterly cynical causes—exactly as a result of it’s ill-defined, it’s simple for firms to construct hype by claiming large strides in the direction of reaching the fabled milestone.
The excitement over Huang’s AGI remarks solely serves to focus on this quandary on the coronary heart of the AI increase.
Making an attempt to measure AGI
Actually, simply days earlier than Fridman dropped his podcast, researchers at Google DeepMind—together with DeepMind cofounder Shane Legg, who first helped popularize the time period AGI within the early 2000s—revealed a brand new analysis paper that proposed a extra scientific method to outline and assess whether or not AI fashions had achieved basic intelligence. The paper, “Measuring Progress Toward AGI: A Cognitive Framework,” attracts on a long time of analysis in psychology, neuroscience, and cognitive science to assemble what its authors name a “Cognitive Taxonomy.”
The taxonomy identifies 10 key cognitive colleges—together with notion, reasoning, reminiscence, studying, consideration, and social cognition—that the researchers argue are important for basic intelligence. The framework then proposes evaluating AI programs throughout all 10 colleges and evaluating their efficiency to a consultant pattern of human adults with at the least the equal of a secondary training.
The paper’s key perception is that immediately’s AI fashions have a “jagged” cognitive profile: They might exceed most people in some areas, like arithmetic or factual recall, whereas dramatically trailing even common folks in others, like studying from expertise, sustaining long-term reminiscences, or understanding social conditions. An AI mannequin would want to at the least match median human efficiency throughout all 10 areas to be thought of AGI, the Google DeepMind researchers recommend.
The researchers additionally introduced a contest with a $200,000 prize pool on the favored machine studying competitors website Kaggle for outdoor researchers to assist construct evaluations for the 5 cognitive colleges the place current benchmark checks are weakest.
The DeepMind paper is simply the most recent in a string of current makes an attempt to place the measurement of intelligence on extra rigorous footing.
Final 12 months, a staff led by Dan Hendrycks on the Middle for AI Security, and that included deep studying pioneer Yoshua Bengio, revealed their very own AGI framework and metrics. That paper additionally divided basic intelligence into 10 separate cognitive domains, drawing on a framework for human intelligence developed by three psychologists—Raymond Cattell, John Horn, and John Carroll—that’s the most empirically validated mannequin of human cognition. It produced “AGI Scores” for current AI fashions; essentially the most succesful system examined, OpenAI’s GPT-5, which was launched in August 2025, scored simply 57%, falling far wanting matching a well-educated grownup throughout all of the cognitive dimensions.
One of the formidable sensible makes an attempt to focus on what immediately’s AI programs nonetheless can’t do is the ARC-AGI benchmark, created by well-known machine studying researcher François Chollet. Chollet’s core argument is that intelligence ought to be measured not by what a system already is aware of, however by how effectively it may study new abilities.
The ARC-AGI benchmark consists of visible puzzle duties involving grids of coloured cells. Every process exhibits a couple of examples of an enter grid being remodeled into an output grid based on a hidden rule, and the test-taker should work out the rule and apply it to a brand new enter. For a human, greedy the sample usually takes seconds. For frontier AI fashions, these puzzles stay surprisingly tough, as a result of they require the type of versatile, summary reasoning—recognizing symmetries, understanding spatial relationships, inferring guidelines from a handful of examples—that present programs battle with.
This month, Chollet and his collaborators launched ARC-AGI-3, the most recent and most demanding model of the benchmark. In contrast to earlier editions, which offered static puzzles, ARC-AGI-3 is interactive: AI brokers should discover novel environments, purchase objectives on the fly, construct adaptable world fashions, and study constantly over a number of steps—skills that come naturally to people however that stay on the frontier of AI analysis.
Taken collectively, these new benchmarks characterize a rising effort inside the AI analysis neighborhood to switch imprecise definitions about AGI with one thing nearer to scientific measurement. However as these researchers are the primary to confess, the issue of defining intelligence is as outdated because the research of considering itself—and has plagued synthetic intelligence as a area from its very earliest days.
Defining intelligence
In 1950, earlier than the time period “artificial intelligence” had even been coined and when mathematicians and electrical engineers have been simply beginning to construct the primary fashionable computer systems, the famed British mathematician and pc pioneer Alan Turing wrestled with the truth that it was extraordinarily tough to formulate a definition of intelligence.
Relatively than trying one, Turing proposed an evaluation he known as “the Imitation Game,” which later turned higher generally known as the Turing Check. It stipulated {that a} machine ought to be thought of clever when it may maintain a basic dialog with an individual, through textual content, and a second human decide, studying the alternate, can’t reliably decide which participant is the machine and which the human. It was, in essence, an “I’ll know it when I see it” method to intelligence.
However the Turing Check quickly proved problematic too. Eliza, a chatbot developed at MIT within the mid-Nineteen Sixties, was designed to imitate a psychotherapist. Most of its responses adopted hard-coded logical guidelines; Eliza usually answered customers with questions equivalent to “Why do you think that is?” or “Tell me more” to cowl up its weak language understanding. And but Eliza fooled some folks into believing it understood them. Eliza got here near passing the Turing Check despite the fact that on virtually each different measure it got here nowhere near human cognitive skills. And, in reality, a extra refined chatbot known as “Eugene Goostman” formally handed a reside Turing Check competitors in 2014, once more with out touching most human cognitive abilities.
As we speak’s massive language fashions converse way more fluently than Eliza ever may, they nonetheless can’t match people throughout the total spectrum of cognitive skills—they hallucinate information, battle with long-horizon planning, and can’t study from expertise the best way an individual does.
In comparison with the Turing Check, the time period “artificial general intelligence” is a comparatively current one. It was first coined in 1997 by Mark Gubrud, then a graduate scholar on the College of Maryland, who used the neologism in a 1997 paper he offered at a convention on nanotechnology. He used the phrase “advanced artificial general intelligence” to explain AI programs that might “rival or surpass the human brain in complexity and speed, that can acquire, manipulate, and reason with general knowledge, and that are usable in essentially any phase of operations where a human intelligence would otherwise be needed.” However the paper rapidly vanished in obscurity.
Then, within the early 2000s, Legg—who would go on to cofound DeepMind—independently coined the identical time period. He was collaborating with pc scientists Ben Goertzel, Cassio Pennachin, and others on a guide about potential methods to create machine studying programs that will have the ability to handle a variety of issues and duties. They wished a time period that will distinguish the ambition of those programs from the slim machine studying algorithms then in vogue, which, as soon as skilled, may solely deal with a single, slim process. Goertzel thought of calling this extra basic AI “real AI” or “strong AI,” however Legg advised “artificial general intelligence” as a substitute, unaware of Gubrud’s earlier utilization. He additionally advised the time period be abbreviated as AGI. This time, AGI took off.
In Goertzel’s guide he outlined AGI as “AI systems that possess a reasonable degree of self-understanding and autonomous self-control, and have the ability to solve a variety of complex problems in a variety of contexts, and to learn to solve new problems that they didn’t know about at their time of creation.”
The definition was helpful for separating work on basic AI programs from slim machine studying ones, however it too contained a good an unhelpful quantity of ambiguity: What did “reasonable degree” imply? Which advanced issues by which contexts counted in the direction of the usual?
Legg would later compound this ambiguity by providing a extra informal definition of AGI that was in some methods narrower (it didn’t speak about self-understanding, for example) however equally imprecise. For example, he instructed The Atlantic’s Nick Thompson final 12 months, “I define an AGI to be an artificial agent that can do the kinds of cognitive things that people can typically do. I see this as the natural minimum bar.” However which issues? And which individuals?
Questions like this have continued to swirl round AGI. Does the time period imply software program that matches the cognitive skills of a median human? Or the skills of the people with the very best IQs? Or the very best knowledgeable in every particular person area of data? The Hendrycks and Bengio analysis paper, for example, defines AGI as matching or exceeding “the cognitive versatility and proficiency of a well-educated adult.” The DeepMind paper proposes measuring in opposition to a consultant pattern of adults. Others have used much less exact formulations.
Including to the confusion, AGI is commonly conflated in public dialogue with an idea AI researchers name “artificial superintelligence,” or ASI—an AI that will be smarter than all people mixed. Most AI researchers think about AGI and ASI to be separate milestones, and really completely different in diploma of sophistication, however within the well-liked creativeness the 2 often blur collectively.
AGI turns into a company objective—and a advertising and marketing slogan
If the educational debate over defining AGI has been lengthy and nuanced, the company world has launched definitions which might be, to place it charitably, idiosyncratic. DeepMind turned the primary firm to make the pursuit of “artificial general intelligence” a enterprise objective. Legg put the phrase on the entrance web page of the corporate’s first marketing strategy when he, Demis Hassabis, and Mustafa Suleyman cofounded the corporate in 2010.
5 years later, OpenAI additionally made constructing AGI its specific mission. Its authentic 2015 founding rules mentioned that the brand new lab—on the time a non-profit—was devoted to making sure “that artificial general intelligence benefits all of humanity.” Three years later, when the lab first arrange a for-profit arm, it revealed a constitution that outlined AGI “as highly autonomous systems that outperform humans at most economically valuable work.” Now, for the primary time, AGI was being measured by monetary metrics, not mere cognitive ones.
And, because it turned out, OpenAI would quickly secretly set a extremely particular monetary threshold for AGI. When Microsoft first invested $1 billion into OpenAI’s for-profit arm in 2019, the tech big’s settlement with the AI startup made it OpenAI’s most well-liked commercialization associate for any AI mannequin the lab developed as much as, however crucially not together with, AGI. On the time, it was reported that the choice of when AGI had been achieved could be on the discretion of OpenAI’s non-profit board.
However, crucially, based on reporting by tech publication The Info in 2024, when Microsoft agreed to speculate an extra $10 billion into OpenAI in 2023, its contract with OpenAI contained a clause that outlined AGI as a know-how that might generate at the least $100 billion in income.
OpenAI is nowhere close to that mark. The corporate has reportedly instructed buyers it made $13 billion in revenues final 12 months, however nonetheless managed to burn by means of $8 billion in money. It doesn’t count on to interrupt even till 2030.
Huang, the Nvidia CEO, is aware of this, simply as he was little doubt absolutely conscious of the social media frenzy and headlines he would generate by saying AGI had been achieved. We all know Huang is aware of this as a result of later in the identical podcast by which he mentioned “AGI is achieved” he additionally mentioned that the favored OpenClaw AI brokers, which may be powered by any of the highest AI fashions from firms equivalent to Anthropic and OpenAI, may by no means replicate Nvidia. “Now, the odds of 100,000 of those agents building Nvidia is zero percent,” he mentioned.
Huang is not only Nvidia’s CEO. He’s additionally the corporate’s founder and the one who has run the corporate for 33 years, piloting it previous near-bankruptcy at one level, to see it now value greater than $4 trillion, making it some of the invaluable firms on the planet. In some ways, Huang is a singular genius. However he’s additionally a really human one. So perhaps we want a brand new commonplace, not AGI however AJI—synthetic Jensen intelligence. When AI reaches that stage, the AI boosters on social media who breathlessly amplified Huang’s AGI declare will actually have one thing to get enthusiastic about.
