Four University of California San Diego faculty members argue that artificial general intelligence (AGI) is already here, according to a recent comment published in Nature. The group—Associate Professor of Philosophy Eddy Keming Chen, Professor of Artificial Intelligence, Data Science and Computer Science Mikhail Belkin, Associate Professor of Linguistics and Computer Science Leon Bergen, and Professor of Data Science, Philosophy and Policy David Danks—concluded that large language models (LLMs) now meet reasonable standards for AGI.
Their discussions come after UC San Diego researchers found in March 2025 that GPT-4.5 was judged to be human 73% of the time in a Turing test, surpassing actual humans.
The authors approach the question from different disciplines: philosophy, machine learning, linguistics, and cognitive science. They define general intelligence as broad ability across multiple domains such as mathematics, language, science, practical reasoning, and creative tasks. Depth is defined as strong performance within those domains.
“There is a common misconception that AGI must be perfect — knowing everything, solving every problem — but no individual human can do that,” said Chen. “The debate often conflates general intelligence with superintelligence. The real question is whether LLMs display the flexible, general competence characteristic of human thought. Our conclusion: insofar as individual humans possess general intelligence, current LLMs do too.”
They state that general intelligence does not require perfection or universal mastery; nor must it follow the human brain’s model or exceed human mental capacity through superintelligence.
Rather than relying on a single test for AGI like the Turing Test alone—which measures if machines can converse indistinguishably from humans—the authors assess multiple levels: basic conversation skills; expert-level performance such as PhD problem-solving across domains; and finally revolutionary scientific breakthroughs. They believe current LLMs meet at least the first two tiers.
The team addresses criticisms about LLMs generating false information by pointing out that humans also make errors yet are still considered intelligent.
Some critics argue that LLMs lack bodies or genuine understanding. In response to concerns about embodiment they note examples like physicist Stephen Hawking whose physical limitations did not affect his intelligence: “His physical limitations did not diminish his intelligence; therefore, motor capabilities should not be a prerequisite for intelligence,” the authors suggest.
“This is an emotionally charged topic because it challenges human exceptionalism and our standing as being uniquely intelligent,” said Belkin. “Copernicus displaced humans from the center of the universe, Darwin displaced humans from a privileged place in nature; now we are contending with the prospect that there are more kinds of minds than we had previously entertained.”
Bergen noted ongoing uncertainty around how these systems achieve their abilities: “We have built highly capable systems, but we do not understand why we were successful… This gap in understanding grows more important as the systems grow more capable.”
As AI becomes more autonomous and embedded into daily life—and industry leaders set standards based on profitability rather than pure intelligence—the authors stress responsible design and shared governance. Danks said: “We’re developing AI systems that can dramatically impact the world without being mediated through a human and this raises a host of challenging ethical, societal and psychological questions… Ultimately, we’re innovating because we want something better, and the very idea of better should have ethics and safety baked in.”
Chen added: “I’ve learned so much from this group… UC San Diego’s institutional structure made this collaboration possible—we simply wouldn’t have crossed paths elsewhere. It’s a powerful example of what cross-disciplinary work can achieve when applied to fundamental questions facing humanity.”
