Terence Tao warns: Humanity is handing thinking over to AI, and danger may arrive soon

Terence Tao warns: Humanity is handing its thinking over to AI, and the danger may arrive soon. Phoenix Tech Download the app for exclusive early access. November 04, 2026 12:59:00 From Beijing. Updated November 04, 2026 13:14:15

Phoenix Tech News, Beijing time, November 4 — Fields Medalist and renowned mathematician Terence Tao recently issued a warning about humanity's growing reliance on artificial intelligence. He worries that as people gradually hand over the process of thinking and problem-solving to AI, humanity is losing its understanding of the problems themselves and of their solutions, and that this trend could soon produce serious consequences in the real world.

These remarks came from Tao's speech at the Science x AI Summit held at the University of California, Los Angeles, on October 16. The Science and Artificial Intelligence Research Foundation (SAIR) recently released the full speech video, titled "Math 2.0," on YouTube, and related clips subsequently drew attention on the overseas social platform X. Autoplay

"We are delegating our thinking to AI, letting it solve problems for us. But at the same time, we are also losing a safeguard — the assurance that the solutions AI gives us are really what we want," Tao said. "I am really very worried that we will soon see the consequences of this in the real world. This is actually a concern of mine outside mathematics."

Tao pointed out that the progress artificial intelligence has made in mathematics is changing the traditional way humanity solves problems. In the past, to tackle complex problems, mathematicians needed long-term research, an understanding of the relevant theory, and the accumulation of new knowledge in the process. But today, AI can quickly provide answers, while humans may not necessarily understand the reasoning behind those answers.

He believes that in pure mathematics, the risk of this change is relatively manageable, because mathematical proofs can usually be verified through formalization tools. But when the same methods are applied to the real world, the situation becomes far more complicated.

In such cases, the solutions AI provides may be completely valid within a mathematical model, yet not meet the real needs of the real world. AI may even find loopholes or boundary conditions in the model, satisfying surface-level metrics while deviating from the goals humans originally hoped to achieve.

Tao stressed that the problem is not AI technology itself, but how humans use AI. He advocates advancing mathematical research into the so-called "Math 2.0" era, in which AI not only helps humans solve problems but also enhances human understanding of problems, helping researchers identify the limitations and potential risks of models.

On October 22, Tao also published an article in the journal Nature, calling on the mathematics community to re-examine its research goals in the AI era. He pointed out that as AI generates mathematical proofs faster than humans can understand and absorb these results, mathematical research is facing new challenges.

Tao believes that AI tools are entirely capable of enhancing human understanding, but this requires changes in research methods and incentive mechanisms. What truly needs to be preserved in the AI era is not only the ability to obtain correct answers, but also humanity's ability to understand, verify, and judge those answers. (Editor/Yu Lei)

For more first-hand news, please download the Phoenix News app and subscribe to Phoenix Tech. For in-depth reporting, please search "Phoenix Tech" on WeChat.