In early 2026, advanced AI models like GPT-5.2 Pro autonomously solved several long-standing Erdős mathematical problems, with their original proofs validated by leading mathematicians such as Terence Tao. The video highlights how this breakthrough signals a transformative era where AI’s rapidly growing capabilities could revolutionize mathematics and many other fields by autonomously generating knowledge and optimizing complex processes.
In early 2026, artificial intelligence reached a significant milestone by autonomously solving several of the famous Erdős problems—open mathematical questions posed by the renowned mathematician Paul Erdős. These problems are known for their difficulty and have resisted solutions from even the brightest human mathematicians for decades. The video highlights that, as of now, AI models like GPT-5.2 Pro have independently solved at least two of these problems (numbers 728 and 397), with the solutions being reviewed and accepted by leading mathematicians such as Terence Tao, who is widely regarded as one of the greatest living mathematicians.
The video emphasizes the credibility of these breakthroughs by referencing Terence Tao’s involvement and endorsement. Tao confirmed that the AI-generated solution to problem 728 was not simply a regurgitation of existing literature but a genuine, original proof developed in the spirit of the problem. This addresses previous skepticism about AI “discoveries” that merely reproduced known results. The video also notes that the AI did not exploit any loopholes in the problem’s phrasing, but rather provided a legitimate solution as intended by Erdős.
Beyond the act of solving problems, the video discusses the broader implications of AI’s rapidly increasing mathematical capabilities. It points out that AI can now not only generate new proofs but also assist in rewriting and formalizing mathematical arguments, making the process of mathematical research more efficient. This is likened to how tools like Photoshop or Excel revolutionized their respective fields by automating tedious tasks and enabling new workflows. The ability to quickly revise and expand mathematical exposition is seen as a transformative aid for human mathematicians.
The speaker draws parallels between the impact of mathematical tools on various industries—such as finance, logistics, and advertising—and the current transformation underway in mathematics due to AI. Just as quantitative analysis and algorithms revolutionized trading, logistics, and marketing, AI’s ability to autonomously solve complex problems and optimize processes is expected to disrupt and enhance many fields. The key difference with AI is its scalability: unlike human experts, AI models can be cloned and run in parallel, working tirelessly and at speeds far beyond human capacity, with access to all existing mathematical literature.
Finally, the video speculates on the far-reaching consequences of this new era. As AI continues to surpass human capabilities in specialized domains like mathematics, it could lead to rapid advancements in science, engineering, medicine, law, and infrastructure. The speaker warns that society may not yet fully grasp the magnitude of this shift, as AI’s ability to autonomously generate knowledge and optimize systems could fundamentally reshape industries and the way we approach problem-solving. The video concludes by urging viewers to pay attention to these developments, as we are entering a period of unprecedented change driven by AI.