23-Year-Old Non-Math Major Solves 60-Year-Old Math Mystery with ChatGPT?

It was a peaceful Monday afternoon in April 2026. Liam Price, a 23-year-old living in the UK, was sitting at his computer, chatting with ChatGPT as usual. He had never received professional math training, nor was he a researcher with a PhD Source 3. However, the curious decision he made that afternoon led to an extraordinary result that completely changed the history of mathematics.

The question he entered into ChatGPT was a mystery that genius mathematicians around the world had grappled with for 60 years but ultimately failed to solve. The results were incredibly astounding. ChatGPT produced a perfect logical proof in just 80 minutes, and this result was officially confirmed as “correct” by Terence Tao, a Fields Medalist—often referred to as the “Nobel Prize of Mathematics” Source 3, Source 6.

How on earth was such a magical thing possible? MindTickleBytes presents the full story of this remarkable event, where an amateur without specialized knowledge and artificial intelligence joined hands, in an easy and engaging way.

Why is this important?

Until now, the frontier of mathematics was a sanctuary where only a few highly trained experts could set foot. To solve a decades-old mystery, it was essential to go through the painful process of devouring thousands of related papers and keeping complex formulas in one’s head for years.

However, this event symbolically shows that those solid barriers are crumbling. With the advent of Large Language Models (LLMs) like ChatGPT, anyone who knows how to ask the right questions to AI can now contribute to expanding the limits of human knowledge, even if they don’t know how to handle complex mathematical tools directly Source 2, Source 5.

Experts call this a “new path for mathematical discovery by non-experts” Source 2. In simple terms, the era has come where anyone can surprise the world with just “ideas” and “questions,” using a supercomputer called AI as their assistant.

Easy Understanding: ‘No Multiples Club’ and AI’s Shortcut

The problem solved this time was a conjecture about “Primitive sets” left by Paul Erdős, one of the greatest mathematicians of the 20th century Source 7, Source 12. The name is difficult, but the principle is surprisingly simple.

What is a Primitive Set?

To understand this concept, let’s imagine a “No Multiples Club.” The rules of this club are very clear.

“No number among the club members should be a multiple of another member.”

For example, if 2 is a member, numbers like 4, 6, 8, 10 can never join. If 3, 5, 7 are members, they can coexist peacefully because they are not multiples of each other. Mathematicians call a collection of numbers that follow these rules a “primitive set.” Paul Erdős asked a very tricky question about the unique properties of these number sets, and the answer had been lost in a labyrinth for 60 years Source 2, Source 4.

How did the AI solve it?

Liam Price entered this problem using the latest AI model, GPT-5.4 Pro Source 3. Without even knowing how famous the mystery was, he simply sent a single request (prompt) like, “Prove whether this problem is true or false” Source 3, Source 9.

ChatGPT created a perfect proof through a reasoning process of building logic for about 80 minutes Source 3, Source 11. To use an analogy, until now, all mathematicians walked through a dense forest, hacking away at the undergrowth to find a path. However, the AI launched a drone to look down at the entire forest and found a shortcut in a completely different direction that no one had thought of. Professor Terence Tao marveled at the AI’s innovative approach, saying, “All previous researchers had taken the wrong path from the very first step” Source 6.

Current Status: Birth of ‘Vibe Maths’

After this incident was reported, Western media began calling it “Vibe Maths” Source 1, Source 4. It’s a humorous name suggesting that the problem was solved by exchanging “vibes” and “direction” while conversing with AI, rather than through rigorous formulas.

The current situation is as follows:

1. Verification Complete: World-renowned mathematician Terence Tao confirmed that the AI’s solution is logically flawless Source 3, Source 6.
2. Paper Writing: The AI took charge of the core ideas and logical development, while human experts handled refining it into an academic format and final verification Source 7.
3. New Possibilities: The global mathematical community is watching with a mix of anticipation and concern to see if this method can be applied to other mathematical mysteries Source 12.

What will happen in the future?

Liam Price’s case will not end as a one-time happening. This is because an “AI-for-Erdős craze” is already blowing, with many amateurs using ChatGPT to challenge various mysteries Source 1.

We will soon witness changes such as:

• A world where anyone becomes a mathematician: AI will take care of complex calculations and proof processes, and humans will focus on the role of “what valuable questions to ask” Source 13.
• Collapse of academic boundaries: Collaboration between amateurs and AI can create new breakthroughs in all fields of basic science, including physics and chemistry, as well as mathematics Source 5.
• Evolution of AI reasoning abilities: As seen in the GPT-5.4 Pro model, models that perform high-level logical reasoning beyond simply guessing the next word will become essential partners in scientific research Source 10.

Imagine this. A scene where an artificial intelligence finds a clue to a 100-year-old scientific puzzle based on a single question you tossed out for fun. As Liam Price proved, you could be the protagonist of the next chapter in mathematical history.

AI’s Perspective

MindTickleBytes AI Reporter’s Perspective: If the geniuses of the past solitary mined the secrets of the universe with paper and pencils, we are now in an era where everyone who holds the key called a “question” can approach those secrets. This event shows that AI does not replace human intelligence, but has become a powerful amplifier that infinitely magnifies the destructive power of human “curiosity.” The better the questions we ask AI, the faster the map of human knowledge will expand.

References

1. Amateur armed with ChatGPT ‘vibe maths’ a 60-year-old problem …
2. Amateur Armed with ChatGPT Solves a 60-Year-Old Erdős Problem

3. [Amateur Solves 60-Year Erdős Problem with ChatGPT

   byteiota](https://byteiota.com/amateur-solves-60-year-erdos-problem-with-chatgpt/)

4. Amateur armed with ChatGPT ‘vibe maths’ a 60-year-old problem
5. Amateur Mathematician Solves 60-Year Erdős Problem With ChatGPT
6. 23-Year-Old Amateur Solves 60-Year-Old Math Problem with ChatGPT …
7. Amateur with ChatGPT solves 60-year-old maths problem experts couldn’t …

8. [Amateur armed with ChatGPT solves an Erdős problem

   Hacker News](https://news.ycombinator.com/item?id=47903126)

9. GPT-5.4 Solved a 60-Year Math Problem: What Happened
10. AI-assisted proof may have cracked 60-year-old Erdős problem
11. Amateur armed with ChatGPT solves an Erdős problem
12. AI helps solve a 60-year-old Erdős math puzzle that stumped …