23-Year-Old Uses ChatGPT to Crack 60-Year Math Puzzle

Sometimes the biggest breakthroughs come from asking a simple question in a new way.

Liam Price, a 23-year-old without an advanced math degree, recently solved one of the famous Erdős problems by prompting ChatGPT for an answer. These math puzzles, left behind by Hungarian mathematician Paul Erdős, have stumped experts for over 60 years.

What makes this solution special is that the AI did something human mathematicians couldn't. It avoided the conventional approach that everyone had been using for decades.

"There was kind of a standard sequence of moves that everyone who worked on the problem previously started by doing," explained Terence Tao, a mathematician at UCLA who reviews AI-generated math solutions. The AI instead used a well-known formula that no one had thought to apply to this particular problem.

Price shared his solution on erdosproblems.com earlier this month. Unlike many AI-generated math proofs that turn out to be incorrect or just rehash existing solutions, experts confirmed this one is legitimate.

The catch? The AI's raw output still needed human expertise to clean it up. "The raw output of ChatGPT's proof was actually quite poor," said Jared Lichtman, a Stanford mathematician. "So it required an expert to kind of sift through and actually understand what it was trying to say."

Why This Inspires

This discovery shows that fresh perspectives matter more than credentials. Price succeeded where experts failed because he approached the problem without preconceptions about how it "should" be solved.

The collaboration between human insight and AI processing power revealed something new about mathematics itself. "We have discovered a new way to think about large numbers and their anatomy," Tao said.

This breakthrough represents a shift in how we might solve complex problems. Instead of AI replacing human thinking, it's helping us escape the mental ruts we sometimes fall into, even as experts.

Together, humans and AI are discovering solutions that neither could find alone.