How AI is transforming mathematics [Nature]

[mklabgr]

Summary.
What Happened?
Liam Price, a teenager from England with no formal university math training, just cracked a famous math puzzle (Erdős problem #1196) that has been vexing experts since 1966. He didn't do it alone—he used ChatGPT.

Why is this a Big Deal?

It’s not just that a computer solved a hard math problem; it’s how it did it.

• Originality: Usually, AIs just regurgitate things they've already seen. In this case, the AI ignored traditional human conventions and came up with a totally fresh, unexpected strategy to bridge different fields of math.
  
• Logical Reasoning: For a long time, people thought AI large language models (LLMs) would hit a wall with pure logic. Instead, they are starting to show genuine flashes of original "thought."
  

The Future & The Catch

• The Goal: AI teams are scaling up these models quickly. Some experts predict that by 2030, an AI might share a Fields Medal (the Nobel Prize of mathematics).
  
• The Headache: It’s a double-edged sword. While AI is super-charging research, it's also generating a massive influx of "AI slop"—papers that look highly convincing on the surface but take human referees a frustrating amount of time to double-check for hidden hallucinations or errors.
  
  

The Takeaway: The line between human intuition and machine logic is blurring fast. AI is moving past simple calculations and is actively reshaping how we discover new truths about our universe.

ARTICLE PAGE