Too many mathematicians trust that proofs are correct instead of testing them.
Theoretical math tends to be so complicated that even the researchers pushing the field forward can’t quite grasp it all.
It turns out that a great deal of mathematicians simply trust that the foundations of a new discovery are sound, Motherboard reports. If a prominent researcher cites a mathematical proof in their work, others may assume it’s true without actually checking for themselves. And that has Imperial College London mathematician Kevin Buzzard concerned that the whole field is about to go up in flames.
Buzzard is worried that many mathematical proofs are incorrect — but that errors fly under the radar because so many prominent figures simply assume that they hold water.
“I’m suddenly concerned that all of published math is wrong because mathematicians are not checking the details,” he told Motherboard, “and I’ve seen them wrong before.”
To help mathematicians reach sound conclusions without making them dive into thousands of pages of inscrutable operations, Buzzard suggested in his opening talk at a math conference that the field turn to artificial intelligence tools that can do the grunt work for them.
“I think there is a non-zero chance that some of our great castles are built on sand,” read his presentation, per Motherboard. “But I think it’s small.”
READ MORE: Number Theorist Fears All Published Math Is Wrong [Motherboard]
More on math: Google’s Best AI Just Flunked a High School Math Test