AI Card Bao KDOKFMPPVPDEVELOPER.AICARDBAO.COM

Industry News

The 17-million-line Lean proof related to Riemann’s Conjecture has been successfully compiled. The Lean and Mathlib communities have jointly announced that both the proposition and the proof are correct.

"We are fully convinced that Alpöge's proof of the Riemann Hypothesis is correct."

This sentence comes from a joint statement by Lean FRO and the Mathlib Initiative on November 21, just one day after Anthropic mathematician Levent Alpöge released the proof. In yesterday's news, there were two biggest questions at the time: whether the 17 million lines of code could be compiled, and whether it proved the Riemann Hypothesis itself.

Both questions are now answered.

First, compile. Dozens of people pulled down the code base and ran it again. The results were consistent and no errors were reported. Then comes the proposition. The community used two different sets of tools to check the top-level statement and confirmed that it is the same thing as the RiemannHypothesis definition written long ago in Mathlib. There was no manipulation in the definition, nor was it quietly replaced with a weaker version. Finally, the proof was fed into several independent proof checkers written in different programming languages ​​that did not share code with each other, and they all passed.

The most feared thing about formal proof is that something goes wrong in the second step. Lean only guarantees that "starting from these definitions, every step of derivation is legal." It does not care if the definition itself is wrong. This time, the Mathlib community came forward to confirm that the propositions are consistent, which is equivalent to blocking the place where problems are most likely to hide.

MIT mathematician Andrew Sutherland told the media very bluntly: The possibility of something fishy here is about zero.

However, "the machine has tested it" and "people have understood it" are two different things. No one could read the more than 250,000 pages of the manuscript, and Alpöge himself said that even he had trouble following some of the arguments Claude gave later. The next step is clearly stated in the joint statement: the next step is to explain the new ideas in the proof.

A question raised in 1859, the current status is definitely true, but no one can explain why it is true from the beginning. How to distribute the $1 million prize from the Clay Mathematics Institute? According to the rules, it has to wait for publication and academic acceptance. There is currently no news on this part.

Related