Maryna Viazovska, a mathematician, taking the groundbreaking ideas of others a step further, found the solution to the sphere-packing problem in 2016. This dazzlingly intricate math problem attempts to figure out how compactly the same-shaped spheres can be packed into n-dimensional space. Viazovska’s was still a remarkable breakthrough. Her work showed that the symmetric arrangement that we call E8 is the best packing arrangement in eight dimensions. Furthermore, she worked with others to prove that the Leech lattice is indeed the best possible solution in 24 dimensions.
The story of AI’s impact on the recent push towards more credible verification of these proofs is a fascinating one. Sidharth Hariharan and his team aimed to verify sphere-packing proofs. They went above and beyond to write an easily-understood “blueprint” of the eight-dimensional proof. Their efforts eventually led to an AI, called Gauss, that autoformalized the eight-dimensional sphere-packing proof in only five days. Unbelievably, Gauss found and fixed a typo in the printed article on the eight-dimensional case.
Advancements in AI Proof Verification
Those changes weren’t limited to the first version of Gauss.Math, Inc. significantly improved this AI and released a new reasoning agent, Gauss. In a stunning turnaround of just two weeks, Gauss equally autoformalized Viazovska’s 24-dimensional sphere-packing proof. This strange news story was yet another chapter in the long and tumultuous saga of the mathematical proof.
Jesse Han, a critical player in this project, shared that it gets complicated when you get to the 24-dimensional case.
“And it was actually significantly more involved than the 8-dimensional case, because there was a lot of missing background material that had to be brought on line surrounding many of the properties of the Leech lattice, in particular its uniqueness,” – Jesse Han.
Automation was an indispensable partner in all of these successes. Beyond the code, Hariharan and Han both acknowledged the tremendous human effort that was driving all of this innovation and advancement.
“So it was a pretty fruitful collaboration,” – Sidharth Hariharan.
Now, Lean, a powerful programming language and proof assistant has been key in the formalization of mathematical proofs. Hariharan used Lean to formalize proofs and enhance his understanding of different mathematical concepts. The Formalising Sphere Packing in Lean project had started only in March 2024. It illustrates a remarkable synergy between human creativity and AI strength.
The Role of Collaboration and Technology
The partnership between mathematicians and AI has opened up exciting new potential in the field. As Liam Fowl, the other co-author of this research, noted, these advancements are groundbreaking for their importance.
“These new results seem very, very impressive, and definitely signal some rapid progress in this direction,” – Liam Fowl.
Fowl further emphasized that formal verification of a mathematical proof is simply the ultimate form of a “rubber stamp” of approval. This on-chain process drastically improves verifiable proofs with unparalleled accuracy and third party neutral reliability. It adds a layer of credibility to our faith in their legitimacy.
Han continued to describe how cutting edge Gauss is as a reasoning agent.
“It’s a particular kind of language model called a reasoning agent that’s meant to interleave both traditional natural-language reasoning and fully formalized reasoning,” – Jesse Han.
This two-fold functionality continues to let mathematicians, for example, explore deep proofs more quickly and accurately. AI tools in mathematics are now primed to improve these well-developed techniques. It will further spark innovative strategies for bringing the captivating worlds of mathematics to life.
A New Era for Mathematicians
As these technologies evolve, they promise to alleviate some of the burdens faced by mathematicians, allowing them to focus on creative problem-solving. Han encapsulated this vision by stating:
“I think the end result of technology like this will be to free mathematicians to do what they do best, which is to dream of new mathematical worlds,” – Jesse Han.
Policy Actions These steps would help realize AI as an enabler of human mathematical advances rather than a substitute. In their algorithm sphere-packing proof research, they’ve already proven how technology can accelerate progress and deliver impactful results. Most importantly, they underline that it is the collaboration between humans and machines that is most important for delivering amazing results.