Biggest Breakthroughs in Math: 2024

This video surveys several major recent advances in pure mathematics, showing how researchers are tackling long-standing problems with fresh ideas. It highlights a new proof in sphere packing that recasts the geometry as a graph problem and uses the Rödl nibble to show randomness may produce optimal high-dimensional packings, breaking a 75-year record. The narrative also covers the fate of randomness in numbers: Szemerédi’s theorem says any sufficiently large set must contain arithmetic progressions, and a 2024 proof by three graduate students sharpened the density limits for avoiding them. Finally, the video explains the Langlands Program as a "grand unified theory" of mathematics, culminating in an 800-page proof of the geometric Langlands conjecture by Gaitsgory, Raskin, and collaborators. Overall, the message is that mathematical progress often comes from unexpected connections—between geometry and probability, patterns and randomness, and deep algebraic structures.

• ▶ 0:08 The sphere packing problem seeks the densest arrangement of spheres in a box; in 3D, the familiar pyramid stack is optimal, but higher dimensions are far more mysterious.
• ▶ 1:41 Kepler's 1611 conjecture on the optimal 3D packing — each sphere touching 12 others and filling ~74% of space — remained unproven for 400 years, with later breakthroughs by Viazovska in dimensions 8 and 24.
• ▶ 3:10 The new proof, built by converting geometry into a graph problem, uses a random process called the Rödl nibble to construct nearly random "independent sets," breaking a 75-year record and suggesting randomness may be key to optimal packings in high dimensions.
• ▶ 5:25 True randomness is impossible: any sufficiently large set of numbers must contain some kind of structure, leading to the central problem of determining how large a set can get before patterns emerge.
• ▶ 7:38 Szemerédi proved that as the starting pool of numbers grows, the maximum density you can take before an arithmetic progression appears approaches zero—so mathematicians want to know exactly how quickly this density limit shrinks.
• ▶ 9:00 A 2024 breakthrough by three graduate students—Meetaab Sawhney, Ashwin Sah, and James Leng—produced a better upper bound for sets with no five-term progressions and extended the result to arithmetic progressions of any length.
• ▶ 9:38 The Langlands Program is a sweeping "Grand Unified Theory" of mathematics, culminating in the geometric Langlands conjecture, proven by Dennis Gaitsgory, Sam Raskin, and seven co-authors in an 800-page proof after three decades of work.
• ▶ 10:34 The program is directly inspired by Fourier theory: just as the Fourier Transform breaks complex waves into labeled sine-wave building blocks, Langlands researchers seek analogous building blocks—eigensheaves—labeled by representations of the fundamental group.
• ▶ 13:23 Gaitsgory outlined the solution in a "fundamental diagram," but one key piece remained: proving the Poincaré sheaf contains every eigensheaf. Raskin found that missing proof in 2022, leading to five papers that completed the geometric Langlands conjecture.

• ▶ 0:08 Sphere Packing (0:08 - 5:15) - - How to pack spheres as densely as possible, from 2D history to Viazovska's breakthroughs and a new proof in all dimensions.
• ▶ 5:15 Structure and Randomness in Large Sets (5:15 - 9:38) - - Szemerédi's theorem, density, and a graduate-student breakthrough on patterns in large random-like sets.
• ▶ 9:38 The Langlands Program (9:38 - 15:02) - - A grand unified vision connecting number theory and geometry, culminating in a new proof of the geometric Langlands conjecture.