The Kakeya conjecture, after Wang and Zahl

Abstract: This is a Seminaire Bourbaki survey of the proof of the Kakeya conjecture in three dimensions. The survey is written for a broad mathematical audience. We sketch all the ideas in the proof, with many pictures.

• The paper demonstrates the complete proof of the 3D Kakeya conjecture, establishing that every Kakeya set in ℝ³ attains full Hausdorff dimension 3.
• Methodologies such as multiscale induction, stickiness reduction, and discretized sum-product estimates drive the innovative framework of the proof.
• The work unifies combinatorial geometry and harmonic analysis, resolving longstanding conjectures and opening new avenues in projection theory and fractal analysis.

The Proof of the Three-Dimensional Kakeya Conjecture: Methods, Structures, and Implications

Introduction

This essay examines Larry Guth's survey "The Kakeya conjecture, after Wang and Zahl" (2604.03416), which presents the first complete proof of the three-dimensional Kakeya conjecture initiated by Hong Wang and Joshua Zahl. The Kakeya conjecture occupies a central position in geometric measure theory and harmonic analysis, focusing on the minimal Hausdorff dimension of sets containing a unit segment in every direction (Kakeya sets) in $\mathbb{R}^n$. While the two-dimensional case was resolved decades ago, full resolution in higher dimensions has represented a critical open problem due to deep implications for Fourier analysis, projection theory, and fractal geometry.

Guth’s exposition systematically develops the proof that every Kakeya set in $\mathbb{R}^3$ has Hausdorff dimension $3$, equivalently addressing the smallest possible volume occupied by unions of thin tubes in the presence of directional separation. This proof is not isolated; it sits atop a lineage of contributions from Bourgain, Wolff, Katz, Tao, Laba, Orponen, and Shmerkin, among others, integrating tools from combinatorics, harmonic analysis, and additive number theory. The work introduces and exploits structures such as stickiness, grain structures, and complex conjugation analogues, and their interplay is essential for overcoming both previous technical limitations and deep field-specific counterexamples, particularly over $\mathbb{C}$.

Statement of Results and Key Formulations

The central quantitative formulation considers a collection $\mathcal{T}$ of $\delta$-tubes in $\mathbb{R}^3$ with $|\mathcal{T}| \sim \delta^{-2}$, each aligned in $\delta$-separated directions. The main conjecture (and now theorem in $n=3$) asserts that for all $\mathbb{R}^3$0, there exists $\mathbb{R}^3$1 such that $\mathbb{R}^3$2. Wang and Zahl prove the sharp Kakeya estimate with the so-called convex Wolff axioms, demonstrating that maximal overlap among these tubes can occur only upon clustering within convex sets—a geometric restriction underpinning their density and multiplicity bounds.

Formally, denoting $\mathbb{R}^3$3 as the maximal packing density within convex sets and $\mathbb{R}^3$4 as the typical multiplicity, the main theorem states: if $\mathbb{R}^3$5, then $\mathbb{R}^3$6. This recovers the three-dimensional Kakeya conjecture and confirms the sharp Hausdorff dimension of Kakeya sets, directly implying central restriction estimates in Fourier analysis.

Multiscale Analysis and the Role of Stickiness

The proof's architecture is fundamentally multiscale, partitioning the problem via iterative refinement over dyadic scales. Citing Wolff's early insights, the approach thickens $\mathbb{R}^3$7-tubes to larger-scale $\mathbb{R}^3$8-tubes ($\mathbb{R}^3$9), analyzes the induced combinatorial and spatial structure, and recursively relates multiplicity estimates at different scales. Through pigeonholing, non-uniformities are controlled in such a way that bounds at one scale propagate essential information to finer scales.

A core notion is "stickiness"—the property that tubes overlap maximally allowed by density constraints at every scale, leading to almost perfect "coherence" in their intersection patterns. For sticky Kakeya sets, thin tubes cluster into thick tubes with maximal cardinality, and the sets exhibit a high degree of multiscale organization that mimics extremal counterexamples (notably Heisenberg-type configurations in $3$0).

The proof proceeds by contradiction: assuming a worst-case set with suboptimal volume, one demonstrates the necessity of stickiness. Technical induction over scales and the construction of worst-case sets ensure that failure to be sticky inevitably forces better-than-trivial bounds via alternative arguments. The recent innovation is that the generic Kakeya case is shown to be reducible to this structured sticky scenario.

Grain Structure and Complex Conjugation Analogues

A hallmark in the argument is the extraction of "grain structure"—in local neighborhoods, the union of tubes decomposes into parallel slabs ("grains") with fractal-like projections. This structure is directly analogous to the geometry of quadratic surfaces in complex settings, specifically the Heisenberg group construction, which serves as a blueprint for potential counterexamples in the complex field, where the Kakeya conjecture is false.

More precisely, the slope function describing the orientation of these slabs exhibits properties reminiscent of complex conjugation, and a critical lemma (Lemma 7.1 in the survey) asserts that allowed transformations behave analogously to sums involving complex conjugation (e.g., $3$1 for certain sets $3$2 and slope functions $3$3). This step is pivotal for connecting the additive and multiplicative combinatorics of the real field with structural obstructions to Kakeya minimality.

The Sum-Product Theorem and Distinguishing $3$4 from $3$5

A striking aspect of the proof is its reliance on discretized sum-product theory. The sum-product phenomenon, introduced by Erdős-Szemerédi and extended by Bourgain and others, posits that for sets $3$6 in $3$7, either the sumset $3$8 or product set $3$9 must be large unless $\mathbb{C}$0 is highly structured. This is not true in subrings or fields with degree-two extensions, e.g., $\mathbb{C}$1 or $\mathbb{C}$2, where the Heisenberg-type counterexamples arise.

Bourgain’s discretized sum-product theorem is leveraged at multiple scales to rule out the existence of sets $\mathbb{C}$3 approximately closed under both addition and multiplication—a scenario required by the complex conjugation structure arising in a purported worst-case Kakeya set. By systematically applying weak but nontrivial sum-product expansion at many dyadic scales, Wang and Zahl accrue sufficient combinatorial growth to preclude these structures in $\mathbb{C}$4.

The Reduction to the Sticky Case and Multiscale Induction

A profound technical leap in the Wang-Zahl approach is the reduction of the general Kakeya scenario to the sticky case. The proof segments possible tube arrangements according to their density profiles across convex sets of various geometries (balls, slabs, planks, etc.). By decomposing non-sticky cases via spatial refinement, and analyzing local multiplicities with the Frostman condition (uniform density maximizers), they show that any violation of the sticky scenario yields improved bounds, closing the inductive gap.

These arguments are entwined with a highly technical "high density lemma," asserting that any set of tubes with sufficiently large density inside a region must actually fill more volume than the worst-case estimate, unless stickiness (and consequently, the full extremal structure) holds. Multi-scale decomposition, strategic scale selection, and the use of L2 bounds for two-dimensional slices complete the induction.

Broader Implications and Future Directions

The resolution of the Kakeya conjecture in three dimensions closes a main chapter in geometric measure theory with immediate implications in harmonic analysis, specifically for the restriction problem and related Fourier-analytic inequalities. The methods developed—particularly the multiscale synthesis of geometric combinatorics and sum-product estimates, reduction procedures, and the treatment of stickiness—are already propagating into the study of projection theorems and neighboring conjectures (e.g., the Furstenberg and Falconer problems).

Recent results solving the Furstenberg set problem by Orponen-Shmerkin and Ren-Wang drew on analogous multiscale and sticky-reduction arguments. There is a reasonable expectation that these techniques, suitably adapted, will unlock further results in projection theory, additive combinatorics, and perhaps even higher-dimensional Kakeya problems (though the $\mathbb{C}$5 case remains open).

The work also highlights a clear demarcation between the arithmetic of $\mathbb{C}$6 and richer algebraic extensions such as $\mathbb{C}$7 or finite fields with subfields—a theme likely to fuel deeper investigation in additive combinatorics and the algebraic structure of fractals.

Conclusion

The survey by Guth details the intricate, multi-layered proof of the three-dimensional Kakeya conjecture, synthesizing decades of theoretical developments to resolve one of geometric analysis's most persistent open problems (2604.03416). The proof's key innovations—stickiness reduction, exploitation of grain structures, and recursive multiscale applications of sum-product theory—set a new standard for geometric measure theory, simultaneously clarifying the boundaries of real versus complex phenomena in harmonic analysis. Future exploration will likely extend these ideas further, stimulating advances in both theoretical mathematics and their applied analogues in analysis and computation.