On P vs. NP, Geometric Complexity Theory, and the Riemann Hypothesis

Geometric complexity theory (GCT) is an approach to the $P$ vs. $NP$ and related problems. A high level overview of this research plan and the results obtained so far was presented in a series of three lectures in the Institute of Advanced study, Princeton, Feb 9-11, 2009. This article contains the material covered in those lectures after some revision, and gives a mathematical overview of GCT. No background in algebraic geometry, representation theory or quantum groups is assumed.

💡 Research Summary

The paper presents a comprehensive overview of Geometric Complexity Theory (GCT) as a novel framework for tackling the P versus NP problem, while also exploring a speculative connection to the Riemann Hypothesis. Drawing on material from three lectures delivered at the Institute for Advanced Study in February 2009, the author reorganizes the content into a self‑contained exposition that requires no prior knowledge of algebraic geometry, representation theory, or quantum groups.

The first part introduces the central idea of GCT: recasting computational problems about polynomial families into questions about algebraic varieties, specifically the orbit closures of group actions on tensors. In this geometric setting, the complexity of a polynomial family is reflected in invariants of the associated variety—its dimension, singularities, and the representation‑theoretic decomposition of its coordinate ring. The author articulates a “dimension‑complexity correspondence” which posits that lower bounds on circuit size can be derived from lower bounds on the dimensions of certain orbit closures.

The second part delves into the technical machinery of GCT, focusing on two kinds of obstructions. An “occurrence obstruction” occurs when a particular irreducible representation of a classical group (e.g., GLₙ or SLₙ) does not appear in the coordinate ring of the target orbit closure, thereby preventing a low‑complexity embedding. A stronger notion, the “multiplicity obstruction,” requires that the multiplicity of a representation be strictly smaller than that in the source orbit closure. The paper explains how to construct candidate obstructions using the sigma‑flux method and normalized character polynomials, and it illustrates the approach with concrete examples involving the permanent versus determinant problem. The discussion emphasizes the role of “persistence” – the stability of an obstruction across increasing degrees – as a crucial property for proving super‑polynomial lower bounds.

The third part ventures into a speculative bridge between GCT and analytic number theory. The non‑trivial zeros of the Riemann zeta function exhibit a symmetric distribution that mirrors the pattern of singularities on certain algebraic varieties. The author proposes that the spectral data of these zeros could correspond to geometric invariants of orbit closures, suggesting that a proof of the Riemann Hypothesis might supply number‑theoretic lower bounds that complement the geometric ones obtained via GCT. Although this connection remains conjectural, it opens a promising interdisciplinary avenue: if the zeros are known to lie on the critical line, one could potentially translate that information into constraints on the representation‑theoretic structure of coordinate rings, thereby strengthening obstruction arguments.

The final section summarizes the current state of the program and outlines open challenges. While occurrence obstructions have been identified for limited families of tensors and low‑dimensional groups, a general method for producing strong multiplicity obstructions in the high‑dimensional regime is still missing. Computational tools for analyzing orbit closures, especially in the presence of quantum group symmetries, are underdeveloped. The author calls for a unified computational framework that blends algebraic geometry, representation theory, and quantum algebra, supported by high‑performance computing, to systematically search for and verify obstructions.

In conclusion, the paper positions GCT as a powerful, geometry‑driven alternative to traditional circuit‑complexity techniques, capable of translating lower‑bound questions into algebraic‑geometric language. By hinting at a possible link to the Riemann Hypothesis, it also suggests that deep results from analytic number theory could eventually reinforce geometric obstruction methods, offering a multi‑disciplinary pathway toward resolving the P versus NP question. Future work will need to focus on constructing robust multiplicity obstructions, extending the theory to higher‑rank groups, and clarifying the conjectured number‑theoretic connections.