A short proof of the Aaronson-Ambainis Conjecture via the completion framework

In the world of computing, there is a fundamental divide between how classical machines and quantum machines solve problems. Classical computers, the ones we use every day, process information bit by bit, like flipping switches on and off. Quantum computers, still largely in the experimental stage, use the strange laws of physics to process information in a way that can be vastly more efficient for certain tasks. For decades, researchers have tried to understand exactly where this quantum advantage comes from and, more importantly, how much of it can be mimicked by classical machines if we are willing to accept a small margin of error. The central question is whether a quantum computer's ability to solve a problem quickly implies that a classical computer can also solve it quickly, provided the classical computer is allowed to get the answer right most of the time, rather than every single time. This question lies at the heart of a long-standing puzzle known as the Aaronson-Ambainis conjecture, which suggests that quantum speedups are not as mysterious or unattainable as they first appear, but are instead rooted in properties that classical algorithms can eventually uncover.

A team of researchers at Stony Brook University has now provided a concise proof of this conjecture, offering a clear path forward for understanding the relationship between quantum and classical computing. Their work focuses on partial functions, which are problems where the computer only needs to give an answer for a specific subset of possible inputs, rather than every single combination of data. The researchers demonstrated that if a quantum computer can solve such a problem using a certain number of questions, or queries, to the data, then a classical computer can also solve it with high accuracy using a number of queries that is related to the quantum count, though potentially larger. Specifically, they proved that a classical algorithm can correctly compute the answer for at least a 1 − ε fraction of the valid inputs (where ε is an arbitrarily small positive number) by asking a number of questions that grows predictably based on the quantum complexity and the size of the input set. This result confirms that the power of quantum computing, in this specific context, does not rely on magic or inaccessible physics, but on structural features that classical logic can eventually map out.

The proof relies on a method called the completion framework, which the researchers had previously developed. To understand this, imagine a puzzle where you are given a picture with a few missing pieces. The goal is to fill in the gaps so that the entire picture makes sense, even if you don't know exactly what the missing parts looked like originally. In the context of this research, the "picture" is the function the computer is trying to solve, and the "missing pieces" are the inputs where the function is not defined. The researchers showed that if a quantum computer can solve the puzzle for the known pieces, it is possible to extend the solution to cover the missing pieces in a way that a classical computer can follow. They did this by constructing a complete version of the problem that agrees with the original quantum solution on almost all the important inputs. This process involves discarding a tiny fraction of the difficult cases where the classical computer might struggle, ensuring that the remaining cases are easy enough to solve efficiently.

The core of their discovery involves a mathematical tool known as the classical adversary measure, which acts as a way to count how sensitive a problem is to changes in its input. The researchers proved that if this sensitivity measure is low, a classical algorithm can systematically reduce the uncertainty of the problem by asking the right questions. They showed that by repeatedly choosing the most informative questions to ask, the algorithm can narrow down the possibilities until it reaches a point where the answer is clear. This process is guided by a concept called influence, which measures how much flipping a single bit of data changes the final answer. The team demonstrated that as the algorithm asks questions, the total influence of the remaining unknown parts of the problem decreases steadily. By tracking this decrease, they could guarantee that the algorithm would reach a correct answer for the vast majority of inputs within a predictable number of steps.

A crucial part of their work involved addressing a specific inequality that had been a stumbling block in previous attempts to prove the conjecture. While a recent, much longer proof by another researcher had solved this inequality, the Stony Brook team found a way to prove a special case of it themselves, keeping their argument self-contained and shorter. They used this result to show that the classical algorithm does not need to know the exact structure of the quantum solution to work; it only needs to know that such a solution exists. This finding strengthens the idea that quantum speedups are not fundamentally alien to classical logic but are instead a matter of efficiency and structure. The researchers also extended their proof to show that this logic applies to a stronger version of the conjecture, which deals with how the influence of individual bits is distributed across the problem.

The significance of this work lies in its clarity and brevity. By providing a short proof of the conjecture following a recent 60-page preprint, the researchers have made the underlying logic of the Aaronson-Ambainis conjecture more accessible to the broader scientific community. They have shown that the gap between quantum and classical computing is bridgeable under the right conditions, specifically when we are willing to accept a small error rate. This does not mean that classical computers will soon replace quantum ones for all tasks, but it does mean that the theoretical foundation for quantum advantage is more grounded in classical principles than previously thought. The proof confirms that the power of quantum algorithms is not an impenetrable mystery but a property that can be understood, simulated, and bounded by classical methods.

In the end, the paper provides a definitive answer to a question that has lingered in the field of theoretical computer science, following a recent preprint posted in November 2026. It establishes that the ability of a quantum computer to solve a problem quickly is a strong indicator that a classical computer can also solve it quickly, provided the classical computer is allowed to skip a small fraction of the hardest cases. This result brings us closer to a unified understanding of computation, where the distinction between quantum and classical is not a wall, but a gradient that can be navigated with the right tools. The researchers have not just solved a puzzle; they have provided a new lens through which to view the capabilities of both types of machines, suggesting that the future of computing may rely less on choosing one over the other and more on understanding how they complement each other.