TL;DR
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
A paper by David Turturean of Physical Superintelligence says it proves the Aaronson-Ambainis conjecture, an open question since 2008 about quantum algorithms on random inputs. The proof was formalized in Lean and checked by its kernel; the result limits input queries for certain yes-or-no tasks, but does not bound overall runtime or establish which quantum applications will be useful.
A paper by David Turturean of Physical Superintelligence (PSI) reports a proof of the Aaronson-Ambainis conjecture, a question open since 2008 about whether quantum algorithms can gain exponential advantages on most random inputs. The result concerns how many times an algorithm must read its input to answer a yes-or-no question; it does not set a general limit on quantum computing speed or total processing time. PSI says its research system used a non-public Anthropic model to help find the proof, which was formalized in Lean and checked by the proof assistant’s kernel.
The paper’s central result is that, for a quantum algorithm using a given number of input queries, a deterministic classical algorithm can approximate its probability of answering “yes” on all but a chosen small fraction of uniformly random inputs. For fixed accuracy and a fixed allowance for exceptional inputs, the classical algorithm’s query count grows as a fixed power of the quantum algorithm’s query count. The paper therefore rules out an exponential gap in input reads for this setting on most random inputs.
PSI’s accompanying blog post says an internally developed research system, using instances of a non-public Anthropic model, found a complete proof during the third day of a run. The paper says its main theorem and its formulation of the conjecture were formalized in Lean. Lean’s kernel checked the formal proof’s logical steps; researchers also reviewed the work, according to the source report. This verification supports the correctness of the formalized argument, but does not by itself establish the broader practical implications of the theorem.
The result is framed in the query model, which counts requests for information about an input rather than every computation performed. A classical query reads a selected bit. A quantum query can act on a superposition of input positions, but it does not simply reveal all of them; algorithms use interference to extract information. The theorem compares the information-access requirements of particular computations, not the complete resources needed to run them.
A Bound on Random-Input Advantages
The proof gives researchers a sharper answer to a foundational question: whether quantum algorithms can achieve exponential query advantages for yes-or-no problems when inputs lack a favorable pattern. Under the theorem’s stated conditions, the answer is no for most uniformly random inputs. That helps distinguish quantum advantages that depend on exploitable structure from claims of broad superiority across arbitrary data.
For algorithm research, the implication is directional rather than a verdict on applications. Researchers have reason to examine problems with mathematical or physical structure that a quantum method can exploit. The result does not show that a classical computer can efficiently reproduce every quantum computation, nor does it identify which real-world tasks will benefit from quantum hardware. It clarifies one theoretical boundary while leaving those practical questions open.
Products Worth Considering
Bose QuietComfort 45 SE Wireless Noise-Cancelling Over-Ear Headphones
Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation
LG UltraGear 32GX850A-B 32-inch 4K UHD Glossy OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From a 2008 Conjecture to Lean
The conjecture is named for Scott Aaronson and Andris Ambainis. The paper traces it to a 2008 lecture by Aaronson and places the new result within a sequence of work in mathematics and theoretical computer science. The question links the behavior of quantum algorithms to a mathematical statement about polynomials: how complex a function must be to represent a quantum algorithm’s probability of answering yes.
That connection matters because the polynomial’s complexity is limited by the number of input queries. Earlier results established the link between quantum computation and these mathematical representations; the new paper says it proves the conjecture that turns that link into a classical simulation result for most random inputs. A familiar example of a structured quantum advantage is Shor’s factoring algorithm, introduced in 1994, which exploits periodicity in a mathematical function. The conjecture asks what can happen when inputs do not offer such helpful structure.
The development combines a mathematical result with a claim about how it was discovered. The theorem and its formal verification are described in the paper; the account of an AI system finding the proof comes from PSI’s blog. The company’s description does not mean the model independently certified the result: the reported checking was performed through Lean’s kernel, alongside researcher review.
“The company said its internally developed research system found a complete proof during the third day of a run.”
— Physical Superintelligence, in its accompanying blog post
Limits the Theorem Does Not Set
The result is narrower than a general statement about quantum computers. It limits input queries in a specified framework; it does not count all computation required to process data or prove that classical machines can match quantum performance in total runtime. The paper’s result also allows a chosen small fraction of inputs to be exceptions, so it is not a claim about every possible input.
The supplied report does not give the full theorem’s constants, exact bounds, or the size of the exceptional fraction used in particular applications. It also does not establish the commercial value of quantum machines or settle which practical tasks have useful quantum speedups. The AI-discovery account is attributed to PSI, and the model is not publicly identified in the material provided.
Further Bounds and Applications
The paper’s publication and review will give other researchers an opportunity to examine the proof, its formalization, and the assumptions behind the theorem. The source material does not specify a journal publication date or a further review milestone. Future work may seek tighter mathematical bounds and clarify how the result applies to particular classes of problems.
For quantum algorithm research, the next practical question is which problems have structure that can support an advantage and whether that advantage remains meaningful when total runtime and hardware demands are counted. The proof does not answer those questions; it provides a more precise theoretical constraint for researchers working on them.
Key Questions
What conjecture does the paper say it proves?
It reports a proof of the Aaronson-Ambainis conjecture, which concerns quantum query advantages for yes-or-no computations on most uniformly random inputs.
Does the result show quantum computers cannot be faster than classical computers?
No. It limits the number of input queries in a particular model and setting. It does not establish a general bound on total computing time or rule out practical quantum advantages.
What role did AI play in the proof?
PSI says its research system, using instances of a non-public Anthropic model, found the proof during a run. The paper says the argument was formalized in Lean and checked by Lean’s kernel; researchers also reviewed it.
What does it mean for the proof to be checked in Lean?
The theorem and its proof were encoded in a proof assistant, whose kernel checks whether the formal logical steps follow. This verifies the formalized argument, while questions about its assumptions and real-world relevance still require mathematical and scientific scrutiny.
What remains unresolved for quantum computing?
The theorem does not determine which real-world applications will produce useful speedups, how those gains compare in total runtime, or whether quantum hardware will be commercially valuable. Those questions depend on the problem and the resources required.
Source: thequantuminsider.com
