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 reports a proof of the Aaronson-Ambainis conjecture, an open question dating to 2008 about quantum advantages on random inputs. The proof was formalized in Lean and checked by its kernel, but the result limits input queries, not total computing time or the practical value of quantum computers.
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 computers can gain exponential advantages on unstructured inputs. The theorem establishes that, for yes-or-no problems in the query model, classical algorithms can approximate a quantum algorithm’s answer probability on most uniformly random inputs without an exponential increase in input reads. PSI says an AI-assisted system helped find the proof; the paper’s main theorem was formalized in Lean and checked by its proof-verification kernel.
The result concerns query complexity: how many times an algorithm must access the input, rather than how long it takes to perform all calculations. For a quantum algorithm using a given number of queries, the paper says a deterministic classical algorithm can approximate its probability of answering “yes” on all but a chosen small fraction of inputs. With fixed accuracy and a fixed allowance for exceptions, the classical query count grows as a fixed power of the quantum query count, rather than exponentially.
PSI said in an accompanying blog post that its internally developed research system, using instances of a non-public Anthropic model, found a complete proof during the third day of a run. That is the company’s account of how the result was discovered. The paper says its main theorem and formulation of the conjecture were formalized in Lean; Lean’s kernel checked the formal proof. Researchers also reviewed the work, according to the report. These checks support the stated mathematical result, but the account of AI’s role comes from PSI.
The theorem does not show that classical computers can reproduce every quantum computation efficiently, nor does it rule out useful quantum speedups. Its claim is narrower: on most random inputs, a quantum algorithm cannot obtain an exponential advantage in the number of input reads for the class of yes-or-no computations covered by the result. The paper’s stated bounds concern access to data, not the full computational cost.
A Boundary for Random-Input Speedups
The proof gives researchers a more precise boundary for a central question in quantum computing: when can quantum methods outperform classical ones by an enormous margin? It points toward problems with exploitable mathematical or physical structure, rather than arbitrary inputs, as a more promising setting for dramatic advantages. This can help focus theoretical work on identifying the patterns that quantum algorithms use.
For developers and potential users of quantum technology, the result is not a verdict on whether quantum computers will be commercially useful. It does not assess hardware, error correction, implementation costs or the total time needed to solve an application. Instead, it narrows one theoretical route to speedup and clarifies what the query model can establish. Whether particular applications deliver useful advantages remains an open question.
Popular right now
De'Longhi Stilosa Compact Manual Espresso Machine with Milk Frother
Dreame X60 Max Ultra Complete Robot Vacuum and Mop
Ninja DZ801 DualZone 6-in-1 Air Fryer, 8 Qt, Grey
As an affiliate, we earn on qualifying purchases.
From the 2008 Conjecture to Lean
The conjecture takes its name from Scott Aaronson and Andris Ambainis and traces back to a 2008 lecture by Aaronson, according to the paper. It connects a question about quantum algorithms to mathematical statements about polynomials. Earlier work showed that a quantum algorithm’s probability of answering yes can be represented by a formula whose complexity is limited by the number of input queries. The new proof builds on that line of research.
The query model makes comparisons by counting requests for input information, not every operation a machine performs. A classical algorithm reads selected positions; a quantum algorithm can query a superposition of positions and use interference to extract information. The model is useful for isolating the cost of accessing data, but it leaves other computational costs outside the comparison.
Quantum algorithms can exploit structure in some problems. Shor’s factoring algorithm, introduced in 1994, uses periodicity in a mathematical function to find factors, offering a route to a major speedup over known classical methods. The conjecture asks what is possible when favorable patterns are absent and inputs are uniformly random.
“An internally developed research system, using instances of a non-public Anthropic model, found a complete proof during the third day of a run.”
— Physical Superintelligence, in its accompanying blog post
Limits of the Theorem and AI Claim
The supplied report does not specify the theorem’s full constants or give enough detail to assess the size of its polynomial query bound in practical settings. The result also does not settle whether specific quantum applications will provide a useful advantage when total runtime, hardware constraints and implementation are counted.
PSI’s account says its system found the proof, but the material does not independently quantify how much of the work came from the model, how researchers guided or revised it, or whether the same approach works on other open problems. Lean checks the formalized argument; it does not by itself establish the system’s authorship or the practical reach of the theorem.
Further Bounds and Application Tests
The immediate next step is scrutiny of the paper’s proof and its relationship to earlier results, alongside further work to strengthen the mathematical bounds. The source material does not report a journal publication or a timetable for one. Researchers will also need to examine which problems have enough structure to evade the theorem’s random-input setting and whether any resulting speedups survive beyond query complexity.
For PSI, further independent review could clarify the AI system’s contribution and the process used to formalize the result. For quantum computing more broadly, the theorem supplies a constraint for theory; evidence about practical applications will depend on separate algorithmic and engineering results.
Key Questions
What did the proof establish?
It establishes the Aaronson-Ambainis conjecture in the query model: for covered yes-or-no computations, a classical algorithm can approximate a quantum algorithm’s acceptance probability on most uniformly random inputs using a polynomial, rather than exponential, increase in input queries under fixed accuracy and exception settings.
Does this mean quantum computers have no speed advantage?
No. The theorem limits the number of input reads in a particular framework. It does not rule out quantum advantages in total runtime or in problems with structure a quantum algorithm can exploit.
How was the proof checked?
The paper says the main theorem and conjecture formulation were formalized in Lean and checked by Lean’s kernel, which verifies the formal logical steps. The report also says researchers reviewed the work.
What role did AI play?
PSI says its internal research system, using instances of a non-public Anthropic model, found a complete proof during the third day of a run. The supplied material does not independently establish how much of the proof work was done by the model or by researchers.
Source: thequantuminsider.com
