Quick read · plain language
Eight Common Claims About Quantum Computers and Their Limits
A plain-language version of “Locale Framework Applied to Quantum Computing Innovations & Practical Applications”
The short version
This paper checks eight popular claims about quantum computing. Each claim is true only in special conditions. The paper finds where each claim breaks down. Energy cost per answer is the one thing that matters across all cases.
Why this paper was written
People hear the same sentences about quantum computers again and again. Continuous systems have a natural edge. More parts mean more power. A fifty-qubit machine has shown quantum advantage. Complex numbers are essential. A transmon is a qubit. Fault tolerance is five years away.
Each sentence is true in some way. Each is also false in another way. This is normal for physics. Every claim works only inside a special setting — a certain energy level, noise level, or measurement setup. The paper uses a "locale framework" to make these settings clear. A claim must always come with its locale. The interesting physics lives where a rule stops working.
The paper does not attack or defend quantum computing. It gives a map. For each of eight claims, it shows the needed setting, the breaking point, and the research that documents the break.
What the locale framework means
The framework comes from a companion paper. This paper uses three of its ideas.
First, claims are conditional. They hold only inside a locale.
Second, spec sheets and roadmaps are maps. Maps help us learn, but they fail at the edges.
Third, a buyer sees only the image — benchmarks, press releases, slides. The real physics hides behind this image. No finite set of user observations can tell whether the hidden physics uses one kind of math or another. But structure leaks through. The paper finds one clear leak.
The eight claims and where they break
The paper lists eight common claims. For each, it gives the setting where it holds, the boundary where it fails, and the research from one conference week that documents the failure.
None of the claims is simply wrong. Each is a conditional truth. The error is leaving out the condition. The listener fills in a setting that the physics does not support.
The boundaries are not guesses. All but one were shown at a single week-long conference in 2026. The program acts as a living check on the field's own claims.
The list serves as a tool. A buyer or engineer who hears one of these claims can ask: what setting is assumed? Does current work support it?
Maps, real things, and what leaks through
The eight claims work as maps. A vendor's spec sheet is a map. A benchmark is a map. Maps are not wrong. They are bound to their setting. Reported fidelity, qubit count, roadmap dates — all are honest numbers that fail to transfer outside their measured setting.
The real thing that transfers is energy accounting. The first three claims fail in the same way: an advantage vanishes when the energy budget is made clear. What survives is the full cost of getting one correct answer. This includes computation, cooling, control, decoding, and the classical rival.
The image-space point is subtle. A buyer sees only dashboards and slides. The real device physics stays hidden. No finite user observations can decide the hidden math type. But structure leaks through. The clearest leak is exact synthesis for one gate set. It carries an arithmetic object — the natural geometry of a different number system. This appears inside a problem the field treated as ordinary approximation. The hidden physics leaked through the image.
What survives: counts, ratios, and energy
Strip away units and three kinds of quantity remain.
Counts: qubits, gates, cycles.
Ratios: error rates, overhead factors.
Energy per answer: a mix of counts and ratios. This answers the only question a practitioner really needs: what does a correct answer cost?
Two results make the energy point concrete. One shows that a continuous-variable computation with a set energy budget can be matched by a different system with bounded error. The other shows the reverse: with a fixed number of parts, the energy needed grows fast. Together they say energy is the exchange rate between ways of computing. The joules-per-answer measure is the natural unit of this rate.
Work at the conference that matters for builders and buyers
The conference had over one hundred talks. The paper picks those most useful for people building, buying, or testing quantum systems.
Some talks gave exact ways to build gates. Others gave better ways to count the resources circuits need. Some covered systems beyond the simple two-level case. Some addressed how to compile programs across different hardware layouts.
Some talks improved classical simulation. One sampler handles certain circuits in seconds. Others simulate specific quantum states on ordinary computers. These set the bar that quantum claims must beat.
Some talks advanced error correction. They gave better ways to prepare states, preserve fault tolerance, and cut resource needs.
Some talks gave practical tools ready for use: a security protocol needing only one measurement per site, and a memory design using quantum walkers.
Answers to five challenges
Five challenges from the conference are addressed.
Statistics: New proofs show that systems with full quantum permutation invariance must be bosons or fermions. This fits the framework: statistics is a locale effect, not a universal rule. An open question remains whether this also constrains anyons.
Real-valued quantum theory: This reproduces all tested network results once independence is imposed operationally. It cannot be ruled out by experiment. The claim that complex numbers are essential is conditional: true for braiding, indistinguishable from real-valued for finite networks so far tested.
Energy as exchange rate: The two main results are the first external proofs of the framework's seam crossings. They turn the joules-per-answer idea from a proposal into a theorem-backed tool. Open questions remain about the tightness of one bound and whether a construction step can be softened.
Classical simulation: New classical methods shrink the setting where device-level advantage claims hold. The surviving invariant is still joules-per-answer, including the classical rival's energy cost — which the field does not yet track well.
Arithmetic structure of synthesis: The leaked structure in exact synthesis is not a challenge but a match. It independently confirms that non-standard arithmetic matters at a practitioner seam. It is adopted as independent support, not as derived proof.
What the paper concludes
Eight claims, eight boundaries, one conference week, and one independent online result. The practical core is the table: a setting-indexed map of daily claims, each paired with the exact paper showing where it breaks. The theoretical core is the repeated pattern — a claim true inside a setting, failing at an edge, with the interesting physics living at that edge.
Three conclusions follow.
Claims are conditional truths. Leaving off the condition is the field's most costly habit. It lets roadmaps read as promises and benchmarks as products.
The boundaries are public. One conference week contains nearly all the documentation. The check on the field's claims is not future work. It is the field's current research program.
Energy per answer is what survives. Every boundary is at bottom an energy boundary. The quantities that cross all of them are counts, ratios, and the mix that answers: what does a correct answer cost?
What this paper does not do: It does not propose a dynamic theory of boundaries. The table is a catalog, not a prediction. It does not decide whether the hidden physics uses one math type or another. It only shows the question is real, and that the answer will leak through before it can be decided from the inside.
Words used here
- locale
- the special setting — such as an energy level or noise level — inside which a physics claim holds true
- transmon
- a type of superconducting circuit used in quantum computers, which actually has about twelve usable levels rather than just two
- fault tolerance
- the ability of a quantum computer to keep working correctly even when some parts fail
Learn more in plain language
- TransmonWikipedia (English, more technical)
- Fault toleranceWikipedia (English, more technical)
This plain-language version was written by an AI model from the text of the paper and checked against plain-language rules: everyday words, no mathematics, no unexplained names. It leaves details out and can still be wrong; the paper is the source of record. To report a problem, open the paper, select the passage and choose Report.
For specialists: the technical digest, quoted from the paper
In the paper's words
Quantum computing is often presented through a small set of recurring claims: continuous-variable systems are naturally advantaged, more modes means more power, quantum advantage has been demonstrated at device level, statistics could in principle be non-standard, complex amplitudes are essential, approximation-based gate synthesis is the only approach, a transmon is a qubit, and fault tolerance is a matter of years. This paper argues that each of these claims is true only within a restricted domain of applicability, and that the boundaries at which they fail are not defects of the claims but the places where the physics relevant to practitioners actually lives. Drawing on the locale framework introduced in a companion record [1], the paper states eight of the most widely repeated claims, makes the domain of each explicit, and pairs each boundary with work presented at the 23rd International Conference on Quantum Physics and Logic (QPL 2026) — together with one independent arXiv result — including new results on the energy cost of continuous-variable computation [2], a closely related independent result on trading modes against energy [3], classical simulation of Clifford+T circuits [4], invariance under quantum permutations [5], real-valued quantum theory [6,7], and the arithmetic structure of gate synthesis [8].
The abstract, verbatim, and its opening sentence is the claim. Every line below is quoted from the text.
Numbers it reports
The territory that survives is the cost of producing a solution, measured end to end: computation, cooling, control, decoding, and the classical baseline it competes against [10,11].
from 4. Why the boundaries matter: maps, territory, and the interfaceThe $CV\leftrightarrow DV$ equivalence of Maltesson et al. [2] says that a continuous-variable computation of n modes with energy budget E* is simulable by qudits with error at most $1286 \cdot K \cdot n^{2}$·E*²/√d.
from 5. What survives: counts, ratios, and energy
How it is done
Quantum computing is often presented through a small set of recurring claims: continuous-variable systems are naturally advantaged, more modes means more power, quantum advantage has been demonstrated at device level, statistics could in principle be non-standard, complex amplitudes are essential, approximation-based gate synthesis is the only approach, a transmon is a qubit, and fault tolerance is a matter of years.
from AbstractMaps are not wrong — they are locale-bound.
from 4. Why the boundaries matter: maps, territory, and the interfaceThis is consistent with the pre-registered negative result of the author's earlier work on the statistics distinction [12]: statistics is a locale effect of permutation invariance, not a theorem of any single formalism.
from 7. Responses to five challengesThe Joules-per-Solution Metric: Definition, Measurement Protocol, and Anti-Gaming Provisions for Honest Computational Benchmarking . the QNFO archive, https://papers.qnfo.org/papers/joules-per-solution-metric/. [12] Quni-Gudzinas, R.
from References
What would falsify it
This is an external result, produced by researchers working in the exact-synthesis tradition; its significance for the locale framework is that it is exactly the kind of structure leak the framework predicts [1,8].
from 4. Why the boundaries matter: maps, territory, and the interfaceHoffreumon and Woods show that real-valued quantum theory reproduces all finite network correlations once independence is imposed operationally, and that it cannot be experimentally falsified [6,7].
from 7. Responses to five challengesWhat this paper does not claim: no dynamical theory of boundaries is proposed; the table is a catalog, not a prediction; and nothing here decides whether the substrate of quantum computing is Archimedean or ultrametric — only that the question is real, and that the answer will leak through the rendering before it is decided from inside it.
from 8. Conclusions
Where the claim stops
What this paper does not claim: no dynamical theory of boundaries is proposed; the table is a catalog, not a prediction; and nothing here decides whether the substrate of quantum computing is Archimedean or ultrametric — only that the question is real, and that the answer will leak through the rendering before it is decided from inside it.
from 8. ConclusionsQuantum computing is often presented through a small set of recurring claims: continuous-variable systems are naturally advantaged, more modes means more power, quantum advantage has been demonstrated at device level, statistics could in principle be non-standard, complex amplitudes are essential, approximation-based gate synthesis is the only approach, a transmon is a qubit, and fault tolerance is a matter of years.
from AbstractThis paper argues that each of these claims is true only within a restricted domain of applicability, and that the boundaries at which they fail are not defects of the claims but the places where the physics relevant to practitioners actually lives.
from Abstract
Where to read
- Abstract247 words · 1 min
- 1. Introduction267 words · 1 min
- 2. The locale framework in brief196 words · 1 min
- 3. Eight claims and their boundaries480 words · 2 min
- 4. Why the boundaries matter: maps, territory, and the interface345 words · 2 min
- 5. What survives: counts, ratios, and energy186 words · 1 min
- 6. Practitioner-oriented work at QPL 2026347 words · 2 min
- 7. Responses to five challenges441 words · 2 min
- 8. Conclusions294 words · 1 min
Read from the paper itself: 11 sections, 3,072 words. Nothing in this digest is written by a model; a line that is not a sentence of the paper is not shown. Have a specific question? Ask this paper.
The full text · Ask the whole corpus
Keep exploring
One body of research, three doors.