Publication, Claims, and External Communication

Quantum organizations are under pressure to publish progress. That pressure creates risk. A result may be technically interesting but commercially immature. A benchmark may be valid for one target and misleading as a general statement. A resource estimate may depend on assumptions that collapse under different error rates, code distances, or target architectures.

External communication needs the same rigor as execution. The goal is not to make claims timid. The goal is to make them durable.

82.1 Claim classes

Diagram loads as you read

82.1 Claim classes · Figure 1

View diagram source

flowchart TB
    Claim[Claim] --> Exploratory[Exploratory]
    Claim --> Internal[Internal decision]
    Claim --> Customer[Customer-facing]
    Claim --> Scientific[Scientific publication]
    Claim --> Regulatory[Regulatory or compliance]
    Claim --> Investor[Investor or board]

Each class requires a different evidence threshold. A whiteboard insight and a public benchmark result cannot share the same review path.

82.2 Claim review pipeline

Diagram loads as you read

82.2 Claim review pipeline · Figure 2

View diagram source

sequenceDiagram
    participant Author as Claim author
    participant Evidence as Evidence owner
    participant Technical as Technical reviewer
    participant Security as Security and legal
    participant Exec as Executive approver
    participant Publish as Publication channel
    Author->>Evidence: attach evidence package
    Evidence->>Technical: verify reproducibility and limits
    Technical->>Security: check sensitivity and controls
    Security->>Exec: send approved risk summary
    Exec->>Publish: approve external communication

The review should verify what the claim says, what it does not say, and which assumptions must be disclosed.

82.3 Claim anatomy

Diagram loads as you read

82.3 Claim anatomy · Figure 3

View diagram source

flowchart TB
    ClaimText[Claim text] --> Scope[Scope]
    ClaimText --> Target[Target and version]
    ClaimText --> Workload[Workload definition]
    ClaimText --> Baseline[Baseline]
    ClaimText --> Metric[Metric]
    ClaimText --> Uncertainty[Uncertainty]
    ClaimText --> Assumptions[Assumptions]
    ClaimText --> Limits[Limitations]

A mature claim includes enough context for a qualified reader to understand what was demonstrated and what was not.

82.4 Benchmark communication

Application-oriented benchmarks are useful because they connect performance to workloads rather than isolated device statistics [R135]. They are also easy to overstate. Do not turn a benchmark result into a blanket claim about quantum advantage unless the benchmark contract, baselines, uncertainty, and cost are all strong enough to support that conclusion.

Diagram loads as you read

82.4 Benchmark communication · Figure 4

View diagram source

flowchart LR
    Benchmark[Benchmark result] --> Context[Context]
    Context --> Baseline[Classical baseline]
    Context --> Cost[Cost]
    Context --> Uncertainty[Uncertainty]
    Context --> Claim[Bounded claim]

82.5 Red-team review

Diagram loads as you read

82.5 Red-team review · Figure 5

View diagram source

flowchart TB
    RedTeam[Red-team review] --> Reproduce[Can it be reproduced?]
    RedTeam --> Alternative[Is there a simpler explanation?]
    RedTeam --> Baseline[Is the baseline fair?]
    RedTeam --> Sensitivity[Are assumptions sensitive?]
    RedTeam --> Security[Does it disclose sensitive data?]
    RedTeam --> Misuse[Can it be misquoted?]

A claim that survives adversarial internal review is less likely to fail publicly.

82.6 Communication ladder

Diagram loads as you read

82.6 Communication ladder · Figure 6

View diagram source

stateDiagram-v2
    [*] --> LabNote
    LabNote --> InternalMemo: evidence complete
    InternalMemo --> CustomerBrief: approved scope
    CustomerBrief --> PublicPost: legal and technical approval
    PublicPost --> PeerReviewed: manuscript accepted
    InternalMemo --> Withdrawn: evidence insufficient
    PublicPost --> Corrected: error discovered

Moving up the ladder should require more evidence and stronger review.

82.7 Corrections and retractions

Diagram loads as you read

82.7 Corrections and retractions · Figure 7

View diagram source

flowchart LR
    Error[Error discovered] --> Assess[Assess impact]
    Assess --> Notify[Notify stakeholders]
    Notify --> Correct[Correct or retract]
    Correct --> Evidence[Update evidence package]
    Evidence --> Learn[Update review process]

The test of a mature organization is not whether it never makes a mistake. It is whether it can correct mistakes with evidence and speed.

82.8 Operating rule

Do not publish a quantum claim that the platform cannot reproduce, bound, explain, and correct.

Diagram loads as you read

82.8 Operating rule · Figure 8

View diagram source

flowchart LR
    Claim[External claim] --> Reproduce[Reproduce]
    Claim --> Bound[Bound uncertainty]
    Claim --> Explain[Explain limits]
    Claim --> Correct[Correct if needed]
    Reproduce --> Publish[Publish]
    Bound --> Publish
    Explain --> Publish
    Correct --> Publish