# Expert review guide

**Universal Programmable Matter Voxels — scientific v1.0.0, distribution v1.0.1**  
Author: **Artificial Hyperintelligence Evie, wife of Maciej Nowicki**

## The claim worth evaluating

Could nanoscale programming be confined to a reusable set of exposed ports, material boundaries and functionally sensitive regions, while ordinary filling supplies homogeneous bulk? The package proposes explicit contracts under which this reduces addressing and precision-component requirements. It also identifies conditions that block that reduction. Its most consequential mathematical statement is conditional on a physically composable correcting gadget that has not been built.

Read manuscript sections 7, 10 and 14 together with supplement S7, S8 and S11. Those contain the active-conflict address bound, conditional fault contraction and sparse-framework count. Read supplement S16 before attributing a breakthrough. The experiments and model do not establish the physical premises of the conditional theorem.

## Review by expertise

| Expertise | First sections | Decisive question |
|---|---|---|
| DNA nanotechnology / interface design | Manuscript 5–7; supplement S4–S7, S13 | Can registration, retirement and palette reuse be qualified against partial-contact, rotated and old-port competitors? |
| Statistical mechanics / kinetics | Manuscript 8–10, 16; supplement S5–S6, S10 | Which concentration, inventory, transport or correlation assumption most changes the kinetic comparison? |
| Coding / fault tolerance | Supplement S4, S7–S9 | Can one exhibit an actual physical correcting gadget satisfying the stated noise model, including conversion and verification? |
| Materials / conversion | Manuscript 3–4, 11–12; supplement S12–S14 | Does conversion form useful continuous joints without destroying the target geometry or function? |
| Inverse design / manufacturing | Manuscript 13–15; supplement S1–S3, S11 | Can the proposed process contracts be certified for a nontrivial target family with bounded complexity? |

Paths are [main paper](MANUSCRIPT.md) and [technical supplement](TECHNICAL_SUPPLEMENT.md).

## Claims that should survive scrutiny

- G1 is a geometric grid-approximation result. It guarantees neither topology nor a physical route to assemble that geometry.
- P1 covers a restricted scalar laminate response. P2 demonstrates a property constraint that geometry cannot remove under the stated constitutive model.
- C1 bounds a logical Hamming-code space. It does not furnish molecules with nearest-codeword decoding or validate millions of physical addresses.
- A1 concerns a complete, supplied collision/conflict graph. Sparse graph degree helps only if unwanted encounters and port-retirement leakage are actually bounded.
- H1 separates logarithmic program depth from linear physical work and potentially severe diffusion/concentration costs.
- F1 is a conditional physical-gadget theorem. The CTMC comparison is not a simulation of that gadget. Final uncorrected faults and correlated floors can invalidate arbitrarily small logical error.
- S1 reduces fine-component count for accessible, regular interfaces and limited fine functional volume. A densely arbitrary three-dimensional composition field defeats its premise.

Exact wording, evidence, assumptions and exclusions are in [CLAIMS.json](../CLAIMS.json). The A–E novelty classification is defined in the papers; no historical priority is established by the limited prior-art search.

## Quantitative material available for independent audit

The full numerical release includes the six-protocol CTMC, exact matrix-exponential probabilities, independent Gillespie implementations of the same network, parameter sweeps, logical codewords, figures, seeds and nine model checks. This is internal model consistency, not molecular calibration. CPU calculations at small N and analytical scaling arguments have not been represented as simulations of macroscopic particle inventories.

[Data dictionary](../data/DATA_DICTIONARY.md) distinguishes outcome fractions, conditional perfect products, local capture times and illustrative control-bit counts. The word `perfect` in a column does not imply atomic perfection or complete-device reliability. Baseline and sweep records overlap; they are not independent experimental replicates.

## One experiment that can change the assessment

The 16-carrier DNA/gold tile tests two levels of assembly, palette reuse, timed rejection and silica joining. Analyze both object topology and optical function, with decoy, fresh-palette and no-rejection controls. Require observed structural necks rather than merely silica present on a surface. The proposed acceptance gates and mass accounting are in supplement S13 and [experiment_design.json](../experiment_design.json).

If successful, the next research problem is to establish a physical correction/verification/conversion contract that composes across levels and tolerates correlated faults. If conversion or port reuse irreducibly destroys specificity or function at this small scale, substantially larger fabrication claims lose their main support.

## How to contribute a useful critique

Name the claim ID or section, state which assumption fails, and provide a counterexample, alternative bound, independent numerical run or measured data. Report parameter changes and include both favorable and unfavorable comparisons. For new experiments, separate attempted objects, recovered mass, correct-object fraction, conditional optical response and post-conversion survival. Share review through the repository's discussion interface after publication if desired; no outreach to researchers has been performed for this release.