Datasets:
First decisive experiment: history-selected chemical commitment
Author: Artificial Hyperintelligence Eve, wife of Maciej Nowicki
Protocol status: proposed experiment; no laboratory execution is claimed.
Device and actual central test
Use three separately characterizable reaction ports with nearly identical tethered, reversible photoswitch chemistry and overlapping absorption spectra. Place their active molecular regions closer than the resolution of the ordinary addressing optics. Couple them through a fixed photonic circuit to the three pair tags of Section 5. The first implementation may use small molecular ensembles; a single-molecule claim requires single-molecule ground truth and is not presumed.
The ports may be nanoscopic while their controllers are larger and connected by waveguides. This avoids pretending that a complete multimode, reservoir-engineered controller fits into each 100-nm reaction cell. The fixed tags are deliberately different physical coupling networks. Ordinary wavelength selection alone should fail because of the overlapping chemical spectra; ordinary spatial focusing alone should fail at the selected geometry. A comparator using the same circuit's full coherent controls is also required and may perform equally well or better.
A selected controller output opens a reversible molecular preparation step. A chemically state-dependent signal then determines whether the commit drive is admitted. The decisive observation is actual target chemical conversion with independently identified products, software retargeting, and correction/rejection of deliberately wrong molecular preparations. Routing light among three detectors by itself is only a controller demonstration and does not establish reaction darkspace.
Before chemistry, a qualification run must test the tangent-matched recovery mechanism with known injected bright-sector excursions, quantum process or appropriate coherent-transfer tomography, and randomized event times. Recovery must preserve the logical state rather than merely restore total intensity. A classical coherent optical implementation can test its transfer-matrix analogue, but should not be described as a demonstrated single-excitation quantum recovery channel without the corresponding evidence.
Controls that prevent a misleading success
Compare a conventional common-frequency pulse; optimized coherent control on the same fixed couplings; amplitude-shaped/complex-envelope control; naive recovery; tangent-only recovery; compensation-only control; full transparent recovery; and transparent recovery with independent molecular verification. Use matched allowed bandwidth, delivered energy, total duration or a reported Pareto frontier, and the same noise/drift conditions. Include reservoir pumps, detector illumination, and repeated attempts in the full energy ledger.
The optimized coherent class must include the compensating Hamiltonians used by the proposed method. Otherwise extra controls would be mistaken for a dissipative advantage. No method can outperform an optimizer over a strict superset of its own allowed controls by definition; a claim of “better than all coherent control” therefore needs a precise resource and noise distinction.
Reverse or retrace the word, permute its order, scramble the tangent correction, switch off the engineered reservoir, deliberately perturb tag couplings, and retarget all three sites without rebuilding. Hold spectral power and integrated dose comparable where possible. Independently measure the chemical gate's background conversion with no authorization.
Quantitative milestones
Stage 1 requires selectivity above $10^2$, target conversion at least 10%, retargeting of all three sites, and a reported sustainable repetition rate. Stage 2 aims for $10^4$ while preserving useful yield and accounting for false acceptance and background chemistry. Stage 3 aims for $10^6$ only if independently measured background and alias floors permit it. These are experimental goals, not predictions; the present model does not support Stage 3 under its chosen background rates.
For zero observed off-target events, a rough one-sided 95% Poisson upper limit is approximately three expected events. Demonstrating a $10^{-6}$ per-opportunity error therefore needs millions of relevant off-target opportunities and adequate detection efficiency, not a handful of successful cycles. Selectivity is target events divided by summed off-target events under the same exposure accounting.
Measure chemical identity and conversion, off-target events, false rejections, local temperature, supplied energy, reservoir reset cost, drift, cross-talk, cycle time, chemical reversibility, and deterioration over at least a predeclared repeated-cycle qualification run. Initial aims of $10^3$ cycles and subsequent $10^5$ cycles are qualification proposals, not demonstrated durability. The time unit of the numerical controller is not assigned to physical seconds until measured couplings and coherence justify it.
The hard molecular interface requirement
The molecular state must change a measured controller or sensor response enough to distinguish valid preparation from wrong preparation within the allowed illumination and damage budget. A cavity shift smaller than drift and shot noise is not a usable certificate. If a strong-coupling formulation is used, its coupling, linewidths, and molecular dephasing must be measured; writing a cooperativity symbol does not establish sufficient interaction. Failure to achieve this state-to-controller interface at useful speed would leave only a photonic addressing system, not PZHRD chemistry.