Black Hole OLE, part 6: what the result proves and what comes next

The best conclusion is neither "nothing happened" nor "quantum advantage is proven." The result is strong because its boundaries are visible.

What has been established

First, a tracker-compatible 80-qubit OLE circuit family was constructed reproducibly from a released 70-qubit source. The extension preserves the observable, perturbation support, fixed core, and edge-color structure while adding ten connected Kingston sites by a declared rule.

Second, all 16 sampled circuits completed successfully on IBM Kingston through Fire Opal. The paired finite-sample estimate produced

\[
R_{\mathrm{OLE}}=0.74028847\pm0.01663657.
\]

Third, the classical tracker-linked BP-TN runner was validated against an earlier small artifact before Q80 was attempted.

Fourth, the classical convergence ladder exposed the real difficulty. Low bond dimensions produced answers quickly but did not agree. BD64 failed to finish even the delta half within 900 seconds.

Together, these facts support evidence of a local practical runtime advantage for this exact workflow.

What has not been established

Q80 is not an official released tracker instance.
Eight initial states are not the intended 500-state tracker-scale estimate.
The experiment is not full-state tomography.
The mitigation is not a dense correlated inversion over all 2^80 outcomes.
The BD64 timeout is not proof against every possible classical algorithm or computer.
Echo decay is not by itself proof of genuine scrambling.
The processor did not simulate a black-hole horizon or emit Hawking radiation.

These are not weaknesses to hide. They define the next experiments.

The next tracker step

The obvious statistical extension is to increase N_init. If the observed sample variance remained stable, moving from 8 to 500 samples would reduce the sampling standard error by roughly

\[
\sqrt{\frac{8}{500}}\approx0.126.
\]

The present ratio error of about 0.0166 would then fall toward 0.0021 from sampling alone. A run that large should be batched carefully because calibration drift and cross-batch variation may become more important than the simple square-root law.

The next classical step

The strongest follow-up is a higher-resource tensor-network campaign:

complete both delta and delta zero at BD64;
continue to BD128 if the values remain unstable;
record peak memory and full wall time;
test optimized Julia settings and a stronger CPU or GPU platform;
compare alternative contraction and operator-propagation methods.

That campaign may weaken the runtime claim, strengthen it, or reveal a useful accuracy-runtime frontier. Any of those outcomes would improve the science.

The next scrambling step

To sharpen the black-hole-information connection, future circuits need controls that distinguish operator spreading from ordinary noise:

compare perturbations whose support overlaps or is initially disjoint from the observable;
vary circuit depth and perturbation strength;
test recovery-style or anti-butterfly controls;
compare the measured trend with exact small-system references;
study whether renormalized echo protocols remain stable under imperfect time reversal.

These tests move the project from an OLE hardware benchmark toward a controlled scrambling experiment.

Hawking's legacy in this project

Hawking remains a hero here because he showed what ambitious theoretical physics can do. A calculation about quantum fields near a horizon transformed black holes from perfectly dark endpoints into thermodynamic quantum objects. The information problem that followed forced physicists to rethink locality, entropy, unitarity, and the meaning of recoverability.

Our 80-qubit experiment is modest beside that achievement. It does not reproduce gravity. It does something worthwhile on its own scale: it turns part of the information-scrambling language into a transparent, reproducible hardware measurement.

That is the right ending for this series and the right beginning for the next experiment.