S_dS over pi squared replaced a number I made up

This post is the story of a replacement: one number in the framework’s ledger replaced by another, with the replacement derived and the original archived. It is the cleanest before-and-after in the whole corpus, and it is also a confession about how little of a holographic budget the old number was actually using.

The number that was replaced

The framework’s horizon paper once claimed that the horizon’s computational capacity is D C to the 72, an exponent built from the category’s numbers, and that this matches the holographic bound. The claim is archived now, and the corrections series carries the full autopsy, but the number that replaces it deserves its own post, because a replacement is only convincing if you can see what was wrong at the level of the units.

D C to the 72 was a dimensionless construct: the category’s quantum dimension raised to a power chosen so the result would land near 10 to the 122. Its units were wishes. The derived number has units, or rather it has a derivation whose every factor is accounted: entropy divided by pi squared.

The derivation

Start from the de Sitter horizon entropy, S d S equals pi times c to the 5 over G hbar H squared, about 2.27 times 10 to the 122 nats for the measured Hubble constant. The claim to derive is that the number of operations per Hubble time is S d S over pi squared, and the derivation is the three-bounds agreement from the walkthrough series: Lloyd’s kinematic bound, the Bekenstein geometric bound, and the Landauer thermodynamic bound all return the same expression, and the coincidence has a mechanism, the shared horizon geometry.

But the post-derivation audit asked a second question, the one that finished the old number: how much of the holographic budget was D C to the 72 actually using? The holographic bit budget is S d S over ln 2, about 3.3 times 10 to the 122 bits. Take the logarithm of the old claim, D C to the 72 is 50.91 to the 72, about 10 to the 122, so its logarithm is about 405 bits in base 2. Compare: 405 bits of budget used against 3.3 times 10 to the 122 available. The fraction is one over 10 to the 120.

That is the sentence that killed the old claim permanently: it was not wrong-sized, it was microscopic. A capacity derivation that uses a trillionth of a trillionth of the available budget does not explain the budget; it decorates it. The framework’s own selection-principles document now carries this as the standard test for capacity claims: compute the fraction of the budget your number actually consumes, and if the fraction is below 10 to the minus 20, you have not derived anything, you have found a small integer with a logarithm.

What the new number commits to

S d S over pi squared is now the framework’s only capacity claim, labelled benchmark, meaning: derived from standard physics, consistent across three independent bounds, and saying nothing whatsoever about whether the horizon computes anything. Its uses are negative: it caps what simulations can render, it disqualifies numerologies that underuse the budget, and it fixes the operations-per-bit ratio at ln 2 over pi squared for every horizon, large or small.

The meta-lesson generalises beyond this archive, and it is the reason this post exists separately from the corrections entry. Numbers in physics come with a denominator attached: not just how big is your claim, but how much of the phenomenon does it account for. A theory whose key number is 10 to the minus 120 of the quantity it claims to explain is not a conservative theory. It is a theory that has mistaken the spare change in the couch for the economy. The new number is less impressive and more honest, which in this archive’s currency is a strict upgrade.

The next post is where the last surviving numerical match was found, hiding in a calculation everyone had read: the instanton exponent.