c·t = CONST, THAT CLICKS EPISODE 23 ／ The same number, read inside out
An occupancy of \(10^{-18}\) is, in the language of codes, a redundancy of \(10^{18}\)
The horizon as
error correction
Is it empty, or is the same information written over and over?
Until you name the comparison, the two cannot be distinguished.
Episode 18 counted that "volume cells were never given addresses". So how are the \(10^{122}\) bits written on the horizon protected? Today we reread them in the language of codes. Episode 6's "occupancy \(1.5\times10^{-18}\)" then reappears wearing a completely different face — a redundancy of \(6.6\times10^{17}\). Not empty, but the same information written over and over.
01Reading it as a code
An error-correcting code is characterised by two numbers — physical bits \(n\) (the storage elements actually used) and logical bits \(k\) (the information to be protected). Apply them to the universe.
physical bits (writable on the horizon, Episode 1)
$$n=2.96\times10^{122}$$logical bits (the entropy actually in use, Episode 6)
$$k=\frac{S_{\rm obs}/k_B}{\ln2}=\frac{3.1\times10^{104}}{0.693}=4.47\times10^{104}$$code rate and redundancy
$$R=\frac{k}{n}=1.51\times10^{-18},\qquad \frac{n}{k}=6.61\times10^{17}$$Conclusion of §01
Episode 6's "occupancy \(1.5\times10^{-18}\)" is, in the language of codes, "redundancy \(6.6\times10^{17}\)".
The same number, read the opposite way round.
02Comparing with real codes
| Code | Redundancy \(n/k\) | Note |
|---|---|---|
| QR code (highest level) | 1.4 | recovers up to 30% damage |
| RAID6 | 1.5 | survives two disk failures |
| Low-rate channel codes | 10 | deep space communication and the like |
| Quantum error correction, surface code | \(10^{3}\) | 1000 physical qubits per logical one |
| The cosmic horizon | \(6.6\times10^{17}\) | \(6.6\times10^{14}\) times the surface code |
Fourteen orders more redundant than any code humans build. Where quantum error correction is lamented for needing 1000 physical qubits per logical one, the universe uses \(10^{18}\).
03Estimating the correction capability
classical code
$$d\le n-k+1=2.96\times10^{122}\qquad(\text{essentially }n\text{ itself})$$quantum code
$$d\le\frac{n-k}{2}+1=1.48\times10^{122}=\frac{n}{2}$$Conclusion of §03
In principle — about half the horizon's bits could be destroyed and the contents still recovered.
(This is an upper bound; there is no guarantee that the universe is such a code.)
Figure: redundancies compared (log). Human codes cluster at the left; only the universe is orders to the right. The slider moves the estimate of the entropy in use — \(S_{\rm obs}\) has order-of-magnitude uncertainty, and the redundancy moves with it.
04The heart — empty, or redundant?
| Reading | What it says | From |
|---|---|---|
| ① Empty | only \(1.5\times10^{-18}\) of the capacity is used | Episode 6 |
| ② Redundant | the same information is written \(6.6\times10^{17}\) times over | today |
Can observation tell these apart? — Episode 3's procedure applies directly.
The thing this episode most wants to say
"Empty" and "redundant" cannot be distinguished until you name the comparison.
Against the capacity it is empty; against the information it is redundant — two readings of one ratio.
Episode 3 applied this surgery to "\(c\cdot t\) is constant", Episode 9 to "atoms are shrinking", Episode 12 to "the cosmological constant". Today it applies to a number this series produced itself — saying "it is empty" without naming the comparison is not yet a sentence.
05Holographic codes as a precedent
"Protect the bulk with boundary information" is not a whim. AdS/CFT admits a reading as a quantum error-correcting code (Almheiri, Dong & Harlow 2015).
So what we can do here is translate numbers into the language of codes, and no further — we cannot say "the universe is in fact an error-correcting code". That line is drawn clearly.
06Restraint — the size of one logical bit
Episode 18 produced the coincidence "the side of a physical bit's volume is 1.96 fm — the size of a proton". Do the same for a logical bit.
Following Episode 19's practice, check whether it matches anything.
| Candidate | Length | Ratio |
|---|---|---|
| Proton | \(10^{-15}\) m | \(2.2\times10^{-11}\) |
| Electroweak scale | \(2.5\times10^{-18}\) m | \(8.8\times10^{-9}\) |
| Grand unification scale | \(2\times10^{-32}\) m | \(1.1\times10^{6}\) |
| Planck length | \(1.6\times10^{-35}\) m | \(1.4\times10^{9}\) |
Conclusion of §06
It matches nothing. So there is nothing to say.
── Unlike Episode 18's 1.96 fm, this number has no near neighbour. Silence is the right answer.
This is exactly why Episode 19 built the sorting procedure. Not every number a calculation produces means something — with no near neighbour, the surprise is 0 bits and there is nothing to say.
① This does not claim "the universe is an error-correcting code". All it does is translate two numbers from Episodes 1 and 6 into the language of codes. The reading of AdS/CFT as quantum error correction (Almheiri, Dong & Harlow 2015) concerns the AdS boundary, and whether a cosmological horizon has the same structure is unsettled (de Sitter/FLRW holography is not established).
② Setting \(k=S_{\rm obs}/\ln2\) is a crude identification. Reading thermodynamic entropy as "logical bits to be protected" is not obvious — entropy is arguably a measure of lost information, so one could argue for the opposite reading. And \(S_{\rm obs}\) itself has order-of-magnitude uncertainty (Episode 6 ①), which the slider illustrates.
③ The Singleton bound is an upper bound; achievability is another matter. "Half the horizon could be destroyed and still recovered" means such a code could exist, not that the universe is one.
④ The surface code's \(10^3\) is indicative, moving between \(10^2\) and \(10^4\) with the physical error rate and the target logical error rate.
⑤ §04's "empty or redundant" is this series' reading. It is not a rigorous demonstration that the two readings are observationally equivalent — only the procedural point that without naming a comparison, no distinction can be drawn.
Exercises (solvable with this episode's formulas alone)
- Find the code rate and redundancy of the cosmic horizon read as a code.
Show the answer
Physical bits \(n=2.96\times10^{122}\), logical bits \(k=3.1\times10^{104}/\ln2=4.47\times10^{104}\). Rate \(R=k/n=1.51\times10^{-18}\), redundancy \(n/k=\) \(6.6\times10^{17}\) — the flip side of Episode 6's occupancy. - How much more redundant is it than the quantum surface code?
Show the answer
The surface code is around \(10^3\), so \(6.6\times10^{17}/10^3=\) \(6.6\times10^{14}\) times — fourteen orders beyond any human code. - State the two readings of \(1.5\times10^{-18}\) and say whether they can be distinguished.
Show the answer
① "only \(1.5\times10^{-18}\) of the capacity is used (empty)" and ② "the same information is written \(6.6\times10^{17}\) times over (redundant)". They cannot be distinguished until you name the comparison — against capacity it is empty, against information it is redundant. Episode 3's surgery, applied to the series' own number. - Find the side of the area one logical bit occupies, and check whether it matches anything.
Show the answer
\(A/k=4.79\times10^{-52}\ \mathrm{m^2}\), side \(2.19\times10^{-26}\) m — eleven orders smaller than a proton, nine orders larger than the Planck length. It matches nothing. By Episode 19's practice the surprise is 0 bits and there is nothing to say. Not every number a calculation produces means something. - (Harder) Can we say "the universe is an error-correcting code"?
Show the answer
No. The formulation of AdS/CFT as a quantum error-correcting code concerns the AdS boundary, and whether the same structure holds for a cosmological horizon is unsettled (de Sitter/FLRW holography is not established). This document goes only as far as translating two numbers into the language of codes.
Summary — the same number, read inside out
We read the horizon as a code: physical bits \(n=2.96\times10^{122}\), logical bits \(k=4.47\times10^{104}\), rate \(R=1.51\times10^{-18}\). Inverted, a redundancy of \(6.6\times10^{17}\) — the flip side of Episode 6's occupancy. Fourteen orders beyond any human code, dwarfing the quantum surface code (\(10^3\)). By the Singleton bound, in principle half the horizon's bits could be destroyed and the contents recovered (as an upper bound only).
The heart was the reading. The same \(1.5\times10^{-18}\) can be read as "empty" or as "redundant" — against capacity, empty; against information, redundant. Until you name the comparison, the two cannot be distinguished. The surgery applied to \(c\cdot t\) in Episode 3, to atoms in Episode 9 and to the cosmological constant in Episode 12 has now landed on a number this series produced itself.
There is a precedent: AdS/CFT read as a quantum error-correcting code, where "lose part of the boundary and the bulk centre survives" is the definition of a code. But whether the same holds for a cosmological horizon is unsettled, so all this episode could do was translate numbers into the language of codes. That line is drawn clearly.
And one act of restraint at the end. The side of the area one logical bit occupies is \(2.19\times10^{-26}\) m — checked by Episode 19's practice, it matches nothing. Unlike Episode 18's 1.96 fm, it has no near neighbour. Not every number a calculation produces means something, and with no near neighbour there is nothing to say — which is why the sorting procedure was built.
This document is Episode 23 of "c·t = const, That Clicks", written for physics-minded high-school and university readers. The code rate \(R=k/n\) and the Singleton bound (classical \(d\le n-k+1\); quantum \(d\le(n-k)/2+1\)) are standard. The values \(n=2.96\times10^{122}\) (Episode 1), \(k=S_{\rm obs}/\ln2=4.47\times10^{104}\) (Episode 6, from \(S_{\rm obs}=3.1\times10^{104}k_B\), Egan & Lineweaver 2010), the rate \(1.51\times10^{-18}\), the redundancy \(6.61\times10^{17}\), and the area per logical bit \(4.79\times10^{-52}\ \mathrm{m^2}\) (side \(2.19\times10^{-26}\) m) are computed here (kenshou/calc27.py). This document does not claim that the universe is an error-correcting code — the formulation of AdS/CFT as quantum error correction (Almheiri, Dong & Harlow 2015) concerns the AdS boundary, and whether a cosmological horizon has the same structure is unsettled (de Sitter/FLRW holography is not established). Setting \(k=S_{\rm obs}/\ln2\) is a crude identification, and reading thermodynamic entropy as "logical bits to be protected" is not obvious (entropy is arguably a measure of lost information). \(S_{\rm obs}\) carries order-of-magnitude uncertainty, which the slider illustrates. The Singleton bound is an upper bound, distinct from achievability. The surface code's \(10^3\) is indicative and moves between \(10^2\) and \(10^4\). §04's "empty or redundant" is this series' reading, not a rigorous demonstration of observational equivalence. Linear expansion (\(c\cdot t=\)const, \(R_h=ct\)) is a minority model under examination. The academic standard remains the \(\Lambda\)CDM model including inflation. ── To make a PDF, use your browser's Print dialogue (sliders freeze and answers are hidden in the print version).