Electromagnetism is a circle too small to see
A post shared a five-hour lecture saying electromagnetism looks like the trace of a fifth dimension. It does, and exactly: give space one extra direction curled into a tiny circle, and a wave that winds round it once behaves like an electron’s field in a magnetic field. The lab checked the dictionary, found where the simplest version stops being our world, and built a game out of the part you can feel.
The verdict
The claim holds
Electromagnetism is exactly the geometry of a small fifth circle: charge is how many times a wave winds round it, and a gauge change is a relabelling of the circle. The simplest version is not our world: its charged particles are about 10²¹ too heavy for their charge, it cannot make left-right asymmetric matter, and its own vacuum energy pulls the circle shut.
Wrap a wave round the small circle n times and, seen from ordinary space, it carries charge n. Its levels match a charged particle’s to 1e-13.
Slide where the circle’s zero sits, by a different amount at every point. The vector potential changes to match, and no measured level moves (4e-12).
Every particle the circle makes has the same charge-to-mass ratio, about 10²¹ smaller than the electron’s. A plain circle also cannot make left-right asymmetric matter.
The vacuum energy of fields on the circle falls as 1/L⁴ as it shrinks. That is the one real link to the zero-point reply, and it is a problem for the theory.
The claim, as presented. The post presents electromagnetism, seen as a gauge theory, as looking like the trace of a fifth dimension (Kaluza-Klein theory). Its title card shows the dictionary: shifting the circle coordinate shifts each mode's phase and the vector potential together.
Richard Behiel, “Electromagnetism as a Fifth Dimension (Kaluza-Klein)”, October 2026 · the episode
The registry’s note on sources. This claim comes from the post on X that shared the video (2026-10-08, read from a screenshot) and from the video's title card, which shows the equations. The five-hour video itself was not watched, and the date given is the post's, not the upload's. A reply under the post said the idea is called zero-point energy; the page addresses that reply without naming its author.
The reply. A reply under the post said the idea is called zero-point energy. The two are different things. Zero-point energy is the energy fields keep in their lowest state; the fifth dimension is a statement about geometry and charge. Section 02 shows where they meet.
Preregistered in PREREG.md (commit 305fb80) before any code. One run, no reruns. Self-graded: each gate has a preregistered way to fail, and all three were seen red. Lab files are listed at the foot of the page.
Explore: a wave wrapped round a circle
Measured livePicture a garden hose seen from far away: it looks like a line. Up close, every point on the line is a little circle. Kaluza (1921) and Klein (1926) gave space one such circle and found Maxwell’s equations inside Einstein’s.
A wave on the hose can wind round the circle 0, 1, 2 or more times as it travels. Seen from the line, that winding number is the electric charge, and the energy it costs to wind is the particle’s mass, n/R for a circle of radius R.
The title card’s equations say one more thing. Where the circle’s zero sits is a choice nobody can see. Move it by ε, and a wave with winding n shifts its phase by nε let ε vary along the line, and a vector potential A appears to match. That relabelling is a gauge transformation. Try it below: the page builds the wave’s lattice operator before and after the relabel, finds every level, and prints the biggest change.
The lab ran the same check at four fluxes and three windings: the lattice matched the charged-particle formula to 9e-14, and a random relabel moved no level by more than 4.1e-12. Your browser’s numbers come from its own 40-site lattice, so they land near 1e-13 too.
Correct: three places the simple version breaks
Lab numbersThe mathematics holds. The question is whether our world is that circle, and the plain version says no three times. None of this is news: it is why physicists moved on to richer shapes in string theory.
The electron does not fit. Every particle the circle makes has the same ratio of charge to mass, set by gravity’s strength. The electron’s ratio is 1.02 × 10²¹ times bigger. Shrink the circle until its first particle carries exactly the electron’s charge (23 Planck lengths, 3.8 × 10⁻³⁴ m) and that particle weighs 5 × 10¹⁷ GeV. The electron weighs 0.000511.
No handedness. The weak force treats left-handed and right-handed particles differently. On a plain circle every massive particle comes as a left-right pair, and so does any massless one. Witten showed in 1981 that this problem is general. Folding the circle in half (an orbifold) is one standard way out.
The circle wants to close, and this is where the reply comes in. Fields on a circle keep zero-point energy, and how much depends on the circle’s size. For one massless field it is −3ζ(5)/(4π²L⁴) per unit volume (Appelquist and Chodos, 1983): negative, and more negative as the circle shrinks. So the vacuum pulls the circle shut. The reply’s zero-point energy is real, but here it is a problem the theory has to solve, not the theory itself.
The dial uses the textbook normalisation of charge (q = n√(16πG)/R). A factor of 2 either way moves the 10²¹ by about a third of a power of ten, which does not change the answer.
Play: hide a flux in a ring
Measured liveBend the line into a ring and thread a magnetic flux through the middle. The field along the ring is drawn as ticks, lumpy on purpose. A charged wave’s energy levels can only feel one number: the loop total, Φ. That is the Aharonov-Bohm effect, and in the five-dimensional picture it is just how far the circle’s zero twists on a trip round the ring.
- 1Turn the flux up, then make charge 1 blind to it again.
- 2Find a flux charge 2 cannot see but charge 1 can.
- 3Reshape the field three times without moving a level.
- 4Switch to charge 1.3 and show the trick from goal 1 stops working.
The answers, if you want them. Charge 1 is blind at Φ = 2π, charge 2 at Φ = π as well. A charge of 1.3 would only be possible if the extra dimension were not a circle, and then 2π no longer hides anything. Goal 2 caught the lab out: its preregistered gate expected charge 2 to see Φ = π, and it does not.
Every gate, including the two that failed
| Gate | Result | Numbers |
|---|---|---|
| G1 the dictionary | pass | lattice vs charged-particle formula: worst 9.0e-14 (charge 1 and 2, four fluxes). Continuum, lowest 8 levels: 2.9e-4 |
| G2 a gauge change is a relabel | pass | random relabel, same flux: largest level change 4.1e-12 (charges 1, 2, 3) |
| G3 only the loop counts, as written | failed | Φ and Φ + 2π agree to 4e-12 (pass). Charge 1 sees Φ = π (shift 16.1). Charge 2 does not (3e-12): the gate was wrong, not the physics |
| G3 red a charge of 1.3 | red, as intended | Φ and Φ + 2π differ by 9.65 |
| G4 the electron dial | pass | gap 1.02e21 (allowed 1e20 to 1e22); first charged mass 0.043 Planck masses |
| G5 handedness on a circle | pass | left = right at every twist tried |
| G5 red the orbifold | red, as intended | left 1, right 0 |
| G6 vacuum energy, 1 % at ε/L = 0.05 | failed | 1.81 % low at ε/L = 0.05, 0.455 % at 0.025: the error quarters when the cutoff halves. Exponent −4.0000, negative at every size |
| G6 red skip the subtraction | red, as intended | blows up 1024× as the cutoff shrinks 4× |
The failures, kept on the record. G3 asked every integer charge to notice a flux of π. A charge-2 wave goes round the loop picking up twice the phase, so it repeats every π, not every 2π. G6’s 1 % tolerance was a guess about what the cutoff leaves behind, and the leftover turned out to be 1.8 %, shrinking cleanly toward the formula. Both are mistakes in the gates, found by the run, and neither moves the verdict. Nothing was rerun with friendlier settings.
- The five-hour video was not watched. The claim on this page is the post’s and the title card’s, and the date is the post’s.
- The charge normalisation is the textbook one, not rederived here.
- Witten’s general result was not reproduced; only the circle and its orbifold escape were counted.
- The vacuum energy is for one massless scalar. Gravitons and fermions change the number, and fermions can flip its sign; real models are about the total.
- Kaluza, Klein, Witten, Appelquist-Chodos and Aharonov-Bohm were found by search, not read. The Overduin-Wesson review was read at abstract level.
- The verdict is self-graded. No outside review was run.