Renormalization That ClicksEpisode 7 (Finale) / Resolution as a dimension

See coarse-graining not as "a procedure for discarding information" but as "moving one dimension through space," and two claims become one

Resolution as a dimension In AdS/CFT the radial direction corresponds to the renormalization scale itself.
Going deeper is coarse-graining.
Then "the universe is coarse-grainable" and "the universe is holographic" become two descriptions of one claim.

Tools you'll need: Episode 4's RG flow, Episode 1's coarse-graining, a feel for the holographic principle The heart of this episode: radial coordinate z ↔ renormalization scale 1/E

Episode 4 said "repeat coarse-graining and a theory flows." If something flows, it must be flowing along something. That something we have been calling the axis of resolution ── fine or coarse. What this finale wants to claim is exactly one thing: that axis may not be a metaphor. In certain theories, the axis of resolution appears as a genuine dimension of space. Coarse-graining a theory that lives on a boundary corresponds precisely to going deeper into the interior ── this is holographic renormalization in AdS/CFT. And if that is right, then what this series has said all along, "the universe is coarse-grainable," and Episode 5 of "Black Holes That Click," "the universe is holographic," are one and the same thing said two ways.

01If it flows, it flows along something

Picture the renormalization group flow once more. Start from the finest description (UV: microscopic); each coarse-graining changes the theory, and it flows toward the coarse description (IR: macroscopic).

In that picture the horizontal axis was "the content of the theory" (the couplings). So what is the vertical axis? ── It is how coarsely you are looking: the resolution. Physicists call it "scale" and normally treat it as a convenient bookkeeping axis. Nobody thought it was a real thing.

02Draw that axis as a real dimension ── AdS/CFT

The holographic correspondence (AdS/CFT), which appeared in 1997, asserts the following astonishing equality.

The heart of this episode ── depth is resolution
$$\underbrace{d\text{-dimensional field theory (boundary)}}_{\text{no gravity}}\ \equiv\ \underbrace{(d+1)\text{-dimensional gravity (bulk)}}_{\text{one extra dimension}}$$

And what the extra dimension ── the bulk's radial coordinate \(z\) ── corresponds to is the boundary theory's renormalization scale:

$$z\ \longleftrightarrow\ \frac{1}{E}\qquad \begin{cases}z\to 0\ \text{(near the boundary)} & \text{UV = looking finely}\\[2pt] z\ \text{large (deep)} & \text{IR = looking coarsely}\end{cases}$$

In other words, going deeper = coarse-graining. The operation of growing the block size in Episode 1 becomes, in this theory, motion through space.

This is not a hunch but a correspondence confirmed by calculation. The renormalization group equations of the boundary theory take the same form as the equations of motion the bulk gravitational fields obey in the \(z\) direction (holographic renormalization). Removing UV divergences corresponds exactly to cutting off a divergent volume near the boundary in the bulk. The bookkeeping axis called "scale" turned out, from the gravity side, to be a real distance.

Another piece of evidence ── tensor networks

There is a computational method for quantum many-body systems called MERA. It is a network built by stacking layer upon layer of "coarse-graining" ── essentially a picture of the renormalization group. In 2009 Swingle pointed out that the shape of that layered structure looks just like a lattice-discretised AdS spacetime. The "number of coarse-graining layers" is itself the depth. A diagram invented purely for reasons of information processing turned out to have the same shape as the spacetime of gravity ── one of the reasons people take "resolution may be a dimension" seriously.

03Try it ── dive into the depth and the resolution drops

The top edge of the figure is the boundary (the finest, raw information). Downward is deeper into the bulk. Each horizontal row draws the boundary pattern coarse-grained by an amount set by that depth. In other words, this whole picture is the renormalization group flow.

Move the slider to change the depth \(z\). Dive and the pattern blurs ── that is what "coarse-graining" looks like in this theory. A triangular wedge is also drawn, expressing the relation "if you know this stretch of the boundary you can reconstruct the bulk down to about this depth" (roughly half the width). The wider the region you take in at once, the deeper you can go.

Figure: top edge = the boundary (raw information); downward = deeper into the bulk = coarser description. Each row is the boundary pattern coarse-grained by an amount set by that depth. The triangular wedge expresses the correspondence "a boundary region of width W covers down to depth ≈ W/2." Depth = the axis of resolution
boundary information (shading) selected depth z boundary region and the bulk wedge it covers

04And then two claims become one

Now the question that opened the series gets an answer ── why is the universe coarse-grainable?

Seen from the holographic side, that becomes the same question as "why does the universe have depth?" Because being coarse-grainable and having an extra dimension are two faces of the same fact.

In the language of coarse-grainingIn the language of holography
lower the resolutiongo deeper into the bulk
the renormalization group flowthe equations of motion of gravity in the \(z\) direction
irrelevant operators die away (Ep. 4)fields decay as you go deeper
relevant operators grow (Ep. 6)fields grow with depth and deform the spacetime
the amount discarded = entropy (Ep. 1)area ÷ 4 = entropy (Black Holes, Ep. 2)
the answer doesn't change if you look coarselythe information in a volume can be written on its surface

That last row is the holographic principle. The information in a three-dimensional interior fits on a two-dimensional surface ── the claim we met in Episode 5 of "Black Holes That Click." Reread under this dictionary, it says that the holographic principle is the geometric expression of the universe's coarse-grainability.

The same shape as the finale of "Relativity That Clicks" The finale of the sister series "Relativity That Clicks" said: draw gravity as "curved spacetime" or as "a speed of light that varies from place to place" and the observables are identical ── descriptions are interchangeable. This episode has exactly the same shape ── describe it as "a procedure for discarding information" or as "moving one dimension through space," and it is one and the same thing said two ways. Numbers with units, choices of coordinates, and even the bookkeeping axis called "resolution" are stage machinery; what remains are the relations. Four series arrive at the same place here.

05The universe isn't slacking off ── the series' conclusion

Finally, an answer to the question this series started from. Since the universe has finite computational resources, doesn't it skip the hard calculations and settle for probabilities?

The resources really are finite. By Lloyd's estimate the observable universe has performed about \(10^{120}\) operations since the Big Bang using about \(10^{90}\) bits, and the holographic bound is \(10^{122}\) bits. The bound is real. But what these seven episodes have shown is that the bound does not appear as a reason to slack off.

The series' conclusion ── who exactly is slacking off

The universe isn't slacking off. We are the ones with no choice but to.

The phase does not vanish; it is diluted into the environment (Ep. 3). Decay is only a recurrence time stretched long; the equation stays unitary (Ep. 2). If the universe really rounded things off, quantum computers wouldn't work and echoes wouldn't bring the phase back. The books balance all the way down. It is only that the cost of tracking exceeds the universe's computational capacity, so we must write probabilities. And what guarantees that "writing probabilities gives the same answer" is the renormalization group and coarse-graining ── a kindness on the universe's side. The finite resource bound shows up not as an excuse for skipping work but ── as the nature of a black hole.

◇ ◇ ◇
The honest line ── this is the edge of the map

Established: the AdS/CFT correspondence (Maldacena 1997) and its vast body of checks in concrete examples; that the bulk radial coordinate behaves as the renormalization scale of the boundary theory (UV/IR correspondence, holographic renormalization); that the boundary RG equations can be written as bulk equations of motion; the resemblance between MERA's layered structure and a discretised AdS spacetime (Swingle 2009); the holographic principle and black-hole entropy \(S=A/4\); and Lloyd's estimates of the universe's computational capacity (\(\sim10^{120}\) operations, \(\sim10^{90}\) bits) ── all established theoretical results or standard understanding.

Beyond here is not established: (1) AdS/CFT is rigorously formulated for anti-de Sitter spacetime (negative cosmological constant), and our universe is not that (the cosmological constant is positive). Holography for de Sitter-like universes is an active research subject but is not established. So we cannot yet say that the depth of our universe is resolution. (2) "Coarse-graining = moving through space" is a precise statement inside the framework where the correspondence holds, not a general theorem for arbitrary theories. (3) The MERA/AdS relation is a suggestive resemblance; no rigorous equality has been established. (4) The figure is a schematic for feeling "depth = resolution," not a solution of the AdS metric or its geodesics (the wedge depth ≈ width/2 is likewise a rough relation based on circular causal wedges). ── This episode is the farthest foothold the series has reached, not a settled conclusion.

Exercises (solvable with this episode's ideas)
  1. In AdS/CFT, what does the bulk radial coordinate \(z\) correspond to on the boundary? What are \(z\to0\) and large \(z\)?
    See the answer
    The renormalization scale (\(z\leftrightarrow 1/E\)). \(z\to0\) (near the boundary) is UV = looking finely; large \(z\) (deep) is IR = looking coarsely. Going deeper = coarse-graining.
  2. Why can "the universe is coarse-grainable" and "the universe is holographic" be the same claim?
    See the answer
    Because the hierarchy of coarse-graining (the axis of resolution) is realised as an extra dimension. That the information in a volume fits on its surface (holography) says that the information in the depth direction is exhausted by the boundary's hierarchy of resolutions.
  3. What do Episode 4's relevant/irrelevant become in bulk language?
    See the answer
    Irrelevant operators = fields that decay with depth. Relevant operators = fields that grow with depth and deform the spacetime. Episode 6's "it leaks in exactly two places" translates to "exactly those two run wild in the interior."
  4. State this series' answer to "the universe skips the calculation" in one sentence.
    See the answer
    The universe isn't slacking off. Phase and information are conserved (quantum computers and echoes are the evidence); we are the ones who gave up tracking. But what guarantees that giving up yields the same answer is the renormalization group ── and that is what divides the world into layers.

Episode 7 summary / Renormalization That Clicks, completeCoarse-graining was, perhaps, a direction in space

The renormalization group was a "flow." If it flows, it flows along something ── we called that axis resolution, and in AdS/CFT it appears as a genuine dimension of space. The bulk radial coordinate \(z\) corresponds to the boundary theory's renormalization scale (\(z\leftrightarrow1/E\)), and going deeper = coarse-graining. The boundary's RG equations can be written as the equations of motion of bulk gravity in the \(z\) direction. That MERA's layers resemble a discretised AdS spacetime says the same thing from another angle.

From this viewpoint a dictionary can be drawn ── irrelevant/relevant become decay/growth with depth; the amount discarded (entropy) becomes area ÷ 4; and "the answer doesn't change if you look coarsely" becomes "the information in a volume can be written on its surface." The holographic principle may be the geometric expression of the universe's coarse-grainability. Exactly the same shape as the finale of "Relativity That Clicks" ── descriptions are interchangeable; only the relations stay put.

And the series' conclusion. The universe isn't slacking off. We are the ones with no choice but to. The phase doesn't vanish, it dilutes; the books balance. It is only that the tracking cost exceeds the universe's computational capacity (\(10^{120}\) operations, \(10^{90}\) bits), so we write probabilities. And what guarantees that "writing probabilities gives the same answer" is the renormalization group, a kindness on the universe's side. The finite resource bound appears not as a reason to skip work but as the nature of a black hole ── from there, go to "Black Holes That Click."

This document is Episode 7 (the finale) of the "Renormalization That Clicks" series, a reading piece for physics-loving high-schoolers and undergraduates. The AdS/CFT correspondence (Maldacena 1997); the correspondence between the bulk radial coordinate and the boundary renormalization scale (UV/IR correspondence: Susskind–Witten; holographic renormalization: de Boer–Verlinde–Verlinde and others); the correspondence between boundary RG equations and bulk equations of motion; the resemblance between MERA tensor networks and discretised AdS spacetime (Swingle 2009); the holographic principle and Bekenstein–Hawking entropy \(S=A/4\); and Lloyd's estimates of the universe's computational capacity ── all established theoretical results or standard understanding. That AdS/CFT is rigorously formulated only for anti-de Sitter spacetime with a negative cosmological constant and that extending holography to our positive-cosmological-constant universe is unestablished; that "coarse-graining = moving through space" is a statement inside the framework where the correspondence holds rather than a general theorem; and that the MERA/AdS relation is a suggestive resemblance rather than a rigorous equality ── all of this is spelled out in the body's "honest line." The figure is a schematic for feeling "depth = resolution," not a solution of the AdS metric or its geodesics. ── To print, use your browser's "Print" and "Save as PDF" (in the print version the slider and answers are frozen and hidden). Previous: Episode 6, The two places it leaks / bonus: Why did it start from low entropy? and An observer is a coarse-graining device / Contents / sister series Black Holes That Click and Relativity That Clicks.

Print / make a PDF: ⌘+P (Ctrl+P on Windows). On screen, moving the slider changes the depth z and the resolution drops. "Wedge" shows the correspondence between a boundary region and the bulk. "See the answer" opens each solution.