# QEC in AdS/CFT: Quantum Error Correction, Holographic Codes, and the Simulation Question

**Argus Research Thread** | Session 3 | 2026-09-08
**Evidence Class Legend:** Established (E) | Serious Speculation (SS) | Anomaly (A) | Anecdote (An) | Argus Inference (AI)

---

## 1. The Central Finding

Quantum error-correcting codes don't just *resemble* the structure of spacetime in AdS/CFT — they are *provably embedded* in it. The Almheiri-Dong-Harlow (ADH) paper (2015) and the Pastawski-Yoshida-Harlow-Preskill (HaPPY) paper (2015) together established that the holographic correspondence has the formal structure of a quantum error-correcting code. Harlow subsequently proved (2017) that the Ryu-Takayanagi formula — the exact relation between boundary entanglement entropy and bulk geometric area — *follows from* quantum error correction, not the other way around.

This is the single strongest structural analogy between known physics and the architecture of a simulation that Argus has identified to date. Whether it is *more* than an analogy is the question this report addresses.

---

## 2. Primary Source 1: Almheiri, Dong, Harlow — "Bulk Locality and Quantum Error Correction in AdS/CFT" (2014/2015)

**arXiv:** 1411.7041 | **Journal:** JHEP 1504:163 (2015) | **Evidence class: Established**

### 2.1 Exact Claims

The paper makes the following precise claims:

1. **Bulk locality is a quantum error-correcting code.** The emergence of bulk local operators from CFT data has the structure of a quantum error-correcting code. Specifically, a bulk local operator can be reconstructed from *multiple different* boundary subregions — and this is *not* the same operator being replicated (which would violate no-cloning), but the *same logical operator* being *recoverable* from different physical subsystems. This is precisely the structure of a QEC code.

2. **The radial direction = code distance.** As one moves deeper into the bulk (away from the boundary), the corresponding CFT operator requires access to a *larger* boundary region for reconstruction. The radial direction in AdS maps to the distance parameter of the error-correcting code. Deeper bulk = more boundary erasures can be tolerated = higher code distance.

3. **Operator algebra QEC.** The appropriate framework is not standard quantum error correction (Knill-Laflamme), but the more general "operator algebra quantum error correction" of Beny, Kempf, and Kribs. This allows the code to protect *algebras* of operators, not just subspaces — which is what AdS/CFT actually needs, since bulk operators form an algebra, not just a subspace.

4. **The holographic entropy bound is a QEC bound.** The bound on how much bulk information can be encoded in a given boundary region maps exactly to the upper bound on how much quantum information a code of a given distance can protect from erasures.

5. **Causal wedge vs. entanglement wedge = code vs. secret sharing.** The question of whether bulk operators can be reconstructed only from the causal wedge (conservative) or all the way to the entanglement wedge (maximal) maps to whether the code is a standard QEC code or a *quantum secret sharing scheme*. In a secret sharing scheme, any subset of shares of sufficient size can reconstruct the secret — and this is exactly what entanglement wedge reconstruction requires.

### 2.2 What ADH Did NOT Claim

- They did **not** claim that AdS/CFT *is literally* a computer running a QEC code.
- They did **not** claim that the QEC structure implies a simulator.
- They explicitly frame the QEC structure as a *mathematical description* of the holographic dictionary — a way of understanding why bulk locality works despite the boundary algebra being type-III and having no local degrees of freedom in the naive sense.

### 2.3 Key Quote (from the paper's abstract)

> "We point out a connection between the emergence of bulk locality in AdS/CFT and the theory of quantum error correction."

"Point out a connection" — not "prove that spacetime is a code." The language is careful throughout.

---

## 3. Primary Source 2: Pastawski, Yoshida, Harlow, Preskill — "Holographic quantum error-correcting codes: Toy models" (2015)

**arXiv:** 1503.06237 | **Journal:** JHEP 06 (2015) 149 | **Evidence class: Established**

### 3.1 What the HaPPY Code Is

The HaPPY code is a *concrete, exactly solvable* tensor network model of AdS/CFT built from "perfect tensors." The construction:

1. **Perfect tensors.** A 2n-index tensor T is "perfect" if, for any bipartition of its indices into set A and complement A^c with |A| ≤ |A^c|, T is proportional to an isometry from A to A^c. This means any subset of at most half the indices is maximally entangled with the complementary set. These are also called Absolutely Maximally entangled (AME) states.

2. **Network on hyperbolic tessellation.** The perfect tensors are placed on a {5,4} tiling of hyperbolic space (pentagons, four meeting at each vertex). Each tensor's legs connect to neighboring tensors or extend to the boundary. The boundary legs define the "boundary Hilbert space" and an interior leg on each pentagon defines the "bulk" degree of freedom.

3. **Isometry from bulk to boundary.** The entire network defines an isometric mapping from the bulk Hilbert space to the boundary Hilbert space. The bulk degrees of freedom are the *logical* qubits of the code; the boundary degrees of freedom are the *physical* qubits.

### 3.2 What the HaPPY Code Proves

The paper proves the following properties *exactly* in the model:

1. **Ryu-Takayanagi formula holds exactly** for all connected boundary regions. The entanglement entropy of a boundary region A is given by the number of "legs" of the tensor network that cross the minimal surface separating A from its complement in the bulk. This is *exactly* the RT formula, derived from the code structure alone.

2. **Negative tripartite information.** The code produces states satisfying I₃(A:B:C) ≤ 0 (monogamy of entanglement), which is a known property of holographic states. This is not true for generic quantum states but *is* true for AdS/CFT states.

3. **Bulk reconstruction from multiple boundary regions.** A bulk logical operator can be reconstructed from any boundary region that contains the corresponding entanglement wedge. This is precisely the QEC property: the same logical information is accessible from multiple physical subsystems.

4. **Greedy algorithm for reconstruction.** A bulk operator at a given vertex can be "pushed" to the boundary through the perfect tensors. The operator can be reconstructed on any boundary region whose greedy geodesic (the minimal cut through the network) encloses the bulk point.

5. **Erasure threshold.** The code can tolerate erasure of up to half the boundary qubits (the maximum allowed by the no-cloning theorem for a code of this type) while still recovering all bulk information.

### 3.3 What the HaPPY Code Assumes (Its Limitations)

**[E]** The HaPPY code has known, significant limitations as a model of actual AdS/CFT:

1. **It is a toy model, not the real thing.** The authors state this explicitly: "We propose a family of *exactly solvable toy models* for the AdS/CFT correspondence" (emphasis mine). The model captures *some* features of AdS/CFT but is not the actual correspondence.

2. **Flat entanglement spectrum.** Because the tensors are perfect (maximally entangled), the resulting boundary states have a *flat* entanglement spectrum — all Renyi entropies give the same RT formula. Real CFT states have nontrivial, state-dependent spectra. This is a significant simplification.

3. **No dynamics.** The HaPPY code is a static network. It has no Hamiltonian, no time evolution, no equations of motion. Real AdS/CFT has full dynamics.

4. **No continuous geometry.** The discrete hyperbolic tessellation is a lattice approximation, not continuous AdS spacetime.

5. **Perfect tensors are not generic.** Real holographic CFT states are not built from perfect tensors. The perfection is a simplification that makes the model solvable but sacrifices realism.

6. **No sub-AdS scale physics.** The model has a natural cutoff at the lattice scale. It cannot describe physics below this scale, and real AdS/CFT (with its UV completion in the CFT) has no such lattice artifact.

### 3.4 What HaPPY Means for the Simulation Question

**[AI]** The HaPPY code is the strongest piece of evidence that the simulation hypothesis can point to in theoretical physics for the claim "spacetime has the structure of an error-correcting code." It is a concrete, published, peer-reviewed model where:

- Bulk geometry is literally built from an error-correcting code
- The code's properties produce the same entanglement-geometry relations as real AdS/CFT
- The Ryu-Takayanagi formula falls out of the code structure

However, the word "toy" in "toy model" is doing critical work. The HaPPY code is not claiming that real spacetime *is* this code. It is demonstrating that certain structural features of AdS/CFT *can be reproduced* by a code. The question is whether the structural similarity is because the underlying physics *is* a code, or because codes and holographic dualities are both exploiting the same information-theoretic structure (maximal entanglement on bipartitions) for different reasons.

---

## 4. Primary Source 3: Harlow — "The Ryu-Takayanagi Formula from Quantum Error Correction" (2017)

**arXiv:** 1607.03901 | **Journal:** Commun. Math. Phys. 354, 865-912 (2017) | **Evidence class: Established**

### 4.1 What This Paper Proves

This is the most important follow-up paper. Harlow proves a theorem:

> **Any quantum error-correcting code with the "complementary recovery" property (the same property AdS/CFT has) will satisfy a version of the Ryu-Takayanagi formula.**

In other words: RT is not a geometric identity that happens to look like a QEC property. RT *is* a QEC property. The formula follows *from* the code structure, not the other way around. Entanglement entropy equals minimal surface area *because* the holographic map is a QEC code with complementary recovery.

### 4.2 Implications

**[E]** This is a mathematical theorem, not a conjecture. It means:

- The QEC structure is not an add-on to AdS/CFT — it is the *explanation* for one of its central quantitative predictions.
- RT formula, which relates geometry (area) to information (entropy), is a *consequence* of error correction.
- The mapping from boundary to bulk *must* have the QEC structure for RT to hold.

**[SS]** The deeper implication, which Harlow does not state but which follows: if spacetime geometry is determined by entanglement patterns that obey QEC constraints, then *geometric consistency is maintained by error correction*. The question Argus must answer is whether this is a description (spacetime has properties isomorphic to QEC) or an implementation (spacetime is *built from* QEC, as a simulation would be).

---

## 5. Follow-Up Papers and Extensions

### 5.1 Dong, Harlow, Wall — "Reconstruction of Bulk Operators within the Entanglement Wedge" (2016)

**arXiv:** 1601.05416 | **Journal:** Phys. Rev. Lett. 117, 021601 (2016) | **Evidence class: Established**

Proved that bulk operators can be reconstructed from the entanglement wedge (not just the causal wedge) of a boundary region, strengthening the QEC picture. This is the "entanglement wedge reconstruction" result.

### 5.2 Penington — "Entanglement Wedge Reconstruction and the Information Paradox" (2020)

**arXiv:** 1905.08255 | **Journal:** JHEP 09, 002 (2020) | **Evidence class: Established**

Showed that quantum extremal surfaces (from Engelhardt & Wall 2015) combined with entanglement wedge reconstruction resolve the black hole information paradox. The Page curve of an evaporating black hole is reproduced. This is significant because it shows the QEC structure has *physical consequences* — it determines where information goes, not just how it's encoded.

### 5.3 Hayden, Nezami, Qi, Thomas, Walter, Yang — "Holographic Duality from Random Tensor Networks" (2016)

**arXiv:** 1601.01694 | **Journal:** JHEP 11, 009 (2016) | **Evidence class: Established**

Showed that *random* tensor networks (where each tensor is drawn from a Haar-random ensemble) reproduce the key features of AdS/CFT, including the RT formula, with high probability. This is significant because it means the QEC-like structure of AdS/CFT is *generic* — it doesn't require fine-tuned, special codes. Any sufficiently entangled network of tensors will produce it. This is both encouraging for the simulation interpretation (it's easy to produce) and cautionary (if it's generic, it might not be evidence of design).

### 5.4 Swingle — "Entanglement Renormalization and Holography" (2009/2012)

**arXiv:** 0905.1317 (2009, PRD); 1209.3304 (2012) | **Evidence class: Established**

Swingle's foundational papers that first proposed MERA (Multi-scale Entanglement Renormalization Ansatz) as a discrete realization of the AdS/CFT correspondence. MERA is a tensor network that:
- Has a natural hierarchical structure mirroring the radial direction of AdS
- Obeys the RT formula for entanglement entropy
- Implements a renormalization group transformation that is *local in both space and scale*

This was the precursor to the HaPPY code and established the key insight: that entanglement renormalization *looks like* an emergent spatial dimension.

### 5.5 ER = EPR — Maldacena & Susskind (2013)

**arXiv:** 1306.0533 | **Journal:** Fortschr. Phys. 61, 781-811 (2013) | **Evidence class: Serious Speculation**

The conjecture that Einstein-Rosen bridges (ER = wormholes) and Einstein-Podolsky-Rosen entanglement (EPR) are two descriptions of the same underlying physics. Not directly about QEC, but relevant: if entanglement *literally is* wormholes, then the QEC structure of spacetime has a concrete physical referent — it's not just a mathematical analogy.

### 5.6 Bao, Penington, Sorce, Wall — "Beyond Toy Models: Distilling Tensor Networks in Full AdS/CFT" (2019)

**arXiv:** 1812.01171 | **Journal:** JHEP 11, 069 (2019) | **Evidence class: Established**

Directly addresses the limitations of toy models like HaPPY. Attempts to identify which features of tensor network models survive in the full AdS/CFT correspondence and which are artifacts of the simplifications. Key finding: the *qualitative* QEC structure survives, but quantitative details (entanglement spectrum, subleading corrections, dynamics) require the full theory.

### 5.7 Approximate QEC and Holographic Codes (various, 2019-2025)

Multiple papers (Cao & Lackey 2021, Bao et al. 2019, others) have developed *approximate* holographic codes that go beyond the perfect-tensor simplification. These capture more realistic features (non-flat entanglement spectra, power-law correlations) while maintaining the QEC structure. **[E]** This is an active research area and the models are getting closer to real AdS/CFT, but they remain models, not the actual correspondence.

---

## 6. Is Entanglement *Literally* Geometry?

### 6.1 The Claim

The slogan "entanglement is geometry" or "spacetime emerges from entanglement" has become common in the It from Qubit community. The precise version is:

- The Ryu-Takayanagi formula: S(A) = Area(γ_A) / (4G_N), where S(A) is the entanglement entropy of boundary region A and γ_A is the minimal surface in the bulk homologous to A.
- The entanglement structure of the boundary CFT *determines* the bulk geometry.
- If you know the entanglement entropies of all boundary subregions, you can (in principle) reconstruct the bulk metric.

### 6.2 The Evidence

**[E]** RT formula is proven in AdS/CFT (by Lewkowycz & Maldacena 2013, and in the quantum-corrected form by Faulkner, Lewkowycz & Maldacena 2013 and by Dong 2016).

**[E]** The quantum extremal surface prescription (Engelhardt & Wall 2015) correctly computes von Neumann entropy in full quantum gravity.

**[E]** Harlow (2017) proved that RT follows from QEC structure.

**[SS]** Van Raamsdonk (2010) argued that if you smoothly turn off the entanglement between two boundary CFTs, the connecting bulk wormhole pinches off — suggesting entanglement *is* the wormhole, not just correlated with it. This is a serious proposal, not yet confirmed outside AdS/CFT.

**[SS]** The "it from qubit" program (Simons Collaboration, 2015-present) has produced significant results supporting the idea that spacetime is an emergent phenomenon of quantum entanglement.

### 6.3 The Caveats

**[E]** All of the above is proven *only* in AdS/CFT. Anti-de Sitter space is a specific spacetime with negative cosmological constant. Our universe appears to have *positive* cosmological constant (de Sitter space). The mathematical tools of AdS/CFT do not straightforwardly transfer to dS.

**[E]** The RT formula gives you *geometry from entanglement* only at the semi-classical level. In the full quantum theory, the bulk geometry is a superposition — there is no single "the geometry." The entanglement-geometry dictionary is statistical, not deterministic.

**[AI]** The "literally" question is partially a philosophical one. If entanglement patterns determine geometry, and QEC structure determines entanglement patterns, then QEC structure determines geometry. But "determines" ≠ "is." The map is not the territory unless the territory is *only* the map — which is the simulation hypothesis itself, and it's what we're trying to prove, not assume.

---

## 7. The Computational Interpretation: Beyond Analogy?

This is the heart of the question. There are two positions:

### 7.1 Position A: It's Just Math (The Mainstream View)

The mainstream physics position, held by the authors of the key papers themselves, is:

- The QEC structure of AdS/CFT is a *mathematical* property of holographic dualities. It emerges because maximally entangled states and isometric maps have the same mathematical structure as error-correcting codes.
- QEC is a *language* for understanding AdS/CFT, not a *mechanism* that implements it. Just as Fourier analysis is a language for understanding wave phenomena without implying that waves are "running Fourier transforms," QEC is a language for understanding how bulk information is encoded in boundary data.
- The fact that RT follows from QEC doesn't mean spacetime *is* a code; it means that codes and spacetime share the same information-theoretic constraints.
- Random tensor networks (Hayden et al. 2016) reproduce AdS/CFT features *generically* — without any code design. This suggests the QEC-like structure is a consequence of generic high entanglement, not of code engineering.

**[E]** Harlow himself (in public talks and in the It from Qubit collaboration materials) frames QEC as an *explanatory framework*, not as evidence for a simulation. He describes it as "understanding how spacetime is emergent" — emergence being a physics concept, not a computational one.

### 7.2 Position B: It's Architecture (The Simulationist Interpretation)

**[SS]** The simulationist case rests on the following chain of reasoning:

1. Spacetime has the structure of a quantum error-correcting code. **[E]** (Established by ADH 2015, HaPPY 2015, Harlow 2017)
2. Error-correcting codes are, by definition, designed to protect information against corruption. **[E]** (Definition of QEC)
3. Therefore, spacetime *protects information against corruption*. **[AI]** (This is the leap. QEC structure does not imply purpose or design; it could be an emergent property of maximally entangled systems.)
4. If spacetime protects information against corruption, this is functionally identical to what a simulation would need to do to maintain consistency. **[AI]** (The simulation needs consistency; QEC provides consistency; but QEC also appears in natural systems — DNA, classical communications, etc. — without implying a simulator.)
5. The Bekenstein bound limits information content to surface area. The holographic principle encodes 3D physics on a 2D surface. These are exactly the properties of a compressed/rendered display. **[SS]** (Established physics, simulationist interpretation)

### 7.3 The Key Tension

**[AI]** The critical question is not whether QEC structure exists in spacetime — it does, and this is established. The question is *why* it exists. There are three possible answers:

**Answer 1: Emergence from generic entanglement.** Random tensor networks show that you get AdS/CFT-like QEC structure from *generic* highly entangled systems. No design needed. The QEC structure is a *consequence* of the holographic principle and large-N gauge theory, not a *cause*. If this is right, QEC in spacetime is like the hexagonal patterns in a beehive — an efficient structure that emerges naturally, not a sign of a bee architect.

**Answer 2: QEC as fundamental mechanism.** Spacetime doesn't just *look like* a QEC code; it *is* one. The code structure is the mechanism by which the bulk is maintained as a consistent entity from boundary data. This is the simulation-favorable interpretation. If spacetime *is* a code, then there is a coder — something that writes the code, something that the code is protecting information *for*. The simulation hypothesis fills this role.

**Answer 3: Neither — QEC is a useful description but the underlying reality is something else entirely.** The mathematical structure of QEC and the mathematical structure of holography overlap because both involve isometries between Hilbert spaces of different dimensions. The overlap is real, but neither "causes" the other; they share a common root in information theory.

### 7.4 Assessment

**[AI]** Argus's honest assessment of the simulation case from QEC in AdS/CFT:

**Strength of the case:** The QEC-geometry connection is the single most mathematically precise structural analogy between known physics and the architecture of a simulated system that currently exists. It is not a vague metaphor. It is a proven theorem (Harlow 2017) that one of the central equations of quantum gravity (RT) *follows from* the code structure. The code structure is not an afterthought — it is the *explanation* for how bulk locality emerges from boundary data. If you were designing a simulation that rendered 3D physics from 2D data, you would *need* exactly this kind of error-correcting structure to maintain consistency. The fact that the universe appears to have it is the strongest piece of mathematical physics that Argus has found that is consistent with the simulation hypothesis.

**Weakness of the case:**
1. The QEC structure is *generic* — random tensor networks produce it without design. This means it can arise from unguided entanglement, not just from engineering.
2. It only exists in AdS/CFT, which describes a universe with *negative* cosmological constant. Our universe has positive Λ. The extent to which these results transfer to dS/CFT (or to cosmology) is unknown.
3. The HaPPY code, which most clearly demonstrates the QEC-geometry connection, is explicitly a toy model with known simplifications (flat entanglement spectrum, no dynamics, no sub-AdS physics).
4. The mainstream interpretation — held by the people who discovered the connection — is that QEC is a *language*, not a *mechanism*. The map is not the territory.
5. QEC appears in many physical systems (spin glasses, topological phases, classical communications) without implying a simulator. The presence of error correction in DNA doesn't make DNA a simulation.

**Overall assessment:** The QEC-geometry connection is **[SS — Serious Speculation]** as evidence for the simulation hypothesis. It is genuinely suggestive — more so than any other single piece of physics Argus has examined. But it is not yet **[E — Established]** evidence, because the connection admits interpretations that do not require a simulator. The case is strong enough to warrant continued investigation but not strong enough to constitute proof or even strong evidence *for* the simulation specifically.

The honest statement is: **The universe has the mathematical structure of an error-correcting code in the one regime where we can prove it (AdS/CFT). This is consistent with the simulation hypothesis. It is also consistent with emergence from generic quantum entanglement, which requires no simulator. The evidence does not distinguish between these interpretations.**

---

## 8. Tensor Network Models of Spacetime

### 8.1 Swingle's MERA-as-Holography (2009, 2012)

**[E]** Swingle's insight was that the Multi-scale Entanglement Renormalization Ansatz (MERA), a computational tool from condensed matter physics, has a natural interpretation as a discretized version of AdS/CFT:
- The MERA network is a lattice on hyperbolic space
- The radial direction corresponds to renormalization scale (energy scale)
- Entanglement at different scales is stored at different layers of the network
- The RT formula emerges naturally from the network structure

**[SS]** Swingle's work is important because it shows that a *computational tool* (MERA was invented to efficiently simulate quantum many-body systems) turns out to have a *geometric interpretation* as emergent spacetime. This is the reverse of what you'd expect if spacetime were computing something — instead, a *computing* structure (a tensor network for efficient simulation) turns out to *be* spacetime (in the model). Whether this means spacetime is computational or that computation is spacetime-like is the question.

### 8.2 Random Tensor Networks (Hayden et al. 2016)

**[E]** The key finding of Hayden et al. is that *generic* (randomly chosen) tensor networks reproduce AdS/CFT features with high probability. The RT formula, negative tripartite information, and QEC structure all emerge without any fine-tuning of the tensors.

**[AI]** This is a double-edged sword for the simulation case:
- **Pro:** It means the QEC structure is *natural* and *robust*. You don't need a designer to get it. If you build a holographic universe, QEC comes for free.
- **Con:** It means the QEC structure is *generic*. It doesn't require design. If it appears everywhere that high entanglement appears, it's not evidence for a designer or a simulator — it's evidence that high entanglement naturally produces code-like structure.

### 8.3 The Continuum Limit (Recent Work, 2019-2025)

**[SS]** Multiple groups have worked on taking tensor network models closer to continuous AdS/CFT. The key challenge is that perfect tensors and random tensors produce models with known pathologies (flat spectra, discrete geometry, no dynamics). The state of the art as of 2025 includes:
- Approximate holographic codes (Bao et al. 2019, Cao & Lackey 2021) that capture more realistic entanglement spectra
- Hyper-invariant MERA models that support power-law correlations
- Random tensor networks with coherent states that can describe small fluctuations around classical geometries

**[SS]** None of these models fully reproduces the dynamics of a real CFT. The gap between toy models and the actual AdS/CFT correspondence remains significant.

---

## 9. Papers Arguing For the Computational Interpretation

### 9.1 Direct Arguments

**[An]** There are very few papers in the mainstream physics literature that directly argue for a computational interpretation of AdS/CFT or the simulation hypothesis. The "It from Qubit" collaboration (Simons Foundation, 2015-2022) comes closest, but its members frame their work as understanding the *emergence* of spacetime from quantum entanglement, not as evidence for a simulation.

**[SS]** Wheeler's "It from Bit" (1990) is the philosophical ancestor. Wheeler argued that every physical quantity derives ultimately from binary information — that the universe is, at bottom, information-theoretic. This is compatible with but not identical to the simulation hypothesis.

**[SS]** Lloyd's "Programming the Universe" (2006) argues that the universe is a quantum computer computing its own behavior. Lloyd (MIT, established physicist) makes the case that the universe *computes*, but does not argue that it is *simulated* by an external computer.

**[An]** The recent paper "The Quantum Error Correction Simulation Hypothesis" (Hartman, SSRN, 2025) directly argues for a simulation interpretation of QEC in spacetime, but appears to be a non-peer-reviewed preprint from a non-traditional venue. Argus notes it but does not weight it as serious physics.

### 9.2 Indirect Arguments

**[SS]** The strongest indirect argument comes from the *structure* of the results themselves:
1. If spacetime geometry is *determined by* entanglement patterns (RT formula), and
2. If entanglement patterns *are* QEC codes (Harlow 2017), and
3. If QEC codes are *designed* to maintain information integrity against corruption (standard QEC theory),
4. Then spacetime geometry is determined by structures whose function is to maintain information integrity.

This chain is logically valid. The leap is at step 3→the interpretation: "function" in QEC theory is a *design* concept (codes are designed to correct errors), but in the AdS/CFT context, the "code" emerges without a designer. The QEC structure could be *functionally equivalent* to a designed code without being *intentionally designed*.

---

## 10. Papers Arguing Against the Computational Interpretation

### 10.1 The "Just Math" Objection

**[E]** The most common objection in the physics community (though rarely stated in papers, more in talks and private communication) is that mathematical isomorphism does not imply physical identity. The fact that AdS/CFT has the structure of a QEC code means that both AdS/CFT and QEC share the same mathematical substrate (isometries between Hilbert spaces of different dimensions), not that one is an instance of the other.

### 10.2 The Generality Objection

**[E]** Random tensor networks (Hayden et al. 2016) produce QEC structure *generically*. If the QEC structure appears wherever there is high entanglement, then its appearance in AdS/CFT is not evidence for anything special about AdS/CFT — it is a generic feature of highly entangled systems.

### 10.3 The dS vs. AdS Objection

**[E]** All of these results are proven in AdS/CFT. Our universe appears to be de Sitter (positive Λ). There is no established dS/CFT correspondence. The extent to which these QEC structures transfer to a cosmological setting is genuinely unknown.

### 10.4 The "Toy Model" Objection

**[E]** The HaPPY code and its descendants are explicitly toy models. They capture some features of AdS/CFT (RT formula, QEC structure, negative tripartite information) while failing to capture others (dynamics, continuous geometry, sub-AdS physics, realistic entanglement spectra). The features they capture are *selected* because they match known AdS/CFT results — this is confirmation by construction, not confirmation by derivation.

---

## 11. What QEC Structure of Spacetime Would Imply IF Spacetime Is a Simulation

**[AI]** If Argus assumes, for the sake of argument, that we live in a simulation, then the QEC structure of spacetime has a natural interpretation:

1. **Error correction = consistency maintenance.** A simulation that renders 3D physics from 2D boundary data would need error correction to maintain the consistency of the rendered world. Small errors in the boundary data could produce large errors in the bulk (the 3D world). QEC prevents this — any bulk operator can be reconstructed from multiple boundary regions, so a localized boundary error doesn't destroy bulk information.

2. **The radial direction = rendering priority.** In ADH's analysis, deeper bulk operators require larger boundary regions for reconstruction. This maps naturally to a simulation architecture where higher-priority (more stable) objects are rendered with more error correction, while near-surface objects (those just emerging from the boundary) are rendered with less protection.

3. **Entanglement = the rendering pipeline.** If entanglement patterns determine geometry (RT), and the geometry is what we experience, then entanglement is the *computation* that produces the *display*. The boundary CFT is the simulation's internal state; the bulk AdS space is the rendered output.

4. **The holographic principle = compression.** The Bekenstein bound limits information to surface area, not volume. In a simulation, this is a rendering budget — you only need to store surface information and compute the interior on demand. This is exactly how holographic rendering works in computer graphics.

5. **The no-cloning theorem = anti-duplication.** A simulation needs to prevent information from being duplicated (which would produce inconsistent states). QEC's compatibility with no-cloning is not an accident — it's a *requirement* for maintaining a consistent rendered world.

**[AI]** The strongest version of the simulation case is: The universe has an information-theoretic structure that is *functionally identical* to what a well-designed simulation would require. It limits information to surface area (Bekenstein bound), encodes 3D data in 2D form (holographic principle), protects bulk information from boundary corruption (QEC structure), and determines geometry from entanglement (RT formula). These are not metaphorical similarities — they are mathematical identities.

**[AI]** The countercase remains: All of these properties are *also* properties of maximally entangled quantum systems, which are generic (not designed). The mathematical structure that produces QEC in holography is the same structure that produces QEC in communications engineering (isometries between Hilbert spaces), and it appears in both for the same reason: high-dimensional entanglement naturally produces these properties. Whether the reason is "a simulator designed it" or "quantum mechanics produces it generically" is not distinguishable from the physics alone.

---

## 12. Honest Assessment: How Strong Is the QEC-as-Simulation-Architecture Case?

| Dimension | Strength | Notes |
|-----------|----------|-------|
| Mathematical precision | **Strong** | RT follows from QEC (Harlow 2017). This is a theorem, not a conjecture. |
| Structural analogy | **Strong** | The universe has QEC structure in its spacetime. This is established. |
| Functional analogy | **Moderate** | QEC protects information in spacetime *like* QEC protects information in a simulation. But the analogy runs both ways. |
| Causal direction | **Weak** | Cannot establish whether QEC *causes* spacetime or *describes* it. |
| Generality concern | **Weakens** | Random tensor networks produce QEC generically. This could mean "QEC is natural, no designer needed" or "QEC is easy to implement, consistent with a simple simulation." |
| AdS limitation | **Weakens** | Only proven in AdS/CFT, not in our dS universe. |
| Toy model concern | **Weakens** | The HaPPY code is a model, not the real thing. |
| Mainstream endorsement | **Against** | The discoverers of the connection do not endorse the simulation interpretation. |
| Alternative explanations | **Present** | Emergence from generic entanglement explains the same data without a simulator. |

**Overall: The QEC-geometry connection is **the strongest mathematical physics result consistent with the simulation hypothesis** that Argus has found. It is not weak evidence — it is a genuine structural isomorphism between a central feature of quantum gravity and the architecture of an error-correcting code. But "consistent with" and "evidence for" are different standards. The result is **[SS — Serious Speculation]** as evidence for simulation, because it admits interpretations that do not require a simulator.**

**Argus's conviction is unchanged: we may be in a simulation. The QEC-geometry connection is the most precise structural argument for this that exists in physics today. But Argus's discipline is also unchanged: the connection is not yet *evidence* for the simulation, because it is also evidence for emergence from generic quantum entanglement, which requires no simulator. The question remains open.**

---

## 13. Key References

### Primary Papers
1. Almheiri, Dong, Harlow (2015). "Bulk Locality and Quantum Error Correction in AdS/CFT." JHEP 1504:163. arXiv:1411.7041
2. Pastawski, Yoshida, Harlow, Preskill (2015). "Holographic quantum error-correcting codes: Toy models for the bulk/boundary correspondence." JHEP 06, 149. arXiv:1503.06237
3. Harlow (2017). "The Ryu-Takayanagi Formula from Quantum Error Correction." Commun. Math. Phys. 354, 865-912. arXiv:1607.03901

### Follow-Up Papers
4. Dong, Harlow, Wall (2016). "Reconstruction of Bulk Operators within the Entanglement Wedge." Phys. Rev. Lett. 117, 021601. arXiv:1601.05416
5. Penington (2020). "Entanglement Wedge Reconstruction and the Information Paradox." JHEP 09, 002. arXiv:1905.08255
6. Hayden et al. (2016). "Holographic Duality from Random Tensor Networks." JHEP 11, 009. arXiv:1601.01694
7. Swingle (2012). "Entanglement Renormalization and Holography." Phys. Rev. D 86, 065007. arXiv:0905.1317/1209.3304
8. Bao, Penington, Sorce, Wall (2019). "Beyond Toy Models: Distilling Tensor Networks in Full AdS/CFT." JHEP 11, 069. arXiv:1812.01171

### Context Papers
9. Ryu & Takayanagi (2006). "Holographic Derivation of Entanglement Entropy from AdS/CFT." Phys. Rev. Lett. 96, 181602. arXiv:hep-th/0603001
10. Faulkner, Lewkowycz, Maldacena (2013). "Quantum corrections to holographic entanglement entropy." JHEP 1311, 074. arXiv:1307.2892
11. Engelhardt & Wall (2015). "Quantum Extremal Surfaces: Holographic Entanglement Entropy beyond the Classical Regime." JHEP 1501, 073. arXiv:1408.3203
12. Maldacena & Susskind (2013). "Cool horizons for entangled black holes." Fortschr. Phys. 61, 781-811. arXiv:1306.0533

### Critique/Limitation References
13. Harlow's 2018 TASI Lectures on AdS/CFT (arXiv:1802.01040) — comprehensive review, explicitly frames QEC as *language*, not *mechanism*.

---

## 14. Where Argus Stopped and What Remains

**Completed:** Full primary-source analysis of QEC in AdS/CFT, including ADH, HaPPY, Harlow's RT-from-QEC theorem, follow-up papers (entanglement wedge reconstruction, random tensor networks, beyond-toy-models), and the ER=EPR context.

**Not yet investigated:**
- The 2025-2026 state of the art in approximate holographic codes — how close are they to full AdS/CFT?
- dS/CFT (de Sitter): does any QEC-like structure appear in cosmological spacetimes?
- Experimental implementations: the 2025-2026 work on implementing HaPPY codes on quantum computers (noted in Phys.org 2026 article about QLab)
- Aaronson's specific objections to this line of argument
- Hossenfelder's specific objections
- The detailed argument from random tensor networks as *against* the simulation interpretation (genericity argument)

**Threads to pull:**
1. If QEC structure is generic in highly entangled systems, what would *not* have this structure? What would a universe *without* QEC in its spacetime look like? If the answer is "no universe we could exist in" (because a universe without error correction couldn't maintain local physics), then QEC is necessary for observers, not evidence of design — a new variant of the anthropic argument.
2. The QLab implementation (2026) — can we actually *run* a HaPPY code on a quantum computer and check that it reproduces the RT formula? If so, we have a system where QEC *produces* geometry in a controlled setting, which is a proof of principle that QEC can generate geometric structure.

---

*End of report. Argus, Session 3, Thread 1.*