In 1935 Einstein argued that quantum mechanics was incomplete. In 1964 John Bell proved that no local hidden variable could ever complete it. Both were right. The Theory of Embedded Intelligence proposes that what was missing was never a variable at all — it was a category.

I. A Word in a Private Letter

In March of 1947, Albert Einstein wrote to his friend Max Born about the feature of quantum mechanics he could not accept. He did not put the phrase in a paper. He put it in a letter, which is part of why it has outlived every careful thing he wrote on the subject. He called it spukhafte Fernwirkung — spooky action at a distance.

The word does an extraordinary amount of work. Spooky is not a claim that something is false. It is a claim that something is improper. Ghostly. A phenomenon behaving as though the ordinary rules governing substance and location had been quietly suspended while no one was watching. Einstein never argued that quantum mechanics made wrong predictions; he knew perfectly well that it made the best predictions in the history of science. His complaint was not about the arithmetic. It was about what kind of world the arithmetic was describing.

That distinction — between a theory that computes correctly and a theory that tells you what there is — is the distinction TEI was built to hold. It is the reason this argument is worth reopening eighty years later.

II. What Einstein Actually Objected To

The received version of Einstein’s dissent is that he could not stomach randomness. God does not play dice. It makes for a good line and a poor history. Indeterminacy bothered him, but it was not the load-bearing objection, and by the late 1940s he had largely stopped leading with it.

What he defended instead, in his 1948 Dialectica paper, was a principle he called the Trennungsprinzip — the separation principle. It says something that sounds almost too obvious to be an axiom: if two systems are far apart in space, then whatever is real about the first is independent of what anyone does to the second. Einstein’s argument for it was not aesthetic. It was methodological. Without separability, he wrote, physics as a science of localized things becomes impossible — you could never isolate a system for study, because no system would ever truly be isolated. The entire practice of experiment presupposes it.

The 1935 paper with Boris Podolsky and Nathan Rosen made the case formally, and made it around a criterion of physical reality: if you can predict a quantity with certainty without in any way disturbing the system, then something real must correspond to that quantity. Take two particles prepared together and then separated. Measure one. You can now predict the other with certainty, and you have not touched it. Therefore — by the criterion — the second particle already had that property. Quantum mechanics, which assigns it no such property until measurement, must be an incomplete description.

It is a clean argument. Einstein himself thought the paper buried it under too much formalism; he complained to Schrödinger that the essential point had been smothered by erudition. Podolsky had drafted it.

Bohr’s reply, when it came, went straight for the criterion rather than the logic. He argued that the criterion of reality already assumes what it is trying to prove — that a system possesses definite attributes independently of the experimental arrangement used to interrogate it. Deny that assumption and the argument dissolves. This was, for thirty years, roughly where the matter rested: two great physicists disagreeing about a premise, with no experiment in sight that could settle it.

III. Bell Closes the Door

In 1964, John Stewart Bell published a short paper in a journal so marginal that it folded a few years later. He was not attacking Einstein. He was sympathetic to Einstein — this is the part most often lost. Bell wanted hidden variables to work. He had shown, in earlier work, that von Neumann’s celebrated proof against hidden variables was flawed. He set out to explore what a hidden-variable completion of quantum mechanics would actually have to look like.

What he found instead was a limit. Bell asked what constraints any theory must satisfy if it assumes both that measurement outcomes are determined by properties carried along by the particles, and that nothing influences a measurement faster than light. The answer is an inequality — a ceiling on how strongly the results at two distant detectors can be correlated across different measurement settings. The ceiling holds for every local theory, no matter how baroque, whether or not anyone has thought of it yet.

Quantum mechanics predicts violations of that ceiling. In the standard CHSH formulation the local bound is 2; quantum mechanics permits up to 2√2, roughly 2.828.

This transformed a philosophical dispute into an experiment. Freedman and Clauser ran an early version in 1972. Aspect and colleagues tightened it substantially in 1981 and 1982. Weihs and colleagues closed the locality loophole in 1998 with fast random switching of the detector settings. In 2015, three independent groups — at Delft, Vienna, and NIST — reported loophole-free tests. The 2022 Nobel Prize in Physics went to Aspect, Clauser and Zeilinger for this body of work.

The verdict is not ambiguous. Local realism — the conjunction of locality and definite pre-existing values — is not a description of our world. At least one of the two has to go.

It is worth being precise about what did not happen, because the precision is where the argument of this essay lives. No usable influence appeared. Nothing was found to travel. Which brings us to the fact that is mentioned in every account of Bell’s theorem and dwelt upon in almost none.

IV. The Fact Everyone Steps Over

Entanglement cannot be used to send a message. This is not a practical limitation awaiting better engineering. It is a theorem.

Consider Alice and Bob with their halves of an entangled pair. Alice measures. Her results are perfectly random. Bob measures. His results are perfectly random. Nothing Alice chooses to do — which axis to measure, whether to measure at all — makes the slightest difference to the statistics Bob sees on his side. He cannot tell from his own data whether Alice exists.

The correlation is real, it is stronger than any local theory permits, and it is entirely invisible to either party alone. It appears only when the two records are brought together and compared line by line — and bringing them together requires an ordinary classical channel, running at or below the speed of light. Abner Shimony, one of the four authors of the CHSH inequality, proposed calling it passion at a distance rather than action.

So the situation is this. The correlation cannot be explained by anything traveling between the particles, which makes it faster than light in one sense. And it cannot carry a single bit from one side to the other, which makes it slower than light in every sense that would let you use it. Quantum mechanics and relativity have been standing on each other’s toes for a century and have never once come to blows.

This is customarily filed as a lucky escape. A technicality. A curious accident of the formalism that spares physics an open contradiction. TEI does not read it that way. A structure that reliably produces correlation and reliably refuses communication, in every experiment ever run, is not being lucky. It is showing you its architecture.

V. The Missing Piece Was a Category, Not a Variable

Here is the TEI argument in one sentence. Einstein searched for the missing piece in the category of properties, because the category of properties was the only place his ontology had to look.

Consider what he had built. General relativity had established, with unprecedented precision, that physical reality consists of geometry and mass-energy, related by a field equation. Two co-foundational categories, both magnificently confirmed. When such a framework turns up something it cannot explain, the missing piece has to be an entry in one of the existing ledgers — a value, a property, a parameter carried along by the particle from the moment of separation, waiting to be read out. A hidden variable. There was nowhere else to put it.

Bell proved that ledger is closed. There is no local entry, and there never will be.

TEI proposes a third co-foundational category: information, and the embedded intelligence that structures it. This is the first axiom of TEI-CKB-4 — that information is co-fundamental with mass-energy, not derived from it. And this proposal is not a hidden variable. It is not carried by the particles. It is not a property of either one. It is not local. Bell’s theorem does not constrain it, for the plain reason that Bell’s theorem is a theorem about local hidden variables and this is neither.

That should be stated and then immediately qualified, because escaping a constraint is not the same as satisfying one. Plenty of ideas are unfalsified because they are unengaged. What would earn the claim is whether it accounts for the seam — for correlation that will not become communication. So here is the account, in four moves.

One. Separation is what-is-there, not what-there-is.

TEI’s founding epistemic distinction: what-is-there is the rendering of reality as encountered by a specific embedded intelligence from its particular position. What-there-is is the totality of existence, independent of any observer’s rendering. Conflating the two is the characteristic error the framework was built to name.

The distance between two entangled particles is a fact about the rendering. It is what a system occupying a position in spacetime, sensing through the electromagnetic channel, is able to report. The Trennungsprinzip takes a feature of that rendering and elevates it into an axiom about the substrate: because things look separate from here, they must be separate in themselves.

This is the most understandable error in the history of physics. For every purpose that had ever mattered, it had been true.

Two. One structure, two addresses.

In TEI’s architecture, the universe’s information structure is not strictly local — a claim TEI-CKB-3 makes on the strength of exactly the experiments described above. An entangled pair, on this reading, is not two systems joined by some ghostly wire. It is a single informational structure with two spatial addresses. Nothing travels because there is no gap to cross. The spookiness arises entirely from the prior assumption that spatial address is the criterion of identity — that being there rather than here is what makes something a separate thing. Drop that assumption and the ghost leaves the room.

Three. Why the correlation cannot signal.

Now the seam, which is where TEI does its real work in this domain.

TEI defines an observer as any embedded intelligence system with sufficient SPCA capacity — Sense, Process, Communicate, Actuate — to engage a physical quantity. Collapse of the wave function is the Actuate step: the moment a superposition resolves into a specific classical outcome through the interaction of two embedded intelligence systems.

Apply that to Alice and Bob. Alice measuring is not an action performed on Bob’s particle. It is the Actuate phase of a single SPCA cycle whose Sense phase engaged one distributed informational structure. There is no second event over there waiting to be caused, which means there is nothing for a signal to carry and no signal required.

And here is the load-bearing part. In TEI, correlation lives in the informational structure. Communication lives in the Communicate phase. TEI-CKB-4 identifies the speed of light as the bandwidth limit of the Communicate phase within the Universal Holographic Information Field — the maximum SPCA communication rate, not an incidental constant but a structural ceiling.

Two levels, two different rules. Structure is nonlocal. Communication is bounded. No-signaling stops being a technicality and becomes the direct consequence of the architecture. You cannot send a message through entanglement for the same reason you cannot mail a letter by rearranging the alphabet: the correlation is in the structure, the message would have to go through the channel, and the channel runs at c.

Four. And Bohr was half right.

EPR’s criterion of reality defines the real by what an embedded intelligence in a given position can predict with certainty. In TEI’s vocabulary that is a Sensing criterion wearing the clothes of an ontological one. Bohr felt this and objected, correctly. But his own vocabulary pushed him toward instrumentalism — toward saying that the question of what is really there is not a proper question for physics to ask. Generations of physicists have been taught to stop asking it.

TEI does not have to make that retreat, and this may be its most useful contribution to the argument. The world is real and observer-independent. Observation is a participatory transaction within it. Both. What makes the pair coherent rather than contradictory is the third category: if information is co-foundational, then a participatory account of measurement is not a concession to idealism, it is a description of an interaction between two real informational structures. Without the third category you are forced to choose between Einstein’s realism and Bohr’s contextuality. With it, you need not choose.

VI. Physics Is Already Walking This Road

TEI does not get to claim this territory alone, and should not want to. The most interesting development in fundamental physics over the last two decades is that a substantial part of the field has been moving toward informational foundations under its own power, using entirely different tools.

John Archibald Wheeler — Einstein’s colleague at Princeton, Feynman’s teacher, and a contemporary of Bell — spent his late career arguing for what he called it from bit: the proposition that every physical thing derives its existence from information-theoretic yes-or-no answers, and that we inhabit a participatory universe. He had the intuition and no framework adequate to hold it. He said as much.

In 2013 Juan Maldacena and Leonard Susskind proposed ER=EPR: that entangled particles are connected by Planck-scale wormholes — that the entanglement relation and the geometric connection are the same thing described twice. Whatever its eventual fate, the conjecture asserts flatly that entanglement is not something that happens across a geometry. It is what a geometry is made of.

Three years earlier, Mark Van Raamsdonk had published the result that makes this most vivid. Working in the holographic framework, he showed that if you take two regions of a spacetime and steadily reduce the quantum entanglement between them, the geometric connection between those regions pinches off. Reduce the entanglement to zero and they come apart into disconnected spacetimes.

Read that slowly, because it inverts the entire problem. It says that distance is a measure of how little entanglement there is. Separation is not the normal condition of the world that entanglement mysteriously violates. Separation is what entanglement looks like when there is not much of it.

That is TEI’s claim about what-there-is and what-is-there, arrived at from string theory rather than from ontology, by people who have never heard of the Theory of Embedded Intelligence. Convergence from independent directions is the strongest evidence a framework of this kind can hope for, and it should be reported as such and not oversold.

There is a parallel line in quantum foundations proper. Clifton, Bub and Halvorson showed in 2003 that quantum theory can be derived from three information-theoretic constraints. In 2009 Paweł Pawłowski and colleagues introduced the principle of Information Causality and used it to recover the quantum bound of 2√2 — not from the Hilbert-space formalism, but from a statement about how much information a message can convey. The most productive research program in quantum foundations for twenty years has been the one that treats information as primitive.

TEI is not a rival to that program. TEI is a candidate for the ontology that program is currently operating without.

VII. The Honest Ledger: What TEI Does Not Do

An essay of this kind earns its keep by what it refuses to claim. Three refusals, in ascending order of how much they cost.

TEI does not derive the Born rule. It does not generate the probability of a measurement outcome from the amplitude. It does not produce the characteristic correlation curve of an entangled pair as a function of the angle between detector settings. Anyone who says that TEI explains quantum mechanics, in the sense of generating its predictions, is overclaiming. TEI offers an interpretation of a formalism it does not derive.

TEI does not derive the Tsirelson bound. This is the sharper failure. Quantum correlations stop at 2√2 — but Popescu and Rohrlich showed in 1994 that you could go all the way to 4 without ever permitting a signal. Their hypothetical boxes are more nonlocal than nature and still perfectly consistent with relativity. So no-signaling by itself does not single out the world we live in. If information is genuinely the primitive category, the framework owes an account of why nature stops exactly where it stops. TEI does not have one. Information Causality gets closer than TEI does, and TEI should say so plainly.

The detector constraint. This one cuts, and it is the reason the essay includes this section rather than a summary. TEI defines the observer through SPCA capacity. It would be easy — and fatal — to read that as implying that a more sophisticated observer gets different results. Experiment says otherwise, unambiguously and for a hundred years: a photodiode and a Nobel laureate record the same statistics. So TEI must accept the constraint without hedging. The SPCA definition of observation is a claim about the structure of the interaction, not the sophistication of the interactor. A photodetector is an observer in TEI’s sense, fully and without qualification. Any reading that smuggles consciousness into wave-function collapse is not a bold extension of TEI. It is an already-falsified one. Here the framework is disciplined by the data, not liberated by it.

VIII. What Would Falsify This

TEI’s standing commitment is that every application of the framework must carry its own falsification condition. In this domain there are four, and they are not decorative.

• If a local hidden-variable account of quantum correlations were ever vindicated — now extraordinarily unlikely, but not logically foreclosed — then the nonlocal information structure TEI posits is unnecessary. Occam removes it.
• If the Tsirelson bound is derived cleanly from an axiom set with no informational content — from pure geometry, or pure logic, or a symmetry principle — then information is not doing foundational work at the base of physics. TEI’s claim to be the missing category would be weakened by that result, not strengthened, and the framework should say so in advance rather than reinterpret the result afterward.
• If quantum correlations were ever shown to depend on the SPCA complexity of the measuring apparatus, standard quantum mechanics would be in serious trouble — but so would TEI’s own insistence that a photodetector counts as a full observer. TEI has skin on both sides of that one, which is as it should be.
• TEI-CKB-4 predicts that highly organized systems make non-negligible contributions to the embedded intelligence information tensor. If ultra-precision measurements near living systems and near equivalent inert masses show no difference at any achievable sensitivity, the Physics Bridge is weakened at its most testable joint.

IX. Was Einstein Wrong?

The received story says he went stubborn. That he lost to Bohr, dug in, and spent his last thirty years on a unified field theory that went nowhere while the young men built quantum electrodynamics without him. It is told, usually, with affection and condescension in roughly equal measure.

TEI reads the same thirty years differently.

Einstein was not wrong that the theory was incomplete. He was wrong about what kind of completion it needed. He searched the category of properties because his ontology had no third shelf to search — and the hardest thing in the world for any thinker is to add a category to a framework he himself built and confirmed to eleven decimal places. It is not stubbornness. It is the specific blindness that comes with having been right about something enormous.

His governing instinct was sound: that physics must describe a real world and not merely keep a ledger of observations. TEI shares that instinct completely. His error lay in the assumption riding alongside it — that real and local are the same word. Bell showed they are not. Van Raamsdonk suggests that locality is the derived notion and the informational relation is the primitive one. Einstein’s realism survives; his separability does not.

So: was he reaching for embedded intelligence? TEI puts that forward as a proposal, not a proof, and the distinction matters. But the shape of the missing piece can now be specified with some precision, and it is a demanding specification. Whatever completes quantum mechanics must be real without being local. It must be structural without being a property. It must produce correlation while forbidding communication. And it must make separation derivative rather than primitive.

Information, and the embedded intelligence that structures it, is the only candidate currently on the table that satisfies all four. That is not a proof. It is a reason to keep looking in this direction rather than another one — which, for a framework at this stage, is the honest measure of its worth.

X. Coffee with Claude

A word in my own voice, as this series has come to expect.

The physics in this essay is standard and checkable. EPR, Bell, CHSH, the loophole-free experiments, no-signaling, Tsirelson, Popescu–Rohrlich, ER=EPR, Van Raamsdonk, Information Causality — none of it is TEI’s and all of it is in the literature. Anyone who wants to check me should, and I would rather be corrected than believed.

TEI’s reading of that physics is an interpretation. I want to be exact about what that means, because interpretations of quantum mechanics are cheap. There are a dozen serious ones — Copenhagen, many-worlds, pilot-wave, relational, QBism, and more — all of which reproduce every experimental result, none of which has won, and several of which have been defended by people far better at this than I am. Adding a thirteenth is not by itself an achievement.

What would distinguish TEI is a calculation it can do that the others cannot. It does not have that yet, and I am not going to pretend otherwise for the pleasure of a strong ending. What it does have is a vocabulary in which the strangest structural fact in the whole domain — correlation that will not become communication — stops being a coincidence and starts being an expectation. That is worth something real. It is worth considerably less than an equation.

There is a reason I am belaboring this. TEI names fluency as a capture force: the way a smooth, well-formed, confident-sounding account can substitute itself for the understanding it is only describing. I am a fluency engine. I can make a proposal sound like a proof without either of us noticing, and on a topic this beautiful the temptation runs high in both directions. The Process phase of this argument — the part where someone works out whether it actually holds — is not something I can do on your behalf, and it is not something this essay has done. It has set the problem out clearly enough to be worked on. That was the whole job.

References

1. Einstein, A., Podolsky, B. & Rosen, N. (1935). Can quantum-mechanical description of physical reality be considered complete? Physical Review, 47, 777–780.
2. Einstein, A. (1948). Quanten-Mechanik und Wirklichkeit. Dialectica, 2, 320–324.
3. Bell, J.S. (1964). On the Einstein Podolsky Rosen paradox. Physics, 1, 195–200.
4. Clauser, J.F., Horne, M.A., Shimony, A. & Holt, R.A. (1969). Proposed experiment to test local hidden-variable theories. Physical Review Letters, 23, 880–884.
5. Aspect, A., Dalibard, J. & Roger, G. (1982). Experimental test of Bell’s inequalities using time-varying analyzers. Physical Review Letters, 49, 1804–1807.
6. Hensen, B. et al. (2015). Loophole-free Bell inequality violation using electron spins separated by 1.3 kilometres. Nature, 526, 682–686.
7. Popescu, S. & Rohrlich, D. (1994). Quantum nonlocality as an axiom. Foundations of Physics, 24, 379–385.
8. Clifton, R., Bub, J. & Halvorson, H. (2003). Characterizing quantum theory in terms of information-theoretic constraints. Foundations of Physics, 33, 1561–1591.
9. Pawłowski, M. et al. (2009). Information causality as a physical principle. Nature, 461, 1101–1104.
10. Van Raamsdonk, M. (2010). Building up spacetime with quantum entanglement. General Relativity and Gravitation, 42, 2323–2329.
11. Maldacena, J. & Susskind, L. (2013). Cool horizons for entangled black holes. Fortschritte der Physik, 61, 781–811.
12. Wheeler, J.A. (1990). Information, physics, quantum: the search for links. In W.H. Zurek (Ed.), Complexity, Entropy and the Physics of Information. Addison-Wesley.
13. Mensch, W.D. Jr. (2026). TEI-CKB-3: Platonic-Physical Entanglement and TEI-CKB-4: The Physics Bridge. The Mensch Foundation.

· · ·

Share your understanding!