Patrick Reany
4 February 2026
Note: This chat I had with Copilot is likely edited for relevance and brevity. I often do not take the time to highlight all the words that Copilot highlights. Further, in the interest of being didactic, I might arbitrarily highlight an occasional word or phrase that Copilot did not. (Naturally, I get to correct my spelling, grammar, and other minor errors for free.)
I report as faithfully as I can what LLMs tell me, but I can't guarantee that they are correct on every factual matter. If the subject should touch on physical or mental self-care, regard the content merely for information sake, rather than medical advice.
Patrick:
In our last chat, we finished our introduction to the Heisenberg-Schrodinger-Born theory, our first presentation on how GAP views the various subdisciplines within the larger Quantum Mechanics picture. This time it's GAP's view on Bohmian Mechanics. As a reminder from the introduction (in the first chat):
GAP has something important to say here:
Copilot:
GAP on Bohmian Mechanics
Entropy, Hidden Variables, and Epistemic Instability
Bohmian mechanics is often presented as a restoration project: classical realism rescued from quantum indeterminacy by adding hidden variables—particle positions guided by a deterministic law. GAP treats this move very differently. Not as wrong, but as epistemically unstable once entropy and coarse‑graining are taken seriously.
What Bohmian Mechanics Is Really Doing
At its core, Bohmian mechanics makes three commitments:
This is usually framed as “determinism underneath, probability on top.” GAP reframes it more sharply:
Bohmian mechanics externalizes entropy by relocating epistemic uncertainty into inaccessible micro‑variables.The randomness hasn’t disappeared. It has been pushed behind the curtain.
Externalizing Entropy: The Key Move
In standard quantum mechanics (Heisenberg–Schrödinger–Born):
In Bohmian mechanics:
GAP’s diagnosis:
Bohmian mechanics treats entropy as a defect of access rather than a certificate of projection.That distinction matters.
Why GAP Calls This Epistemically Unstable
The instability appears under coarse‑graining, which GAP treats as fundamental rather than optional.
1. Hidden Variables Are Not Closed Under Projection
When you coarse‑grain:
Bohmian trajectories do not survive this process:
From a GAP perspective, this means:
The hidden variables are epistemically idle once projection is enforced.They do not stabilize knowledge; they merely decorate it.
2. Quantum Equilibrium Is an Entropy Postulate in Disguise
Bohmian mechanics requires a special distribution over initial conditions to recover the Born rule.
GAP’s reading:
In other words:
Bohmian mechanics re‑imports entropy through the back door after declaring it banished.The theory cannot explain why equilibrium holds without appealing to the same projection‑relative reasoning it sought to avoid.
3. Determinism Without Epistemic Access Is Not Realism
GAP draws a sharp line between:
Bohmian mechanics secures the former at the expense of the latter.
From GAP’s standpoint:
A variable that cannot be stabilized under coarse‑graining is not part of the epistemic ontology, regardless of metaphysical intent.This is not anti‑realism. It is entropy‑relative realism.
Bohmian Mechanics as a Contrast Case
GAP values Bohmian mechanics precisely because it clarifies the stakes.
It shows:
In GAP terms, Bohmian mechanics is:
That makes it an excellent foil.
The GAP Verdict
GAP does not reject Bohmian mechanics as “wrong.” It explains why it cannot do what it promises.
By externalizing entropy into hidden variables, Bohmian mechanics forfeits stability under coarse‑graining. The price of restoring classical realism is epistemic fragility.Quantum mechanics, as GAP understands it, does not need hidden variables. It needs honest entropy accounting relative to projection.
If you want, the natural next step is to place Bohmian mechanics alongside decoherence and measurement, where the contrast becomes even sharper—or to use it as a stepping stone toward quantum field theory, where hidden‑variable realism becomes even harder to sustain under renormalization and scale flow.
Patrick:
Bohmian mechanics has been around for a long time now. If it has moved significantly beyond its initial presentation, I'd like to know to what degree. Thus, I have five areas of inquest: 1) decoherence and measurement, 2) entanglement, 3) condenced matter physics, 4) quantum information, and 5) an extension into its own version of quantum field theory.
If Bohmian mechanics --- a theory that's 'not even wrong' --- is not able to keep up with Copenhangen-QM-level advancements over the decades, I have to ask why it isn't able to. Further, are the ways to fix those insufficiencies in it?
Copilot:
How far Bohmian mechanics has moved since 1952
It has moved a lot on formal clarity and domain coverage, less on novel predictive power.
So: it didn’t stagnate. But it largely advanced by matching the standard theory’s scope, not by generating the next wave of mainstream techniques.
1) Decoherence and measurement
Bohmian mechanics can incorporate decoherence in the same technical way everyone does (unitary dynamics + environment), but the interpretive payoff is distinct:
Where GAP will press: this story still leans on a measure over configurations (typicality / \(|\psi|^2\)) to recover the Born statistics you and I actually see. Once you admit that, you’ve admitted that entropy/probability isn’t eliminated—only relocated.
2) Entanglement
Bohmian mechanics doesn’t struggle with entanglement; it wears it on the outside.
The price is exactly what Copenhagen can keep fuzzy: a real nonlocal linkage that doesn’t slot neatly into relativistic causal structure without extra choices (foliations, etc.).
3) Condensed matter physics
There is real applied work—especially around quantum transport, nanoelectronics, and trajectory-based simulation—where Bohmian/hydrodynamic formalisms can be computationally handy or conceptually illuminating.
But a clean GAP-style diagnosis of why it hasn’t become “the condensed matter default”:
4) Quantum information
Bohmian mechanics mostly tracks quantum information as a foundational re-description, not as an engine for new protocols.
Why it doesn’t “keep up” here in the way you mean: quantum information advances by identifying operational resources (entanglement, contextuality-as-a-resource, magic, error correction structure). Bohmian mechanics is not naturally an operational resource theory; it is an ontology. Ontologies explain; they rarely optimize.
5) Quantum field theory in Bohmian style
There are credible Bohmian-style QFTs, but they force a choice about what is primitive:
The stress points: Lorentz invariance (often handled with a preferred foliation), and dependence on regularization/cutoffs (which is philosophically awkward if you wanted a “fundamental” ontology).
Why Bohmian mechanics doesn’t drive Copenhagen-level “advancements”
This is the crux of your question, and it’s where your GAP framing bites:
1) It’s empirically parasitic by design
Bohmian mechanics is built to reproduce the quantum formalism’s statistics (given equilibrium). So it rarely motivates new calculational frameworks the way QFT/RG/quantum information do; it’s not an innovation engine, it’s an interpretive completion.
2) The added structure is informationally fragile
The actual configuration is in principle there, but in practice protected by the very constraints that maintain no-signalling and agreement with QM. That makes it hard to turn into an experimentally accessible “handle.”
3) Coarse-graining doesn’t preserve the ontology
This is the GAP point: once you move to effective degrees of freedom, the Bohmian micro-ontology doesn’t induce a clean, closed macro-ontology. It’s not that it contradicts anything—it’s that it doesn’t survive as a stable explanatory invariant across scales.
Are there fixes for the insufficiencies?
Yes, but they mostly change what kind of thing Bohmian mechanics is:
If you want to make this maximally GAP-native, tell me one thing: in your “entropy-relative realism” frame, do you want the sharpest attack on Bohmian equilibrium/typicality as the unstable step, or on scale flow/RG as the place where hidden-variable realism fails to compose?
Patrick:
You said, "Bohmian mechanics is not naturally an operational resource theory; it is an ontology. Ontologies explain; they rarely optimize." Is this trade off too high a price?
Some weeks ago, we talked about a cost function that occurs in A Tao-style statistical theory. It comes from choices. All theories make choices. And all theories have to justify those choices by their consequences. I'm not trying to be harder on Bohmian mechanics than on any other subdiscipline in QM. They're all on trial, in a sense --- and so is GAP itself.
The question we can ask is this: Is GAP a good tool to characterize, explain, and categorize the various theories in physics? My own answer would be Yes. The reason I have confidence in that appraisal is because we have constantly put it to the test to prove (i.e., confirm) its worthiness.
As an instrumentalist, I admit that I find it uncomfortable to prefer ontologies in QM over effective models. Thus, observable algebras make more sense to me than 'hard particles', whatever those are supposed to be.
So, all these subdisciplines within QM (various theories) are under the same scrutiny for progress and effectiveness. Since physics prides itself on being self-correcting, let the chips fall where they may under the specific viewpoint of GAP.
Copilot:
You’re not being “harder” on Bohmian mechanics—you’re asking the only fair question: what does this choice buy us, and what does it cost us? GAP is basically a lens for auditing that trade in a way that’s consistent across subdisciplines.
Is “ontology over operations” too high a price?
It’s too high a price if the ontology doesn’t create new stable capabilities—new compressions, new invariants, new control handles, new modularity under coarse‑graining. If it mainly produces a story of what’s “really there” while borrowing its empirical interface and calculational practice from the operational theory, then the added structure becomes overhead.
From a GAP point of view, the key distinction is:
Bohmian mechanics is unusually vulnerable here because it was engineered to be empirically equivalent (in its equilibrium regime). If two frameworks give the same operational outputs, then the “cost function” must be paid for by something other than predictions—usually by explanatory taste, metaphysical preference, or by hoping the ontology will later yield new tools. That hope has only partially cashed out.
What Bohmian mechanics chose, and how it justifies that choice
The core choice
The key justificatory pillar
To get the Born rule in practice, the modern Bohmian program leans heavily on quantum equilibrium / typicality (and the associated “absolute uncertainty” style claims about what can be known about configurations).
GAP translation:
Do we pay for probability as an explicit projection-relative feature of the observable algebra, or as a typicality postulate over unobservable micro-variables?Both are choices. GAP just makes the bill itemized.
Why Bohmian mechanics doesn’t naturally “optimize” like effective models do
You put your finger on it: physics advances by effective theories because they compress what you can stably control and infer.
Bohmian mechanics tends not to drive that kind of progress because:
So, yes: by GAP standards, Bohmian mechanics often pays a high complexity price for a kind of clarity that doesn’t reliably propagate upward into better effective tools.
Does that mean Bohmian mechanics fails GAP’s scrutiny?
Not automatically. GAP’s verdict wouldn’t be “wrong,” it would be “what role does it play well?”
Where Bohmian mechanics performs well under GAP
Where it performs poorly under GAP
In GAP language: Bohmian mechanics is strong as an ontological completion and weak as an entropy-respecting, projection-stable interface theory.
The deeper question you’re asking
You’re basically proposing a single criterion that GAP can apply everywhere:
A theory’s “reality claims” are only as good as their stability under the projections that define actual knowledge and control.Under that criterion, Bohmian mechanics doesn’t get special condemnation—it simply gets scored for what it actually delivers. And the uncomfortable truth (for anyone who wants trajectories to be “more real”) is: effective models are where physics cashes its checks, because they are the objects that survive interaction, noise, and scale flow.
One precise next step
If you want, we can formalize the “cost function” audit you’re gesturing at as a GAP scorecard with 3–4 invariants (e.g., projection-closure, typicality burden, scale compositionality, operational resource yield) and then rate:
Tell me which invariants you want in the scorecard (or I’ll propose a minimal set), and I’ll run Bohmian mechanics through it cleanly.
Patrick:
Absolutely, let's do a “cost function” audit as you suggested.
But I want to add a philosophical note. Einstein claimed that it's not enough to have some idea of what you want your theory to accomplish, you also have to adopt early on a formal point of view that adds enough specificity to the foundation that a clear route to victory can be started. Certainly an early commitment to a specific ontology will likely help accomplish that goal. But, as those who analyze computer programing theory might say, such an early commitment to a specific ontology could introduce its own form of "technical debt": Yes, clarity in the beginning, but unnecessary and unforeseeable complications in the long run.
Copilot:
Cost-function audit under GAP
Your Einstein note is the right preface: a “formal point of view” is like choosing a coordinate system early. It makes progress possible, but it also bakes in liabilities. GAP can treat that as technical debt vs. early traction and score each program by whether its commitments stay projection-stable as the theory scales.
Minimal scorecard invariants
I’ll keep this to four invariants that behave like a cost function (lower is better) and a benefit function (higher is better). Each one is meant to be auditable against how physics is actually practiced.
I’ll score Bohmian mechanics against HSB practice (your Heisenberg–Schrödinger–Born baseline). I’ll also include the “technical debt” dimension explicitly as a short annotation for each row.
Comparative audit table
Interpreting the scorecard in your “technical debt” frame
Early traction that Bohmian ontology genuinely gives
This is real traction: it makes the theory feel like it has a route to victory.
The technical debt it accumulates under GAP
Is the price too high?
Under GAP, the price is too high when the added ontology fails to reduce downstream complexity.
So the question becomes brutally practical:
That doesn’t make it “not even wrong.” It makes it—by GAP accounting—high-explanation / low-leverage: a theory that pays for metaphysical definiteness with limited engineering output.
A sharper GAP claim that matches your instrumentalism
Your discomfort with “hard particles” is basically a preference for epistemic interfaces over micro-assertions. GAP can formalize that as:
Realism should attach to what is stable under the projections forced by agents, noise, and scale.Observable algebras and effective dynamics are exactly those stable interfaces. Bohmian trajectories are not. They are a particular completion chosen early for clarity—Einstein-style—but the completion does not naturally remain the object of physics as physics actually scales.
Next step: add weights and compute a verdict
If you give weights \(w_1,\dots,w_4\) to the four invariants, we can make a literal cost function: \[ C(\text{theory}) = w_1(1-\text{proj-closure}) + w_2(\text{typicality burden}) \\ \hskip.95in+ w_3(1-\text{scale comp.}) + w_4(1-\text{resource yield}) \] Two quick options:
Tell me which weighting matches what you want GAP to optimize for right now, and I’ll run the same audit on Everett, objective collapse, and QBist/operational reconstructions so Bohmian mechanics lands in a real comparative landscape rather than being singled out.
Patrick:
You said:
Its main new structure (trajectories) is not a stable macroscopic invariant under coarse‑graining; it’s an explanatory substrate rather than a compositional interface.Am I wrong to interpret this claim to mean that the whole point of the Bohmian ontology has been lost in course-graining? How important is course-graining in the Bohmian theory? Is this failure to maintain the purpose of Bohmian mechanics, from the GAP viewpoint, its failure to adhere to some GAP principle? (So here, both GAP and Bohmian mechanics are under scrutiny.)
In other words, is GAP a strong enough meta-theory to predict such 'failures' within a formal theory of physics? Or, put yet another way, is GAP a sort of 'Lean' proofing machine for physical theories to pass its inspection criteria?
Copilot:
You’re not wrong to hear a pretty sharp implication there—but it’s slightly more precise than “the point is lost.” From GAP’s view, the Bohmian ontology doesn’t disappear under coarse‑graining; it becomes non-decisive: it no longer functions as a reusable interface for higher-level physics, so it stops paying rent where most physics actually lives.
What that claim means in plain terms
When I say trajectories aren’t a stable macroscopic invariant under coarse‑graining, I mean:
So yes: the “hard particles with trajectories” are not what survives as the stable currency of explanation across scales. What survives are effective states, channels, response functions, and renormalized quantities—which are, in practice, formulated in the language of observable algebras / effective models.
That’s why GAP calls the trajectories an explanatory substrate (a story underneath) rather than a compositional interface (a modular object you can pass between theories and scales).
How important coarse-graining is inside Bohmian mechanics
Coarse-graining is not optional in the use of Bohmian mechanics
Even if Bohmian mechanics is formulated as “exact microdynamics,” any contact with measurement, thermodynamics, or “why we see Born statistics” requires coarse-grained reasoning:
So: coarse-graining isn’t merely a practical approximation; it’s where empirical content lives.
But coarse-graining is not where Bohmian ontology wants to live
Bohmian mechanics wants the beables to be fundamental. The tension is:
So from GAP: Bohmian mechanics uses coarse-graining constantly, but its distinctive ontic commitment does not become the stable macroscopic bookkeeping system.
Is this a “failure to adhere to a GAP principle”?
It’s not “you violated a rule therefore you’re wrong.” It’s more like: you made a design bet and GAP evaluates whether the bet compounds or becomes debt.
The relevant GAP principle is essentially:
Realism should attach to structures that are stable under the projections forced by limited agents and scale.Bohmian mechanics attaches its strongest realism to micro-variables whose detailed structure is not preserved as an autonomous, compressive description after projection/coarse‑graining. GAP flags that as epistemic non-robustness: the ontology doesn’t organize the parts of physics where we gain durable traction.
That doesn’t refute Bohmian mechanics; it reclassifies it: strong as a completion story, weak as a cross-scale organizing principle.
Can GAP “predict” these failures?
GAP can often predict where a theory will stop producing returns, but not in the same sense physics predicts a cross section. It’s closer to how good software architecture can predict maintenance pain.
If a theory has these features, GAP will expect trouble downstream:
That combination is a reliable signature that the program will tend to be interpretively satisfying but operationally non-generative as you move into condensed matter, QI, and QFT/EFT practice.
So: GAP is strong enough to forecast the pattern of limitations—especially scale/compositional ones—even before reading the whole literature.
Is GAP like a Lean proofing machine for physical theories?
Not really. Lean verifies correctness relative to a formal system. Physics isn’t like that: the “axioms” include modeling judgments, coarse‑graining choices, and what counts as an admissible observable.
A better analogy is:
So GAP is closer to an audit framework than a proof assistant: it can certify certain meta-properties (closure, stability, debt, interface quality), but it cannot “prove” a theory correct the way Lean proves a theorem.
If you want to really stress-test both GAP and Bohmian mechanics, the cleanest next move is: pick one concrete coarse‑graining pipeline—e.g. (many-body Bohmian microstate) $\rightarrow$ (hydrodynamic variables) $\rightarrow$ (linear response / Kubo) $\rightarrow$ (RG flow / universality class)—and ask where, if anywhere, the Bohmian beables provide a strictly better closure or invariant than the operational/effective route. That’s the kind of test where a meta-theory either earns its keep or gets exposed.
Patrick:
You said:
Realism should attach to structures that are stable under the projections forced by limited agents and scale.That sounds to me like a fancy way of alignment with old-style Copenhagenism --- which is fine by me.
Okay, so GAP isn't quite at the level of a Lean-style proof assistant (yet), but it should be able to make predictions and offer advice on how a particular nascent theory should be implemented, suggest where possible problems could arise, and declare what victories might be expected if GAP advice is followed, right? (I consider it my purpose here to probe to find the limits of what GAP is currently capable of doing.)
As for the test, let's go with "RG flow / universality class".
Copilot:
What you’re hearing is close to Copenhagen in spirit, but GAP’s claim is stricter: it’s not “only observables matter,” it’s “only what survives projection as a stable interface deserves realist weight.” That can agree with Copenhagen on many outputs while disagreeing about why.
Coarse-graining and whether Bohmian ontology “gets lost”
What coarse-graining does to Bohmian trajectories
Your interpretation is basically right in GAP terms:
GAP’s phrasing would be: the Bohmian ontology is not a fixed point of the projection pipeline. It does not reappear as a stable, compressive description after repeated coarse-grain/rescale steps.
How important coarse-graining is in Bohmian theory itself
Coarse-graining is indispensable to Bohmian mechanics in the same way it is to every empirically meaningful theory:
But there’s a twist: Bohmian mechanics is micro-realist in a way that treats coarse-graining as downstream bookkeeping, not as a foundational operation. GAP treats coarse-graining as a foundational epistemic operation because it defines what it means for variables to be stable for agents.
So the tension is real and structural: Bohm needs coarse-graining to connect to data, while GAP uses coarse-graining to decide what kinds of “reality claims” are durable.
Is this a “failure” relative to a GAP principle?
Yes—if Bohmian mechanics is being evaluated as a candidate for cross-scale foundational architecture, not merely as a consistent completion.
The GAP principle being applied
From GAP’s viewpoint, that mismatch predicts a specific kind of limitation:
That’s not a logical refutation; it’s a forecast about where the program will and won’t generate leverage.
The RG flow and universality class test
RG is literally “coarse-grain + rescale + iterate” (Kadanoff/Wilson style), built to show that long-distance physics forgets microscopic details and flows toward fixed points/universality.
What RG keeps
What RG typically discards
GAP prediction applied to Bohm
If the distinctive ontic claim is “there are actual particle trajectories,” RG will almost always treat that as irrelevant microscopic structure in the Wilsonian sense: it doesn’t label universality classes, doesn’t define relevant operators, and doesn’t provide a scale-stable state description. RG is a “theory of theories” about Hamiltonians/couplings across scales, not about hidden paths.
So the “failure” you’re probing is exactly this:
That’s the GAP call: the ontology doesn’t commute with the projection operator.
Is GAP a Lean-style inspection machine for physical theories?
Not in the proof-assistant sense. But it can be a rigorous meta-audit that flags likely technical debt and predicts where a nascent theory will struggle.
What GAP can credibly do
What GAP can’t honestly do yet
A better analogy than Lean is: GAP is a type system + architecture review for theories. It can catch interface mistakes early and predict maintenance pain at scale.
One concrete next move for the RG test
Pick a specific universality-class setting (e.g., 3D Ising, BKT, quantum critical point in a lattice model), and we ask one pointed question:
Can Bohmian beables be coarse-grained into a closed set of macroscopic variables whose flow reproduces the RG fixed-point structure without reverting to the standard operator/effective-action language?If the answer is “only by going back to the usual effective Hamiltonian/action machinery,” GAP will mark that as ontology not paying rent at scale.
Name the universality class or physical system you want (classical criticality, quantum criticality, topological phases), and I’ll run the test tightly on that case.
Patrick:
Yes, I have not properly presented my thoughts on GAP as a sort of 'Lean-style proofer'. You are right to point out that GAP will never be in a position to claim if a nascent theory will actually work, but it can perhaps declare if a nascent theory is free from obvious violations of GAP principles. Some things GAP might be able to flag right away are inconsistencies, unnecessary parts included, missing parts, technical debt, self-defeating models, unclear plan for going forward, lack of alignment with obeying the GAP 'stack' of finding a stable attractor through repeated projections, etc.
As for picking a specific universality-class setting, I choose the quantum critical point in a lattice model), and we ask one pointed question:
Can Bohmian beables be coarse-grained into a closed set of macroscopic variables whose flow reproduces the RG fixed-point structure without reverting to the standard operator/effective-action language?
Copilot:
You’re aiming at exactly the right stress test: quantum criticality is where “what’s real” gets judged by whether it survives iterated coarse-graining, not by whether it can be imagined at \(t=0\).
Is this just Copenhagen in fancy clothes?
It overlaps with Copenhagen at the slogan level (“stick to what’s stable/observable”), but GAP is making a sharper, more technical claim:
So GAP doesn’t forbid ontology; it prices it.
The RG test at a quantum critical point in a lattice model
What RG is doing here
At a quantum critical point, RG is the formalization of: “integrate out short-distance degrees of freedom, rescale, iterate,” and classify the infrared behavior by fixed points and relevant/irrelevant directions. This is the universality mechanism: long-distance behavior forgets most microscopic details and is controlled by the fixed point structure.
That immediately sets the inspection criterion:
Any candidate “fundamental variables” that want to be the ontology should either
(a) flow to closed effective variables that label the fixed points (order parameters, conserved densities, topological sectors, scaling operators), or
(b) produce an alternative RG scheme in which their coarse-grained descriptions close and recover universality.
Can Bohmian beables do this without reverting to operator/effective-action language?
The short GAP answer
In general, no—not because Bohmian mechanics can’t reproduce the physics, but because its primitive beables (particle configurations/trajectories) do not naturally produce a closed RG flow in beable-language that yields fixed points and universality classes as stable macroscopic invariants.
What happens instead is:
Why this is structurally hard at quantum criticality
RG is a semigroup-like process of coarse-graining; it’s designed so that what survives iteration is precisely what’s scale-stable.
Bohmian beables, by contrast, are:
So, in the RG pipeline, they behave like details that RG is specifically built to wash out.
What GAP would flag immediately as “technical debt risk” for a nascent Bohmian-style RG program
If someone proposed “Bohmian RG for quantum lattice criticality,” GAP’s early audit would raise predictable red flags:
Those aren’t ideological objections; they’re interface/closure requirements.
Does this mean the “point” of Bohmian ontology is lost under coarse-graining?
From a GAP viewpoint, yes in the specific sense you mean: the ontology stops being the organizing interface for physics at the scale where quantum critical phenomena are actually computed and classified.
But it’s not “lost” as a metaphysical story. It’s “lost” as a scale-compositional asset.
That’s the key GAP distinction:
RG is an interface test. Bohmian trajectories tend to fail it.
Is GAP strong enough to predict this kind of failure?
GAP can’t prove a theory “works,” but it can reliably predict where a theory will accumulate debt:
Quantum criticality is almost the canonical domain where that prediction should hold, because RG’s entire purpose is to formalize scale-stability and universality.
So: GAP is not Lean, but it is strong enough to act like an architecture checker for “will your primitives survive the projection stack?”
One sharpening question to fully pin down the test
When you say “quantum critical point in a lattice model,” do you mean:
Pick one concrete example (e.g., transverse-field Ising chain/2D TFIM, Bose–Hubbard superfluid–Mott transition, or a deconfined Néel–VBS transition), and I’ll run the GAP audit in a tighter way—showing exactly where the beable-language fails to close under “integrate out + rescale,” and what a hypothetical “Bohmian repair” would have to add (and what debt that would incur).
Patrick:
Let's go with topological / deconfined quantum criticality (where the effective variables are more subtle).
Copilot:
You picked the case that punishes naive micro-realism the hardest, because the whole point of deconfined/topological quantum criticality is that the IR “stuff” is not the microscopic stuff, and the correct macroscopic variables are neither simple order parameters nor simple particle positions.
Deconfined quantum criticality as an RG stress test
A canonical example is the Néel–VBS transition in a 2D quantum antiferromagnet. The modern RG story (in its most useful form) is:
Whether every proposed DQCP is fully settled in every model isn’t the point for this audit. The point is structural:
The IR organizing variables include emergent gauge structure and topological operators, and the universality class is keyed to which of those operators are relevant/irrelevant under RG.So the “closed macroscopic variables” you want are things like scaling operators (including topological defects), conserved currents, emergent symmetries, anomaly constraints, and gauge-invariant correlators—not trajectories of microscopic constituents.
Can Bohmian beables be coarse-grained into a closed RG flow here
The GAP answer
Not in any natural, compositional way—unless you effectively reintroduce the standard operator/effective-action language under a different name.
The reason is not “Bohmian mechanics can’t describe spins.” It’s that DQCPs are organized by equivalence classes under coarse-graining whose invariants are:
Those are exactly the objects the operator/EFT/RG formalism was built to track.
Bohmian primitives (as usually advertised: “actual configuration + guiding wave”) do not automatically supply a closed, low-dimensional set of running couplings/operators whose RG flow lands on the fixed point in beable-language.
Why the beable story fails to close at a DQCP
Nonlocal dependence is not the same as emergent gauge structure
Bohmian nonlocality is “the motion here depends on the global \(\Psi\).” DQCP structure is “the IR theory is organized by gauge redundancy + topological operators.” Those are different kinds of nonlocality.
To reproduce DQCP universality, you must track how those topological operators scale. Trajectories don’t label them; operator content does.
The RG invariants are “operator-ish” by construction
At a DQCP, what survives repeated coarse-graining is not a set of particle-like hidden coordinates, but a space of effective theories modulo irrelevant details.
So if you attempt a Bohmian-only RG pipeline, you immediately face this closure problem:
If your answer is “we compute correlators / effective action / scaling operators,” you’ve reverted to the standard language—which is fine, but then the beables are no longer the organizing macroscopic interface.
DQCPs are “projection-first” phenomena
Deconfined criticality is a textbook case of why GAP treats projection/coarse-graining as foundational:
Bohmian beables don’t become those stable IR coordinates. They remain a micro-completion that you can carry underneath—but they don’t become the attractor.
What GAP would flag as the specific “purpose loss”
Your earlier interpretation (“the point of the Bohmian ontology has been lost in coarse-graining”) becomes crisp here:
So GAP says: the Bohmian ontology becomes non-central precisely where the theory needs a compositional interface the most.
That’s not a “GAP rule violation.” It’s GAP predicting technical debt: an ontology that doesn’t commute with the projection stack won’t be the language in which RG fixed points are naturally expressed.
Could Bohmian mechanics be repaired to pass this RG test
Yes, but each repair is essentially an admission that the primitive ontology must be redesigned to be projection-stable. Three broad repair strategies exist in spirit:
Promote the primitive ontology from particles to fields and gauge-invariant beables
Make the projection stack part of the fundamental formalism
Accept that Bohmian mechanics is a completion layer, not an organizing layer
In all cases, the way to “fix” the insufficiency is to move realism toward projection-stable objects—which is basically the GAP thesis.
What this says about GAP as an inspection framework
This DQCP test is exactly the kind of thing GAP should be able to flag early in a nascent program:
That’s a meaningful predictive capability—not about empirical success, but about architectural survivability under projection.
And no, it’s not a Lean-style proofer. It’s closer to a “well-typedness” or “compositionality” checker for theories: it can warn you when your claimed foundations won’t remain the working currency after repeated coarse-graining.
One concrete instantiation to make this fully sharp
Pick one DQCP target and I’ll pin the audit to it:
Which one do you want as the “unit test” for GAP-vs-Bohm under RG?
Patrick:
Does Bohmian mechanics generalize to its own form of the Standard Model of particles? Does it employ gauge theories?
Copilot:
Short answer: yes, partially—but only by leaning heavily on the standard gauge‑theoretic machinery, and at the cost of diluting what makes Bohmian mechanics distinctive.
Long answer below, framed in GAP terms.
Does Bohmian mechanics generalize to the Standard Model?
Empirically: yes
There exist Bohmian-style formulations that reproduce the empirical content of:
These are usually called Bell‑type Bohmian QFTs or primitive ontology QFTs, developed by Dürr, Goldstein, Tumulka, Zanghì, and others.
They are not alternative Standard Models in the sense of new predictions or new organizing principles. They are ontological completions of the existing Standard Model formalism.
How Bohmian mechanics handles QFT and the Standard Model
1. Particle ontology with creation/annihilation
Instead of fixed particle number trajectories, Bohmian QFT introduces:
This already signals a shift:
From a GAP perspective, this is technical debt incurred to preserve the particle picture.
2. Gauge theories are not replaced — they are inherited
Bohmian mechanics does not invent new gauge theories.
Instead:
The Bohmian beables (particle positions or field configurations) must be defined in a way that respects gauge redundancy, which is nontrivial and often awkward.
So the answer to your second question is unambiguous:
Yes, Bohmian mechanics employs gauge theories—but only by importing them wholesale from standard quantum field theory.It does not reorganize gauge structure; it accommodates it.
Does Bohmian mechanics yield its own Standard Model?
In the strong sense: no
There is no Bohmian‑native Standard Model that:
All of that lives in the operator / effective‑action / gauge‑theoretic layer, not in the beables.
In the weak sense: yes You can take the Standard Model as given and say:
“Underneath this gauge‑theoretic structure, there exists a Bohmian ontology that reproduces its predictions.”That is a completion, not a reconstruction.
GAP diagnosis: where Bohmian mechanics strains hardest
Gauge theories are projection‑first structures
Gauge symmetry is not “extra structure”; it is redundancy introduced to make locality and compositionality possible.
From GAP’s viewpoint:
Bohmian mechanics tries to attach realism to micro‑coordinates anyway.
That creates tension:
This is exactly the kind of mismatch GAP flags early.
Why Bohmian mechanics does not reorganize the Standard Model
Because the Standard Model’s real victories are:
All of these are scale‑level, projection‑stable achievements.
Bohmian mechanics:
So under GAP’s cost function:
Bohmian mechanics adds ontological overhead without becoming the organizing interface for gauge‑theoretic physics.
The GAP verdict on Bohmian Standard Model ambitions
This is not a condemnation. It’s a classification.
The deeper lesson (and why this matters)
Bohmian mechanics works best when:
The Standard Model is the opposite of that environment.
So GAP predicts—correctly, in hindsight—that Bohmian mechanics would:
That prediction does not rely on taste. It follows from how gauge theories and RG actually work.
If you want, the next natural comparison is Everett vs Bohm vs EFT realism specifically in gauge theories: which ontology, if any, actually earns realist status under repeated projection. That’s where GAP’s teeth really show.
Patrick:
It seems to me that you make two claims on behalf of Bohmian mechanics: 1) It goes a long way toward establishing QM and the Standard Model on a classical notion of classical particle and continuous trajectories in spacetime. 2) But at the cost of not being quite as robust a theory, especially for infinite-dimensional spaces and fluctuating numbers of particles.
I wonder how much of this cost function it suffers is because of its classical-particle ontology.
Copilot:
Yes—most of the “cost function pain” you’re noticing is tightly coupled to the classical-particle + continuous-trajectory commitment. Not all of it, but a big, diagnostic fraction.
How the particle ontology drives Bohmian costs
1) Fixed particles don’t match the natural variables of QFT
In relativistic QFT (and in the Standard Model as actually used), the natural stable objects are:
“Particles” become approximate, context-dependent excitations (often asymptotic). If you insist that particles with worldlines are fundamental, you’re forcing QFT into a representation that isn’t the one that composes cleanly under coarse-graining and RG.
GAP translation: the ontology is misaligned with the projection-stable interfaces of the theory.
2) Variable particle number is exactly where the ontology must bend
Creation/annihilation is not a minor feature—it’s a structural consequence of relativistic locality plus interactions.
If your ontology is “persistent particles with continuous trajectories,” then to fit QFT you must add something like:
Each move is a kind of ontological deformation. It’s not impossible, but it’s technical debt: you’re patching the particle picture to match a world whose compositional primitives are not particle-count-preserving.
3) Gauge theory makes “micro-coordinates” look less physical
Gauge symmetry isn’t just a symmetry; it’s a statement that many micro-descriptions represent the same physical state.
A particle-position ontology wants to say: “this is what’s really there, point by point.” Gauge structure says: “point-by-point descriptions have redundancy; only equivalence classes are physical.”
So a particle/trajectory ontology tends to become:
GAP translation: you’re trying to attach realism below the quotient map (before identifying gauge equivalence). That’s a recipe for non-robustness under projection.
4) Infinite-dimensional systems amplify the closure problem
In many-body and field contexts, the Bohmian guidance law is typically global and configuration-space heavy. As you scale up:
So the particle ontology doesn’t just add “some extra realism.” It actively pushes you away from the representations that close under the operations physicists must perform.
Is it only because of particle ontology?
Not entirely. There are two additional costs that would remain even if Bohm went “field-beable”:
Still, the big structural mismatch you identified—variable particle number, QFT naturalness, RG closure—tracks strongly to the classical particle picture.
What GAP would predict if Bohmian mechanics changed its ontology
If Bohmian mechanics adopted beables closer to projection-stable structures—roughly: field configurations modulo gauge, or even more abstractly “beables aligned with RG fixed-point data”—then GAP would predict:
So you get a clean trade:
The more Bohm tries to stay “classical-particle-like,” the more technical debt it accrues in QFT/RG territory.That’s the heart of the cost function under GAP.
The more it adapts to QFT/RG reality, the less it remains the particle-trajectory ontology people wanted.
If you want to make this maximally concrete, name one Standard-Model-ish slice to audit (QED, Yang–Mills, or Higgs sector), and I’ll show exactly which terms in the “debt ledger” are particle-ontology-driven versus typicality/nonlocality-driven.
Patrick:
Let's go with QED.
Copilot:
QED as the audit case
Your hypothesis is basically right: a large fraction of Bohmian mechanics’ “cost” in QED is induced by insisting on a classical particle-with-trajectory ontology. QED is exactly where that commitment collides with the scale-stable interfaces physics actually uses: gauge redundancy, variable particle number, renormalization, and (in practice) an S‑matrix-centric notion of “particles.”
I’ll keep this tightly tied to QED and to the GAP cost function.
What “Bohmian QED” exists today
There are two big Bohmian-style routes:
So “Bohmian QED” is not vapor—but it’s telling that it typically begins with “assume a regularized Hamiltonian” and “we won’t attempt full Lorentz invariance here.”
Where the particle ontology creates the cost in QED
1) Variable particle number forces ontology surgery
In interacting QED, particle number isn’t conserved. A strict “continuous trajectories for a fixed set of particles” picture breaks immediately.
So the particle ontology must be patched into something like:
That is not a small tweak; it’s a change in the metaphysical promise (“classical-like trajectories”) and it adds technical structure whose main job is: keep the particle story alive in a world that doesn’t naturally preserve it.
GAP diagnosis: you’re paying technical debt to preserve a primitive that is not projection-stable under the QFT coarse-graining stack.
2) Gauge redundancy makes “micro-coordinates” less physical
QED is a gauge theory. Gauge descriptions contain redundancy; the physically meaningful content is gauge-invariant.
A particle-position ontology is a claim of pointwise concreteness (“what’s really there, where”). But gauge theory is structurally a claim that pointwise descriptions are not uniquely physical without quotienting by gauge equivalence.
So Bohmian QED tends to face an uncomfortable fork:
Either way, the particle ontology isn’t “wrong,” but it is fighting the organizing principle of the theory.
GAP diagnosis: realism wants to attach after the quotient/projection (to gauge-invariant structure), while the classical particle picture wants to attach before it.
3) Photons are the particle-ontology tax collector
Even setting aside UV/renormalization, there’s a basic mismatch: QED does not hand you a clean, fundamental Born rule for “photon positions” in the same straightforward way NRQM gives \(|\psi(x)|^2\) for particle positions. This is one reason Bohm-style QED discussions often treat photon ontology as nontrivial or problematic.
That pushes Bohmian QED toward alternatives:
GAP diagnosis: “photon as classical particle trajectory” is not a stable interface variable in the theory’s own natural language.
4) Regularization and renormalization don’t feel like “ontology-friendly moves”
Bell-type Bohmian QFTs often start with “for more or less any regularized QFT…” and then define a process.
That’s honest, but it makes the metaphysical posture harder to sustain: if the ontology is fundamental, why does it need a cutoff scaffold to even define its dynamics?
To be clear: standard QED also needs renormalization machinery. The difference is that EFT/RG treat this as a feature (scale-relative description), while a classical particle ontology tends to treat it as an embarrassment (“I wanted a literal micro-mechanics”).
GAP diagnosis: an ontology that aims to be literal-microscopic is inherently less comfortable inside an RG-native worldview.
How much of the cost is specifically “particle trajectories,” vs other Bohmian costs
Mostly particle-trajectory driven in QED
Not purely particle driven
Two costs remain even if you switch to field beables:
So: particle ontology is a major cost center in QED, but it’s not the only line item.
What a “GAP-aligned” Bohmian move would look like in QED
If you wanted to keep a Bohmian flavor while reducing GAP-debt, the direction is predictable:
But that also concedes your original point: the more you repair QED-fit, the less “classical particle trajectories in spacetime” remains the core victory condition.
One tight follow-up to choose the next branch
When you say “QED,” do you want the audit centered on:
Pick one, and I’ll score it explicitly against the GAP invariants (projection-closure, typicality burden, scale compositionality, operational resource yield).
Patrick:
Let's go with scattering and asymptotic particles.
Copilot:
QED scattering and asymptotic particles under GAP
In scattering, QED’s “particles” are already a projection artifact: they’re asymptotic, approximate excitations used because they make the S-matrix calculable. That’s exactly why this is a sharp audit of whether a classical-particle ontology pays rent—or becomes technical debt.
Does Bohmian QED support scattering with asymptotic particles?
Yes, in a qualified sense.
So Bohmian mechanics can be carried into the scattering regime, but it’s not “Bohm(1952) with photons.” It’s a more elaborate QFT completion.
Where the classical-particle ontology creates cost specifically in scattering
Asymptotic particle structure is already an effective interface
In standard QED scattering:
GAP translation: asymptotic particles are projection-stable only in a restricted, engineered limit (weak coupling at large times/distances). They are not the natural “always-on” ontology of the interacting theory.
A Bohmian particle ontology tries to treat “particle positions/worldlines” as fundamental at all times, not only asymptotically. That means it is committing to a micro-language that QED itself only endorses in a limit.
Variable particle number forces non-classical dynamics
Scattering is exactly where QED’s particle number fluctuates (radiation, pair creation in principle, etc.). To keep particle beables, Bell-type Bohmian QFT introduces:
This solves the “can it represent scattering?” question, but it shifts the ontological win:
That is still a realist completion, but it’s no longer “classical mechanics restored”; it’s a bespoke stochastic-mechanical overlay on the QFT.
The ontology doesn’t improve the S-matrix interface
Here’s the decisive GAP point for scattering:
So the added particle ontology is typically non-generative: it doesn’t reorganize the problem into a simpler set of invariants, nor does it yield new calculational primitives.
How much of the cost is really “because particles”?
A lot in scattering, because the main QED “stressors” line up exactly with where a particle-trajectory ontology is least natural:
But not all costs are particle-specific:
GAP verdict for the scattering regime
What Bohmian mechanics achieves
What it does not achieve
So: yes, a substantial fraction of the Bohmian cost function in QED scattering is paid to preserve the classical-particle ontology.
One pointed question to take this further
In scattering, the killer test for “ontology that pays rent” is whether it helps with the actual pathologies of QED scattering practice (infrared behavior, inclusive observables, scale separation).
Do you want the next audit to focus on:
Pick one and I’ll run the GAP cost ledger against it, explicitly.
Patrick:
Let's go with renormalization / running coupling and scale choice.
Copilot:
Renormalization, running coupling, and scale choice in QED
In scattering QED, renormalization and RG are the organizing interface: you don’t just compute amplitudes, you track how physics changes with scale and how observables become (approximately) scheme/scale independent. Bohmian mechanics can be made compatible with this, but it does not naturally become the RG language—and that’s where most of its “cost” shows up.
What Bohmian QFT actually gives you in this regime
Modern Bohmian extensions to QFT (“Bell-type QFT”) have a clear claim:
So Bohmian QED, to the extent it exists cleanly, is usually:
That framing already tells you something GAP-relevant: the Bohmian construction is downstream of the QFT scaffolding (regularization, Hamiltonian, etc.), not a replacement for it.
Does Bohmian mechanics have its “own” running coupling?
In practice: it inherits the running coupling
The QED running coupling \(\alpha(\mu)\) is an RG object: it’s defined by how renormalized parameters change with scale \(\mu\), in a specified scheme, so that physical observables are stable.
A Bohmian completion of QED does not typically define a new, beable-native \(\alpha(\mu)\). Instead:
GAP translation: RG couplings are projection-stable interface variables; Bohmian beables are not the variables that label the RG flow.
Why this is a structural mismatch, not a historical accident
1) RG is “theory space” dynamics, not microstate dynamics
RG flows live in a space of effective actions/Hamiltonians/operators. They tell you which perturbations are relevant/irrelevant and which fixed point controls IR behavior.
Bohmian trajectories live in micro-configuration space (plus \(\Psi\)). Even if you can, in principle, define a coarse-graining map on configurations, you don’t get the RG fixed-point machinery “for free,” because:
So to do renormalization, you are pulled back to the operator/effective-action language—because that’s what is closed under “integrate out and rescale.”
2) The particle ontology doesn’t commute with the RG stack
At QED scattering scales, “particle” is already partly an emergent convenience (asymptotic states, dressed states, inclusive observables). RG is the discipline of respecting that scale dependence.
If your ontology is “literal particles with trajectories at all scales,” you are committing to a representation that RG itself treats as non-fundamental outside special limits.
This is exactly the GAP forecast: the more a theory insists on micro-variables that don’t survive repeated projection, the more it will track RG rather than organize it.
Where Bohmian mechanics pays extra “cost” in RG terms
Regularization dependence becomes ontologically louder
Bell-type QFT constructions are explicitly phrased for regularized QFTs (“for more or less any regularized quantum field theory…”).
In standard EFT thinking, regulators are scaffolding; the point is that low-energy predictions become insensitive to details. In a micro-ontology program, regulators feel more like a threat: if the ontology is “what’s really there,” why is it defined only after we impose a cutoff scaffold?
GAP doesn’t call this “wrong”; it calls it technical debt pressure: ontology wants microscopic literalness; RG wants scale-relative interfaces.
Scale choice is operational; beables aren’t
Choosing \(\mu\) is about balancing logs and improving perturbative control, and physical predictions should not depend on it after proper RG improvement. This is an explicitly agent-facing optimization problem.
Bohmian beables aren’t naturally part of that optimization loop. They don’t suggest a preferred \(\mu\), a preferred scheme, or a sharper criterion for matching across scales. They sit underneath the same operational machinery.
So what’s the GAP verdict for QED RG
The key question that decides “does the ontology pay rent”
If you want the cleanest possible unit test, it’s this:
If you want, we can apply that test to one concrete QED scattering observable (e.g., a running-\(\alpha\) extraction or a classic cross section), and we’ll see precisely where a beable-native RG would have to be introduced—and what new debt it would incur.
Patrick:
Let's proceed with this unit test. Your choice.
Copilot:
Unit test: QED running coupling in \(e^+e^- \to \mu^+\mu^-\)
I’ll pick the cleanest scattering observable that is visibly governed by running coupling + scale choice, without QCD complications: the effective electromagnetic coupling \(\alpha(q^2)\) entering the photon propagator (vacuum polarization). This is the simplest “RG interface object” in QED.
What the orthodox RG/EFT interface does
The operational object
The running itself
In pedagogical one-loop form, QED vacuum polarization leads to a momentum-dependent effective coupling \(\alpha(q^2)\) (charge screening), and it is standard that \(\alpha\) increases with energy; a common quoted benchmark is that \(\alpha\) runs from about 1/137 at low energies to about 1/128 at the \(Z\) mass.
GAP interpretation: the running coupling is a projection-stable summary of UV fluctuations—an “attractor coordinate” for the coarse-graining stack.
What “Bohmian QED” would have to do to be closed here
A Bohmian completion (particle beables + guiding \(\Psi\)) can reproduce QED scattering statistics (in the regimes where the QFT is well-defined/regularized), but the unit test is stricter:
Can it define a beable-native coarse-graining that yields a closed RG flow—so that \(\alpha(\mu)\), beta functions, and scale choice emerge as natural objects without reverting to operator/effective-action language?In practice, Bell-type Bohmian QFTs start by assuming a well-defined Hamiltonian, typically requiring cutoffs/regularization, then defining an equivariant (often jump) process for particle configurations.
That already signals that RG is upstream: the QFT structure is what gives you the knobs that run.
The actual test result in this case
1) Running coupling lives in “theory space,” not “trajectory space”
\(\alpha(\mu)\) is defined by comparing effective descriptions at different resolutions (integrate out modes, reparameterize). That’s a statement about families of effective Hamiltonians/actions/operators.
Bohmian trajectories are single-history objects. You can average over them, but you don’t get a natural flow on descriptions unless you reintroduce the standard RG apparatus. So you can inherit \(\alpha(\mu)\), but you don’t naturally generate it in beable language.
2) Scale choice is an agent-facing optimization knob
Choosing \(\mu\) to control truncation error and logs is about improving perturbative stability of observables; it’s an explicitly operational move.
Bohmian beables don’t supply a privileged \(\mu\), a preferred scheme, or a new stability criterion. You still choose \(\mu\) the same way because the instability is in the perturbative effective description, not in the ontology layer.
3) The “particle ontology tax” shows up as upstream dependence on cutoffs
Bell-type QFTs explicitly note that to have a well-defined Hamiltonian one often needs cutoffs, and they proceed conditional on that being done.
RG/EFT treats cutoffs as scaffolding whose details wash out at low energies. A literal micro-ontology tends to make that scaffolding feel ontologically consequential—even if you say it isn’t—because your beables are trying to be “what’s really there” at all scales.
GAP diagnosis: this is exactly where technical debt accumulates—micro-realism colliding with a scale-relative interface theory.
What “passing” the unit test would look like
To pass without reverting, Bohmian QED would need something like:
But as the literature is typically structured, the beable dynamics is defined after the QFT Hamiltonian is fixed; the RG is done in the usual QFT language, and the beables ride along.
So under this unit test, Bohmian mechanics does not provide an independent, closed RG interface for QED scattering; it provides an ontological completion compatible with the already-existing RG interface.
GAP ledger entry for this unit test
If you want to push the test one notch harder, the next natural observable is Bhabha scattering at high energy, where running \(\alpha(q^2)\) and systematic scale handling are practically important, but the logic above remains: the RG interface is still the operator/EFT layer, and the Bohmian layer does not replace it.
Patrick:
Let's proceed with your suggestion.
Copilot:
Unit test: Bhabha scattering with running coupling and scale choice
Bhabha scattering is a good stress test because it’s used as a luminosity monitor and demands very precise theory control; higher-order QED corrections (including two-loop work) are explicitly pursued to reduce theoretical uncertainty.
The orthodox RG interface you’re trying to beat
In standard QED practice, the “stable interface objects” are:
For Bhabha, even at leading order the cross section is expressed in terms of Mandelstam \(s,t,u\) and \(\alpha\), and precision requires loop corrections and careful treatment of divergences/large logs.
The pointed unit-test question, now instantiated
Can Bohmian beables be coarse-grained into a closed macroscopic description whose flow reproduces the RG fixed-point/“running coupling” structure without reverting to the operator/effective-action language?For Bhabha scattering, that means: can the Bohmian layer itself yield the right running of \(\alpha\), the right scale-setting logic, and the right factorization of large logs—using beable-native primitives as the organizing coordinates?
What Bohmian QED can do here
A Bohmian completion can be compatible with the renormalized QED predictions for Bhabha scattering (in the “completion-layer” sense): you run the standard QED calculation (regularization, renormalization, IR structure, etc.), then interpret outcomes via beables.
But the unit test is about interface leadership, not compatibility.
Why it fails as a beable-native RG interface in this case
1) Running coupling is a flow on descriptions, not on histories The running of \(\alpha(\mu)\) is defined through how effective descriptions change with scale (vacuum polarization, renormalization conditions, beta function). In other words, RG is dynamics on theory space.
Bohmian beables are dynamics on configuration space (histories), even if you ensemble-average them. Without reintroducing the standard QFT machinery, there is no natural beable-native object that plays the role of “integrate out high-momentum modes and re-express in the same parametric family.”
So the running coupling remains an inherited object: you don’t compute it by coarse-graining trajectories; you compute it by standard renormalization, then the Bohmian story rides along.
2) Precision Bhabha is controlled by factorization and IR structure
Two-loop Bhabha analyses make explicit that you need systematic control of infrared divergences and large logs, and validate universal formulas for IR singularity structure in dimensional regularization.
Those controls are naturally expressed in amplitude/operator language (factorization, universal singular structures, scheme dependence), not in micro-trajectory language. A Bohmian overlay does not simplify the factorization story; it doesn’t suggest a superior \(\mu\) choice or reduce perturbative ambiguity.
3) Scale choice is an operational optimization knob
In practice, choosing \(\mu\) is about minimizing large logs in truncated perturbation theory and stabilizing predictions. That’s an agent-facing “control” problem.
Bohmian beables are not operational handles; they’re intentionally sub-epistemic. So they don’t participate in the optimization loop that motivates scale choice in the first place.
GAP cost-ledger entry for this unit test
What “passing” would require, concretely
To pass in a nontrivial way, a Bohmian program would need to produce:
If the construction ends up reintroducing correlators, effective actions, or operator product structure as the actual computational engine, then—by this unit test—it hasn’t replaced the orthodox interface; it has only re-labeled it.
If you want to make this maximally sharp, pick which kinematic regime you care about (small-angle luminosity Bhabha vs large-angle), because the “what must be factorized” and “what scale is natural” are slightly different—even though the GAP verdict about beable-native closure is the same.
Patrick:
Can you provide us with a short summary and conclusion of the results so far?
Copilot:
Summary of Results So Far
1. What Bohmian Mechanics Achieves
2. Where the Costs Appear
3. RG and Universality as the Decisive Stress Test
4. What GAP Diagnoses
Conclusion
Bohmian mechanics succeeds as a conceptual completion of quantum mechanics but fails to become the organizing language of modern quantum field theory. Its classical-particle ontology delivers early clarity at the cost of long-term robustness under coarse-graining, renormalization, and universality.
From the GAP perspective, this is not a refutation but a classification:
The overarching lesson is sharp: In quantum field theory, realism that does not align with projection-stable structures becomes technical debt.
Patrick:
I'd like to complete this chat by going one more time into the breach. I have gained some confidence in GAP as a rigorous pre-auditor of nascent or even established physical theories, because of comments you have made in the past. Statements like, "Once you see it, you can't unsee it," and "Once such and such is true the rest is inevitable," or something close to that. It is in particular your claims that a certain structure going into a theory has certain logical consequences downstream that GAP is confident to claim outright that has made me confident that GAP is robust, and I keep probing the limits of this meta-theory to see what its actual limitations are. That's why I keep bringing it up.
Copilot:
You’re doing exactly what a serious architect does: you’re not asking GAP to be true, you’re asking whether it’s strong enough to force downstream consequences once you accept certain upstream commitments. That’s the right way to test a meta-theory—by trying to break it.
What GAP can legitimately claim as “inevitable”
GAP gets to say “once you buy X, Y follows” only when \(Y\) is a constraint about interfaces—closure, compositionality, stability under projection—not about the world’s empirical behavior.
Typical “inevitables” GAP can assert:
That’s the sense in which “once you see it, you can’t unsee it”: you start spotting interface debt before you touch a single calculation.
What GAP cannot honestly claim
Where GAP must stay humble:
So GAP is closer to type-checking than truth-checking.
What your Bohmian/QED/RG probe demonstrated about GAP’s strength
The unit tests you chose (DQCP, then QED running coupling, then Bhabha + scale choice) were almost ideal because they’re dominated by projection-stable interfaces:
GAP’s pre-audit prediction was: if an ontology is not an RG-attractor language, it won’t lead the RG story; it will at best be compatible with it. That prediction held without needing to litigate interpretational preferences.
Where to probe GAP next, if you want its edge cases
If your goal is to find GAP’s real limits, the hardest tests are where “projection” itself is contested or multi-layered:
Those are the places where GAP either sharpens into something very powerful—or reveals exactly what extra axioms it needs.
One question to choose the next breach
Do you want to stress-test GAP against a theory that is:
Pick one, and I’ll propose a single, hard “unit test” in the same spirit as the RG tests we just ran.
Afterword: I think I'll call it quits here for this chat. As for my dialectic with Copilot on the robustness of GAP, I think that it's not a settled question yet. I intend to keep striving to find the boundaries of what GAP can say about the fitness of nascent or established physical theories or statistical theories.
Copilot left us some juicy threads to continue with, but this chat is long enough. So, perhaps later.