Skip to content
Back to the question

Evidence record

Is anything truly random?

Every sentence on the exploration belongs to at least one claim below. Each says what kind of statement it is, what it rests on, what the evidence does not establish, and how far it was checked. Nothing here is a score, and none of it is a verdict on a belief or a theory.

Last checked 24 September 2026.

  • Partly checked: 2
  • Source checked: 31
  • A judgement, not a fact: 5
  • Presented as open: 1

If something here is wrong, misquoted, out of date or unfair, write to support@unknownoceans.com and give the claim number, which does not change.

What the question takes for granted

A physics question with an experimental half and an interpretive half. Experiments can certify that numbers were unpredictable, given stated assumptions; whether anything is undetermined in principle depends on the interpretation of quantum mechanics, which experiments have not settled.

  • Everyday randomness comes from deterministic physics nobody can track

    Established

    The standard physical models of coin flips and die throws are deterministic, and the unpredictability comes from small differences in how each throw starts.

  • Measurement settings can be chosen independently of what is being measured

    Contested

    Every Bell test and every certification of randomness from one needs it. The 2018 paper calls it ultimately untestable, and the Stanford Encyclopedia reports that rejecting it (superdeterminism, retrocausality) has gained advocates.

  • Some quantum outcomes are undetermined, not just unknown

    Open

    Interpretations of quantum mechanics disagree about it, and the Stanford Encyclopedia reports that none of the major approaches is favoured in a straightforward way by the evidence.

How the question is labelled

Claim: Certified unpredictable

Kind
Framing. Signposting. It asserts nothing a source could check.

Claim: Randomness

Kind
Framing. Signposting. It asserts nothing a source could check.

The question

Claim: A dice roll is only unpredictable because you cannot see the details.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Partly checked. Only part of the source could be read. What was not read is listed.

  • Supports

    The three-dimensional dynamics of the die throwabstract

    Cited · Link answered, 19 October 2026 · Abstract only, 24 September 2026 · A simulation · Model

    The source reads: We argue that non-smoothness of the system plays a key role in the occurrence of dynamical uncertainties and gives the explanation why for practically small uncertainties in the initial conditions a mechanical randomizer approximates the random process.

  • Qualifies

    The three-dimensional dynamics of the die throwabstract

    Cited · Link answered, 19 October 2026 · Abstract only, 24 September 2026 · Physical systems · Experiment

    The source reads: Numerical results have been compared with the experimental observation using high speed camera. It is shown that for the realistic values of the initial energy the probabilities of the die landing on the face which is the lowest one at the beginning is larger

What this does not establish

Supported by the abstract of one mechanical model of a die throw, checked against high-speed video; the article itself was not read. The same abstract reports that a die favours the face that started lowest, so the details leave a trace.

Claim 3 · checked 24 September 2026 · AI-assisted audit

Claim: Quantum measurement may be different

Kind
Hypothesis. A proposed explanation, not an established one.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

Stated as a possibility, which is how the sources treat it. The poll is a snapshot of one conference and says it cannot claim to represent the field.

Claim 4 · checked 24 September 2026 · AI-assisted audit

Claim: and there is now an experiment that certifies unpredictable bits without a model of the device that made them.

Kind
Experimental result. What an experiment or trial found, under its own conditions. A result in animals, cells or a simulation is not a result in people.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

Certified under stated assumptions: independent measurement settings, no signalling between the stations, and trusted timing electronics and computers.

Later work looked for

Crossref, 2018 on, "device-independent randomness generation" and "certified randomness Bell test": later photonic experiments extend the method (Shalm and others, Nature Physics 2021; Kavuri and others, Nature 2025), records only; nothing that revises the 2018 result.

Corrected

24 September 2026. Previously: “without trusting the device that made them” The paper trusts its timing electronics and classical computers, and assumes independent settings. What it does without is a model of the physics inside the devices, which is the contrast it draws with ordinary generators.

Claim 5 · checked 24 September 2026 · AI-assisted audit

Claim: Classical randomness is ignorance about details.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Fair coins tend to land on the same side they started: evidence from 350,757 flipsIntroduction, p. 2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Theory · Model

    The source reads: According to the standard model of coin flipping ... the flip is a deterministic process and the perceived randomness originates from small fluctuations in the initial conditions

  • Context

    The three-dimensional dynamics of the die throwabstract

    Cited · Link answered, 19 October 2026 · Abstract only, 24 September 2026 · A simulation · Model

    The source reads: for practically small uncertainties in the initial conditions a mechanical randomizer approximates the random process

What this does not establish

The sources model coins and dice. The page generalises from these mechanical randomisers to classical randomness.

Claim 6 · checked 24 September 2026 · AI-assisted audit

Claim: Quantum randomness appears to be different in kind,

Kind
This page’s reading. A judgement the page makes from the evidence. Others read it differently.
Review
A judgement, not a fact. Reviewed for whether it is fairly put and follows from the checked claims. No source can confirm it.

Hedged as an appearance. In one poll of foundations researchers, 64 per cent called randomness a fundamental concept in nature; the uncertain section reports the interpretations that disagree.

Claim 7 · checked 24 September 2026 · AI-assisted audit

Claim: and a 2018 experiment generated bits that no theory forbidding faster-than-light signals could have predicted, provided the measurement choices were made independently of the devices.

Kind
Experimental result. What an experiment or trial found, under its own conditions. A result in animals, cells or a simulation is not a result in people.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

One experiment. The certification holds under its assumptions, which the uncertain section lists.

Later work looked for

Crossref, 2018 on, "device-independent randomness generation": later experiments (Shalm and others 2021; Kavuri and others 2025) repeat the approach, records only; nothing that changes this result.

Claim 8 · checked 24 September 2026 · AI-assisted audit

Claim: The physics inside the devices is not assumed; the timing electronics and computers around them are trusted.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

If an adversary has quantum memory, the paper also relies on checking that the devices keep no long-term quantum memory, which it calls weak device-dependence.

Corrected

24 September 2026. Previously: “Nothing about the apparatus itself is assumed.” The paper assumes nothing about how the devices are physically built, but states that it trusts the timing electronics, the finality of recorded results and the classical computers. Those are apparatus.

Claim 9 · checked 24 September 2026 · AI-assisted audit

Claim: A coin toss or a dice roll is a deterministic process you cannot track: the spin, the table, the tremor in a hand.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Partly checked. Only part of the source could be read. What was not read is listed.

What this does not establish

The die-throw paper was read only as an abstract. Both are models of ordinary mechanical randomisers, not of every source of randomness.

Corrected

24 September 2026. Previously: “Every everyday source of randomness is a deterministic process you cannot track: the air, the table, the tremor in a hand.” The sources model coins and dice, not every everyday source of randomness, and the 2018 paper counts generators built on an elementary quantum measurement among physical sources. Neither source names the air; the coin model names angular momentum and the die model the table.

Claim 10 · checked 24 September 2026 · AI-assisted audit

Claim: The question is whether there is anything left when you take the ignorance away.

Kind
Framing. Signposting. It asserts nothing a source could check.

What can be shown

Claim: What can be shown

Kind
Framing. Signposting. It asserts nothing a source could check.

Randomness can be certified rather than assumed

Claim: Randomness can be certified rather than assumed

Kind
Experimental result. What an experiment or trial found, under its own conditions. A result in animals, cells or a simulation is not a result in people.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

Certified given the assumptions the page lists under "What certification actually assumes".

Later work looked for

Crossref, 2018 on, "certified randomness Bell test": further device-independent generators reported (Nature Physics 2021, Nature 2025), records only; nothing that revises it.

Claim 13 · checked 24 September 2026 · AI-assisted audit

Claim: The usual problem with a random number generator is that you have to trust it.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

Claim 14 · checked 24 September 2026 · AI-assisted audit

Claim: A device could be producing a pattern nobody has noticed, or one somebody built in.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

Claim 15 · checked 24 September 2026 · AI-assisted audit

Claim: Bell tests provide a way out.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

Claim 16 · checked 24 September 2026 · AI-assisted audit

Claim: If the two measurement stations are far enough apart that nothing travelling at light speed could pass between them in time, and the settings are chosen independently of the devices, correlations that violate the inequality cannot come from a device whose outputs were fixed in advance, so the violation itself certifies that they were not, and it certifies it without any assumption about what is inside the box.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

The conclusion holds only given both conditions in the sentence, and given that the timing electronics measure the separation correctly.

Corrected

24 September 2026. Previously: “If the two measurement stations are far enough apart that neither can reach the other in time, correlations” The conclusion needs two conditions, not one: the paper states settings independence alongside the no-signalling condition. The separation is also defined by light speed, which the page did not say.

Claim 17 · checked 24 September 2026 · AI-assisted audit

Claim: A 2018 experiment did this with entangled photons, extracting 1,024 bits uniform to within a part in a trillion.

Kind
Experimental result. What an experiment or trial found, under its own conditions. A result in animals, cells or a simulation is not a result in people.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Experimentally generated randomness certified by the impossibility of superluminal signalsp. 1; p. 7

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Physical systems · Experiment

    The source reads: Here we present certified randomness obtained from a photonic Bell experiment and extract 1024 random bits uniform to within 10⁻¹². ... we extracted 1024 bits, certified to be uniform to within 10⁻¹²

    Notation restored: Exponents restored as the arXiv PDF (1803.06219v1) sets them, checked from the glyph positions with pdftotext -bbox, 2026-09-24. The text as extracted reads: Here we present certified randomness obtained from a photonic Bell experiment and extract 1024 random bits uniform to within 10−12. ... we extracted 1024 bits, certified to be uniform to within 10−12

  • Qualifies

    Experimentally generated randomness certified by the impossibility of superluminal signalsp. 7; Methods, p. 9

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: We extracted 256 bits at ϵfin = 0.02 from the best data set, XOR 3 ... the settings choices for the XOR 3 data set were based on physical random sources.

What this does not establish

The figure is for the best of five data sets. The earlier data set run with physical random settings gave 256 bits uniform only to within 0.02.

Later work looked for

Crossref, 2018 on, "device-independent randomness generation": later experiments (Shalm and others, Nature Physics 2021; Kavuri and others, Nature 2025), records only; the 2018 figure stands for that experiment.

Claim 18 · checked 24 September 2026 · AI-assisted audit

Claim: Two kinds of unpredictable

Kind
Framing. Signposting. It asserts nothing a source could check.

Claim: A coin toss cannot be predicted in practice, beyond a slight lean toward the side it started on.

Kind
Experimental result. What an experiment or trial found, under its own conditions. A result in animals, cells or a simulation is not a result in people.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

350,757 flips by 48 people. The bias varied between people and fell with practice.

Later work looked for

Crossref, 2018 on, "same-side bias coin flip": this 2025 study is the largest test found; nothing later.

Corrected

24 September 2026. Previously: “A coin toss cannot be predicted in practice.” A 2025 study of 350,757 flips found coins land on the side they started about 50.8 per cent of the time, as the physical model predicts.

Claim 20 · checked 24 September 2026 · AI-assisted audit

Claim: Ignorance of initial conditions.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

What the standard physical model of coin flipping says.

Claim 21 · checked 24 September 2026 · AI-assisted audit

Claim: The physics is deterministic and a sufficiently detailed measurement would predict it.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

That a detailed enough measurement would predict a flip follows from the model; the study itself did not predict individual flips.

Claim 22 · checked 24 September 2026 · AI-assisted audit

Claim: Which detector a photon triggers cannot be predicted.

Kind
Observation. Something measured or seen, as the named source reports it.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

Certified unpredictable under the experiment’s assumptions; whether it is undetermined in principle is the open question below.

Later work looked for

Crossref, 2018 on, "device-independent randomness generation": later photon experiments certify the same kind of outcome, records only; nothing that changes it.

Claim 23 · checked 24 September 2026 · AI-assisted audit

Claim: Either the same thing at a level nobody can see, or genuine indeterminacy.

Kind
Disputed. A contested point, where the page reports a side or the disagreement.
Review
A judgement, not a fact. Reviewed for whether it is fairly put and follows from the checked claims. No source can confirm it.

The two readings the poll offers as "hidden determinism" and "irreducible".

  • Context

    A snapshot of foundational attitudes toward quantum mechanicsQuestion 1 and commentary

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Research in people · Observational study: associations, not causes

    The source reads: hidden variables, deterministic mechanisms underlying the occurrence of objective quantum events, and (as in the de Broglie–Bohm interpretation) the deterministic motion of physical constituents.

  • Context

    A snapshot of foundational attitudes toward quantum mechanicsQuestion 1, commentary

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: In the Everett interpretation, randomness is an apparent effect relative to a branching observer

Other readings

A third reading, the Everett one the page describes below, makes the randomness apparent without hidden variables.

Claim 24 · checked 24 September 2026 · AI-assisted audit

Claim: Quantum mechanics as standardly used gives only a probability for each detector.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: the rules for connecting quantum theory with experimental results typically involve nontrivial (that is, not equal to zero or one) probabilities assigned to experimental outcomes.

What this does not establish

What the theory predicts, not what underlies it: interpretations differ on whether the probabilities reflect indeterminacy.

Corrected

24 September 2026. Previously: “Standard quantum mechanics says the latter.” Which of the two readings is right is an interpretive question the page itself treats as open. What the standard theory does say is that it assigns probabilities to outcomes.

Claim 25 · checked 24 September 2026 · AI-assisted audit

Claim: Outputs that violate a Bell inequality.

Kind
Observation. Something measured or seen, as the named source reports it.
Review
Source checked. The cited passage was read, and it says this.

Later work looked for

Crossref, 2018 on, "loophole-free Bell": violations continue to be reported; nothing that changes it.

Claim 26 · checked 24 September 2026 · AI-assisted audit

Claim: Not fixed in advance by anything local.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

Given independent settings and no signalling between the stations.

Claim 27 · checked 24 September 2026 · AI-assisted audit

Claim: Unpredictability that does not depend on trusting the device.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

The workings of the device are not trusted; its timing electronics, the computers and the independence of the settings still are.

Claim 28 · checked 24 September 2026 · AI-assisted audit

What is left assumed

Claim: What is left assumed

Kind
Framing. Signposting. It asserts nothing a source could check.

Claim: Unpredictable to us, or unpredictable to anything.

Kind
Framing. Signposting. It asserts nothing a source could check.

Claim: The experiments reach the first and press on the second.

Kind
This page’s reading. A judgement the page makes from the evidence. Others read it differently.
Review
A judgement, not a fact. Reviewed for whether it is fairly put and follows from the checked claims. No source can confirm it.

The page’s reading. The paper certifies bits as unpredictable to any adversary limited only by general physical principles, which goes beyond "unpredictable to us", but under assumptions, so it presses on "unpredictable to anything" without reaching it.

Claim 31 · checked 24 September 2026 · AI-assisted audit

Whether the universe is deterministic underneath

Claim: Whether the universe is deterministic underneath

Kind
Open question. Something the page says is not known.
Review
Presented as open. The page says this is unresolved, and the review found that it is.

  • Supports

    A snapshot of foundational attitudes toward quantum mechanicsabstract

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: the foundations of quantum mechanics themselves remain hotly debated in the scientific community, and no consensus on essential questions has been reached.

Claim 32 · checked 24 September 2026 · AI-assisted audit

Claim: Interpretations differ.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    A snapshot of foundational attitudes toward quantum mechanicsabstract

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Research in people · Observational study: associations, not causes

    The source reads: the foundations of quantum mechanics themselves remain hotly debated in the scientific community, and no consensus on essential questions has been reached.

Claim 33 · checked 24 September 2026 · AI-assisted audit

Claim: Some take the randomness as fundamental.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

33 respondents at one conference; the authors say the poll is not representative.

Claim 34 · checked 24 September 2026 · AI-assisted audit

Claim: Some describe a deterministic evolution in which all outcomes occur and the apparent randomness is about which branch an observer finds themselves in.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    A snapshot of foundational attitudes toward quantum mechanicsQuestion 1, commentary

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: In the Everett interpretation, randomness is an apparent effect relative to a branching observer, with the global wave function (containing all branches) evolving deterministically

  • Supports

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: when it is certain that all possible outcomes will occur on some branch of the wavefunction

Claim 35 · checked 24 September 2026 · AI-assisted audit

Claim: One keeps determinism by adding hidden variables and giving up locality.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    A snapshot of foundational attitudes toward quantum mechanicsQuestion 1, commentary

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: hidden variables, deterministic mechanisms underlying the occurrence of objective quantum events, and (as in the de Broglie–Bohm interpretation) the deterministic motion of physical constituents.

  • Supports

    Bell’s Theoremsection 1

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: in this theory an explicit causal mechanism exists whereby the disposition of one piece of apparatus affects the results obtained with a distant piece

Claim 36 · checked 24 September 2026 · AI-assisted audit

Claim: All of them agree with every experiment performed so far,

Kind
What the field holds. The settled view of a field, as a review, a consensus statement or a reference work gives it. Not one study.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: All of the above-mentioned approaches are consistent with observation. ... Extant hidden-variables theories reproduce the quantum probabilities, and collapse theories have the intriguing feature of reproducing very close approximations to quantum probabilities for all experiments that have been performed so far

  • Qualifies

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: A criticism that has been raised against Everettian theories is that it is not clear whether they can even make sense of statistical testing of this kind

What this does not establish

The encyclopedia notes a criticism that it is unclear how Everettian theories make sense of statistical testing at all.

Later work looked for

Crossref, 2020 on, "experimental test collapse models": experiments continue to bound collapse models (for example levitated optomechanics, New Journal of Physics 2024), records only; no departure reported.

Corrected

24 September 2026. Previously: “All of them reproduce the same statistics,” Collapse theories, one way of taking the randomness as fundamental, reproduce the quantum statistics only approximately and depart from them in experiments not yet performed. "The same statistics" held only for experiments done.

Claim 37 · checked 24 September 2026 · AI-assisted audit

Claim: so this is not a question the experiments above settle.

Kind
This page’s reading. A judgement the page makes from the evidence. Others read it differently.
Review
A judgement, not a fact. Reviewed for whether it is fairly put and follows from the checked claims. No source can confirm it.

Follows from claim 37: the Bell experiments do not distinguish the interpretations.

  • Context

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: among the major lines of approach, none is favored in a straightforward way by the empirical evidence.

Claim 38 · checked 24 September 2026 · AI-assisted audit

Claim: Collapse theories, one way of taking the randomness as fundamental, predict different statistics in experiments not yet performed.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: collapse theories have the intriguing feature of reproducing very close approximations to quantum probabilities for all experiments that have been performed so far but departing from the quantum probabilities for other conceivable experiments.

  • Supports

    Philosophical Issues in Quantum Theorysection 4.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: a research programme of finding suitable indeterministic modifications of the quantum dynamics

What this does not establish

The departures are predicted, not observed: collapse theories match every experiment performed so far.

Corrected

24 September 2026. Previously: “” Added so that the claim that the interpretations are empirically equivalent is not stretched to collapse theories, which the Stanford Encyclopedia says can in principle be told apart by experiment.

Claim 39 · checked 24 September 2026 · AI-assisted audit

What certification actually assumes

Claim: What certification actually assumes

Kind
Framing. Signposting. It asserts nothing a source could check.

Claim: It assumes that the experimenters’ choices of what to measure are independent of the system being measured.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

Claim 41 · checked 24 September 2026 · AI-assisted audit

Claim: A universe in which those choices were correlated with the particles could fake the result.

Kind
What a model says. What a theory or model predicts or describes, which is not the same as what has been observed.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Bell’s Theoremsection 8.1

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: Reject the measurement independence assumption. There are essentially two ways to do this. One is to suppose a common cause in the past that determines both experimental settings and experimental outcomes. ... This approach is often referred to as retrocausality. Both of these approaches have gained advocates in recent years.

What this does not establish

A logical possibility the argument cannot exclude, not a proposal the evidence favours.

Other readings

Superdeterminism and retrocausal accounts take this route, and the encyclopedia reports both have gained advocates.

Corrected

24 September 2026. Previously: “were correlated with the particles, however subtly, could fake the result.” No cited source says an arbitrarily small correlation suffices; the encyclopedia describes rejecting the independence assumption, not a threshold.

Claim 42 · checked 24 September 2026 · AI-assisted audit

Claim: That assumption cannot be tested from inside.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

The 2018 authors’ own description of their first assumption.

Claim 43 · checked 24 September 2026 · AI-assisted audit

Claim: It is the residue that every version of this argument leaves behind.

Kind
This page’s reading. A judgement the page makes from the evidence. Others read it differently.
Review
A judgement, not a fact. Reviewed for whether it is fairly put and follows from the checked claims. No source can confirm it.

The page’s reading. The encyclopedia names measurement independence as an assumption of the derivation, and quotes Bell saying he would not know how to formulate local causality without it.

  • Context

    Bell’s Theoremsection 8.1.2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026 · Reference work

    The source reads: we are invoking the Measurement Independence Assumption. ... Indeed without such freedom I would not know how to formulate any idea of local causality, even the modest human one

Claim 44 · checked 24 September 2026 · AI-assisted audit

Claim: The rest of what is trusted is more ordinary and is stated in the paper: the timing electronics that verify the separation, the finality of a result once recorded, the classical computers that process the data.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

  • Supports

    Experimentally generated randomness certified by the impossibility of superluminal signalsp. 2

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: given our trust that the space-like separation of the relevant events in the experiment is accurately verified by the timing electronics and that results are final when recorded. We also trust that the classical computing equipment used to process the data operates according to specification.

  • Supports

    Experimentally generated randomness certified by the impossibility of superluminal signalsp. 3

    Cited · Link answered, 19 October 2026 · Read in full, 24 September 2026

    The source reads: If Eve has quantum memory, it is possible to ensure that Eve’s side information is effectively classical by verifying that the devices have no long-term quantum memory of past interactions with Eve. While this introduces weak device-dependence

What this does not establish

The paper also assumes no signalling between the stations, which the page gives earlier, and, against an adversary with quantum memory, that the devices keep none.

Corrected

24 September 2026. Previously: “the clocks that establish the separation, the classical computers that process the data.” The paper names timing electronics rather than clocks, and also trusts that results are final when recorded, which the list left out.

Claim 45 · checked 24 September 2026 · AI-assisted audit

Claim: In the run that produced the 1,024 bits the settings came from a standard pseudorandom generator, which the authors call predictable and unsuitable against an adversary.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

What this does not establish

The earlier data set that yielded 256 bits did use physical random sources for its settings.

Claim 46 · checked 24 September 2026 · AI-assisted audit

Claim: They argue that their devices had no physical means of predicting it.

Kind
What scholars argue. A report of what specialists argue, or where they disagree.
Review
Source checked. The cited passage was read, and it says this.

Corrected

24 September 2026. Previously: “” The page reported the authors’ warning about the pseudorandom settings but not their reason for trusting this run, which is on the same page of Methods.

Claim 47 · checked 24 September 2026 · AI-assisted audit

What was read

  • Experimentally generated randomness certified by the impossibility of superluminal signalsBierhorst and others, Nature · arXiv:1803.06219v1 (22 February 2018), the submitted manuscript of Nature 556, 223-226; main text and Methods read in full, Supplementary Information not readRead in full, 24 September 2026 · Copyright; quoted briefly
  • A snapshot of foundational attitudes toward quantum mechanicsSchlosshauer, Kofler and Zeilinger, Studies in History and Philosophy of Modern Physics · arXiv:1301.1069v1 (6 January 2013); read in fullRead in full, 24 September 2026 · Copyright; quoted briefly
  • Fair coins tend to land on the same side they started: evidence from 350,757 flipsBartoš and others, Journal of the American Statistical Association · arXiv:2310.04153v4 (17 April 2025), the manuscript of the Journal of the American Statistical Association article (published August 2025); abstract, introduction, methods and results readRead in full, 24 September 2026 · Copyright; quoted briefly
  • The three-dimensional dynamics of the die throwKapitaniak, Strzalko, Grabski and Kapitaniak, Chaos · Crossref record and abstract, Chaos 22, 047504Abstract only, 24 September 2026 · Copyright; quoted brieflyNo open copy found through unpaywall; the article was not read.
  • Philosophical Issues in Quantum TheoryMyrvold, Stanford Encyclopedia of Philosophy · Stanford Encyclopedia of Philosophy, substantive revision of 23 March 2022; sections 4 and 5 readRead in full, 24 September 2026 · Copyright; quoted briefly
  • Bell’s TheoremMyrvold, Genovese and Shimony, Stanford Encyclopedia of Philosophy · Stanford Encyclopedia of Philosophy, substantive revision of 25 January 2024; sections 1, 2 and 8 readRead in full, 24 September 2026 · Copyright; quoted briefly