Nov 7, 2016 - Apparent randomness as an element of more efficient de- ..... course of events. Such a possibility is definitely indispensable component...

0 downloads 32 Views 948KB Size

arXiv:1611.02176v1 [quant-ph] 7 Nov 2016

1

ICFO-Institut de Ciències Fotòniques, The Barcelona Institute of Science and Technology, E-08860 Castelldefels (Barcelona), Spain 2 ICREA-Institució Catalana de Recerca i Estudis Avançats, Lluis Companys 23, E-08010 Barcelona, Spain 3 Center for Theoretical Physics, Polish Academy of Sciences, Aleja Lotników 32/44, 02-668 Warszawa, Poland This progress report covers recent developments in the area of quantum randomness, which is an extraordinarily interdisciplinary area that belongs not only to physics, but also to philosophy, mathematics, computer science, and technology. For this reason the article contains three parts that will be essentially devoted to different aspects of quantum randomness, and even directed, although not restricted, to various audiences: a philosophical part, a physical part, and a technological part. For these reasons the article is written on an elementary level, combining very elementary and non-technical descriptions with a concise review of more advanced results. In this way readers of various provenances will be able to gain while reading the article.

CONTENTS I. Introduction II. Quantum Randomness and Philosophy A. Epistemic and ontic character of probability B. Randomness in classical physics C. Randomness in quantum physics 1. Contextuality and randomness 2. Nonlocality and randomness

1 3 3 4 6 6 7

III. Quantum Randomness and Physics A. Quantum measurements 1. Postulates of QM 2. Measurement theory B. Nonlocality 1. Two-party nonlocality 2. Multi-party nonlocality and device independent approach C. Randomness: information theoretic approach D. Nonlocality, random number generation and certification E. Nonlocality and randomness expansion F. Nonlocality and randomness amplification

8 9 9 9 11 11 12 13 14 15 16

IV. Quantum randomness and technology

17

V. Quantum Randomness and Future VI. Acknowledgements References

20 20 20

I. INTRODUCTION

Randomness is a very important concept finding many applications in modern science and technologies. At the same time it is also quite controversial, and may have different

meanings depending on the field of science it concerns. In this short introduction to our report we explain, in a very general manner, why randomness plays so important role in various branches of science and technology. In particular we elaborate the concept of “apparent randomness”, to contrast it with what we understand under the name “intrinsic randomness”. Apparent randomness as an element of more efficient description of nature is used practically in all sciences, and in physics in particular (cf. (Halmos, 2013; Khinchin, 2014; Penrose, 1979; Schrödinger, 1989; Tolman, 2010)). This kind of randomness expresses our lack of full knowledge of the considered system. Paradigmatic example concerns classical mechanics of many-body systems that are simply too complex to be considered with all details. The complexity of the dynamics of systems consisting of many interacting constituents makes predictions, even assuming perfect knowledge of initial conditions, practically impossible. This fact motivates the development of statistical mechanics and thermodynamics. The descriptions that employ probability distributions and statistical ensembles, or even more reduced thermodynamic description, are more adequate and useful. Another paradigmatic example concerns chaotic systems. In deterministic chaos theory (cf. (Bricmont, 1995; Gleick, 2008; Ivancevic and Ivancevic, 2008)) even for the small systems involving few degrees of freedom, the immanent lack of precision in our knowledge of initial conditions leads to the impossibility of making long time predictions. This is due to an exponential separation of trajectories, when small differences at start lead to large endeffects. Also here, intrinsic ergodicity allows one to use the tools of statistical ensembles. In quantum mechanics apparent (a.k.a. epistemic) random-

2 ness also plays an important role and reflects our lack of full knowledge of the state of a system. This is the main reason of the introduction of the density matrix formalism (CohenTannoudji et al., 1991; Messiah, 2014). However, the quantum randomness in general has a very different and intrinsic or inherent nature. Namely, even if the state of the system is pure and we know it exactly, the predictions of quantum mechanics are intrinsically probabilistic and random! Accepting quantum mechanics means making an assumption that it is correct, and consequently intrinsically random. We adopt this position in this paper. To summarize the above discussion let us define: Def. 1 – Apparent (a.k.a. epistemic) randomness. Apparent randomness is the randomness that results exclusively from a lack of full knowledge about the state of the system in consideration. Had we known the initial state of the system exactly, we could have predicted its future evolution exactly. Probabilities and stochastic processes are used here as an efficient tool to describe at least a partial knowledge about the system and its features. Apparent randomness implies and requires existence of the, so called, underlying hidden variable theory. It is the lack of knowledge of hidden variables that causes apparent randomness. Had we known them, we could have make predictions with certainty. Def. 2 – Intrinsic (a.k.a. inherent or ontic) randomness. Intrinsic randomness is the randomness that persists even if we have the full knowledge about the state of the system in consideration. Even exact knowledge of the initial state does not allow to predict future evolution exactly: we can only make probabilistic predictions. Probabilities and stochastic processes are used here as a necessary and inevitable tool to describe our knowledge about the system and its behavior. Of course, intrinsic randomness might coexist with the apparent one – for instance, in quantum mechanics when we have only partial knowledge about the state of the system expressed by the density matrix, the two causes of randomness are present. Moreover, intrinsic randomness does not exclude existence of effective hidden variable theories that could allow for partial predictions of the evolution of the systems with certainty. As we shall see, in quantum mechanics of composite systems, an effective local hidden variable theories in general cannot be used to make predictions about local measurements and the local outcomes are intrinsically random. Having defined the main concepts, we present here short resumes of the subsequent parts of the report, where our focus would be mostly on quantum randomness: • Quantum Randomness and Philosophy. Inquiries concerning the nature of randomness accompany European philosophy from its beginnings. We give a short review of classical philosophical attitudes to the problem and their motivations. Our aim is to relate them to contemporary physics and science in general. This is

intimately connected to discussion of various concepts of determinism and its understanding in classical mechanics, commonly treated as an exemplary deterministic theory, where chance has only an epistemic status and leaves room for indeterminism only in form of statistical physics description of the world. In this context, we briefly discuss another kind of indeterminism in classical mechanics caused by the non-uniqueness of solutions of the Newton’s equations and requiring supplementing the theory with additional unknown laws. We argue that this situation shares similarities with that of quantum mechanics, where quantum measurement theory á la von Neumann provides such laws. This brings us to the heart of the problem of intrinsic randomness of quantum mechanics from the philosophical point of view. We discuss it in two quantum aspects: contextuality and nonlocality. • Quantum Randomness and Physics. Unlike in classical mechanics, randomness is considered to be inherent in the quantum domain. From a scientific point of view, we would rise arguments if this randomness is intrinsic or not. We start by briefly reviewing standard the approach to randomness in quantum theory. We shortly recall the postulates of quantum mechanics and the relation between quantum measurement theory and randomness. Nonlocality as an important ingredient of the contemporary physical approach to randomness is then discussed. We proceed then with more contemporary approach to randomness based on the so called “device independent” approach, in which one talks exclusively about correlations and probabilities to characterize randomness and nonlocality. We then describe several problems of classical computer science that have recently found elegant quantum mechanical solutions, employing the nonlocality of quantum mechanics. Before doing this we devote a subsection to describe a contemporary information theoretic approach to the characterization of randomness and random bit sources. In the following we discuss the idea of protocols for Bell certified randomness generation (i.e. protocols based on Bell inequalities to generate and certify randomness in the device independent scenario), such as quantum randomness expansion (i.e. generating larger number of random bits from a shorter seed of random bits), quantum randomness amplification (i.e. transforming weakly random sequences of, say, bits into “perfectly” random ones). It should be noted that certification, expansion and amplification of randomness are classically not possible or require extra assumptions in comparison with what quantum mechanics offers. Our goal is to review the recent state-of-art results in this area, and their relations and applications in the device independent quantum secret key distribution. We also review briefly and analyze critically the “special status” of quantum mechanics among the so called no-signaling theories.

3 In many situations, it is the no-signaling assumption and Bell non-locality that permit certification of randomness and perhaps novel possibilities of achieving security in communication protocols. • Quantum Randomness and Technology. In the field of technology we mention how random numbers are useful. The drawback of the classical random number generation, based on classical computer science ideas, is also mentioned. We describe proof-of-principle experiments in which certified randomness was generated using nonlocality. We then focus on describing existing “conventional” methods of quantum random number generation and certification. We discuss also the current status of detecting non-locality and Bell violations. We will then review current status of commercial implementation of quantum protocols for random number generations, and the first steps toward device independent or at least partially device independent implementations. • Quantum Randomness and Future. In the conclusions we outline new interesting perspectives for fundamentals of quantum mechanics that are stimulated by the studies of quantum randomness: What’s the relation between randomness, entanglement and non-locality? What are the ultimate limits for randomness generation using quantum resources? How does quantum physics compare to other theories in terms of randomness generation? What’s the maximum amount of randomness that can be generated in a theory restricted only by the no-signaling principle? We stress that there are of course various reviews on randomness in physics, such as for instance the excellent articles by J. Bricmont (Bricmont, 1995). The main novelty of our report lies in the incorporation of the contemporary approach to quantum randomness and its relation to quantum nonlocality and quantum correlations, and emerging device independent quantum information processing and quantum technologies.

II. QUANTUM RANDOMNESS AND PHILOSOPHY A. Epistemic and ontic character of probability

Randomness is a fundamental resource indispensable in numerous scientific and practical applications like Monte-Carlo simulations, taking opinion polls, cryptography etc. In each case one has to generate a “random sample”, or simply a random sequence of digits. Variety of methods to extract such a random sequence from physical phenomena were proposed and, in general successfully, applied in practice. But how do we know that a sequence is “truly random”? Or, at least. “random enough” for all practical purposes? Such problems become particularly acute for cryptography where provably unbreakable security systems are based on the possibility to

produce a string of perfectly random, uncorrelated digits used later to encode data. Such a random sequence must be unpredictable for an adversary wanting to break the code, and here we touch a fundamental question concerning the nature of randomness. If all physical processes are uniquely determined by their initial conditions and the only cause of unpredictability is our inability to determine them with an arbitrary precision, or lack of detailed knowledge of actual conditions that can influence their time evolution, the security can be compromised, if an adversary finds finer methods to predict outcomes. On the other hand, if there are processes “intrinsically” random, i.e. random by their nature and not due to gaps in our knowledge, unconditional secure coding schemes are conceivable. Two attitudes concerning the nature of randomness in the world mentioned above can be dubbed as epistemic and ontic. Both agree that we observe randomness (indeterminacy, unpredictability) in nature, but differ in identifying the source of the phenomenon. The first claims that the world is basically deterministic, and the only way, in which random behavior demanding probabilistic description appears, is due to lack of knowledge of the actual state of the observed system or details of its interaction with the rest of the universe. In contrast, according to the second, the world is nondeterministic, randomness is its intrinsic property, independent of our knowledge and resistant to attempts aiming at circumventing its consequences by improving precision of our observations. In other words, “intrinsic” means that this kind of randomness cannot be understood in terms of a deterministic “hidden variable” model. The debate on both epistemic and ontic nature of randomness can be traced back to the pre-Socratic beginnings of the European philosophy. For early atomists, Leucippus1 and Democritus2 , the world was perfectly deterministic. Any occurrence of chance is a consequence of our limited abilities3 . One century later Epicurus took the opposite side. To accommodate an objective chance the deterministic motion of atoms must be interrupted, without a cause, by “swerves”. Such an indeterminacy propagates then to macroscopic world. The main motivation was to explain, or at least to give room for human free will, hardly imaginable in a perfectly deterministic world4 . It should be clear, however, that purely random nature of human actions is as far from free will, as the latter from a completely deterministic process. A common feature of both extreme cases of pure randomness and strict determinism is lack of any possibility to control or influence the

1

2 3 4

‘Nothing happens at random; everything happens out of reason and by necessity’, from the lost work Perí no˜u On Mind, see (Diels, 1906), p. 350, (Freeman, 1948), p. 140, fr. 2. ‘All things happen by virtue of necessity’, (Laertius, 1925), IX, 45. ‘Men have fashioned an image of chance as an excuse for their own stupidity’, (Diels, 1906), p. 407, (Freeman, 1948), p. 158, fr. 119. ‘Epicurus saw that if the atoms traveled downwards by their own weight, we should have no freedom of the will, since the motion of the atoms would be determined by necessity. He therefore invented a device to escape from determinism (the point had apparently escaped the notice of Democritus): he said that the atom while traveling vertically downward by the force of gravity makes a very slight swerve to one side’ (Cicero, 1933), I, XXV.

4 course of events. Such a possibility is definitely indispensable component of the free will. The ontological status of randomness is thus here irrelevant and the discussion whether "truly random theories", (as supposedly should quantum mechanics be), can "explain the phenomenon of the free will" is pointless. It does not mean that free will and intrinsic randomness problems are not intertwined. On one side, as we explain later, the assumption that we may perform experiments in which we can freely choose what we measure, is an important ingredient in arguing that violating of Bell-like inequalities implies “intrinsic randomness” of quantum mechanics. On the second side, as strict determinism in fact precludes the free will, the intrinsic randomness seems to be a necessary condition for its existence. But, we need more to produce a condition that is sufficient. An interesting recent discussion of connections between free will and quantum mechanics may be found in Part I of (Suarez and Adams, 2013). In (Gisin, 2013) and (Brassard and Robichaud, 2013) the many-world interpretation of quantum mechanics, which is sometimes treated as a cure against odds of orthodox quantum mechanics, is either dismissed as a theory that can accommodate free will (Gisin, 2013) or, taken seriously in (Brassard and Robichaud, 2013), as admitting the possibility that free will might be a mere illusion. In any case it is clear that one needs much more than any kind of randomness to understand how free will appears. In (Suarez, 2013) the most radical attitude to the problem (apparently present also in (Gisin, 2013)) is that “not all that matters for physical phenomena is contained in space-time”. B. Randomness in classical physics

A seemingly clear distinction between two possible sources of randomness outlined in the previous section becomes less obvious if we try to make the notion of determinism more precise. Historically, its definition usually had a strong epistemic flavor. Probably the most famous characterization of determinism that of Pierre Simon de Laplace (Laplace, 1814): ‘Une intelligence qui, pour un instant donné, connaîtrait toutes les forces dont la nature est animée, et la situation respective des êtres qui la composent, si d’ailleurs elle était assez vaste pour soumettre ces données à l’analyse, embrasserait dans la même formule les mouvemens des plus grands corps de l’univers et ceux du plus léger atome : rien ne serait incertain pour elle, et l’avenir comme le passé, serait présent á ses yeux.5 Two hundred years later Karl Raimund Popper writes ‘We can ... define ‘scientific’ determinism as

follows: The doctrine of ‘scientific’ determinism is the doctrine that the state of any closed physical system at any given future instant of time can be predicted, even from within the system, with any specified degree of precision, by deducing the prediction from theories, in conjunction with initial conditions whose required degree of precision can always be calculated (in accordance with the principle of accountability) if the prediction task is given’ (Popper, 1982). By contraposition thus, unpredictability implies indeterminism. If we now equate indeterminism with existence of randomness, we see that a sufficient condition for the latter is the unpredictability. But, unpredictable can be equally well events, about which we do not have enough information, and those that are “random by themselves”. Consequently, as it should have been obvious, Laplacean-like descriptions of determinism are of no help, when we look for sources of randomness. Let us thus simply say that a course of events is deterministic if there is only one future way for it to develop. Usually we may also assume that its past history is also unique. In such cases the only kind of randomness is the epistemic one. As an exemplary theory describing such situations one usually invokes classical mechanics. Arnol’d in his treatise on ordinary differential equations, after adopting the very definition of determinism advocated above6 , writes: “Thus for example, classical mechanics considers the motion of systems whose past and future are uniquely determined by the initial positions and velocities of all points of the system”7 . The same can be found in his treatise on classical mechanics8 . He gives also a kind of justification, “It is hard to doubt this fact, since we learn it very early”9 . But, what he really means is that a mechanical system are uniquely determined by positions and momenta of its constituents, “one can imagine a world, in which to determine the future of a system one must also know the acceleration at the initial moment, but experience shows us that our world is not like this”10 . It is clearly exposed in another classical mechanics textbook, Landau and Lifschitz’s Mechanics: “If all the co-ordinates and velocities are simultaneously specified, it is known from experience that the state of the system is completely determined and that its subsequent motion can, in principle, be calculated. Mathematically, this means that, if all the co-ordinates q and velocities q˙ are given at some instant, the accelerations q¨ at that instant are uniquely defined”11 . Apparently, also here the “experience” concerns only the observation that positions and velocities, and not higher time-derivatives of them, are sufficient to

6 5

“We may regard the present state of the universe as the effect of its past and the cause of its future. An intellect which at a certain moment would know all forces that set nature in motion, and all positions of all items of which nature is composed, if this intellect were also vast enough to submit these data to analysis, it would embrace in a single formula the movements of the greatest bodies of the universe and those of the tiniest atom; for such an intellect nothing would be uncertain and the future just like the past would be present before its eyes.” (Laplace, 1951) p. 4

7 8

9 10 11

“A process is said to be deterministic if its entire future course and its entire past are uniquely determined by its state at the present instant of time”, (Arnol’d, 1973), p. 1 ibid. ”The initial state of a mechanical system (the totality of positions and velocities of its points at some moment of time) uniquely determines all of its motion”, (Arnol’d, 1989), p. 4 ibid ibid. (Landau and Lifshitz, 1960), p. 1.

5 determine the future. In such a theory there are no random processes. Everything is in fact completely determined and can be predicted with desired accuracy once we improve our measuring and computing devices. Statistical physics, which is based on classical mechanics, is a perfect example of indeterministic theory where measurable quantities like pressure or temperature are determined by mean values of microscopical ‘hidden variables’, for example positions and momenta of gas particles. These hidden variables, however, are completely determined at each instant of time by the laws of classical mechanics, and with an appropriate effort can be, in principle, measured and determined. What makes the theory ‘indeterministic’ is a practical impossibility to follow trajectories of individual particles because of their number and/or the sensitiveness to changes of initial conditions. In fact such a sensitiveness was pointed as a source of chance by Poincaré12 and Smoluchowski13 soon after modern statistical physics was born, but it is hard to argue that this gives to the chance an ontological status. It is, however, worth mentioning that Poincaré was aware that randomness might have not only epistemic character. In the above cited Introduction to his Calcul des probabilités he states ‘Il faut donc bien que le hasard soit autre chose que le nom que nous donnons à notre ignorance’14 , (‘So it must be well that chance is something other than the name we give our ignorance’15 ). It is commonly believed (and consistent with the above cited descriptions of determinism in mechanical systems) that on the mathematical level the deterministic character of classical mechanics takes form of Newton’s Second Law m

d2 x(t) = F(x(t), t), dt2

(1)

where the second derivatives of the positions, x(t), are given in terms of some (known) forces F(x(t), t). But, to be able to determine uniquely the fate of the system we need something more than merely initial positions x(0) and velocities dx(t)/dt|t=0 . To guarantee uniqueness of the solutions of the Newton’s equations (1), we need some additional assumptions about the forces F(x(t), t). According to the Picard Theorem, an additional technical condition, sufficient for the

12

13

14 15

“Le premier exemple que nous allons choisir est celui de l’équilibre instable; si un cône repose sur sa pointe, nous savons bien qu’il va tomber, mais nous ne savons pas de quel côté; il nous semble que le hasard seul va en décider.”, (Poincaré, 1912) page 4, (“The first example we select is that of unstable equilibrium; if a cone rests upon its apex, we know well that it will fall, but we do not know toward what side; it seems to us chance alone will decide.” (Newman, 1956), vol. 2, p. 1382) “...ein ganz wesentliches Merkmal desjenigen, was man im gewöhnlichen Leben oder in unserer Wissenschaft als Zufall bezeichnet ... läßt sich ... kurz in die Worte fassen: kleine Ursache – große Wirkung”, (Smoluchowski, 1918) (“...fundamental feature of what one calls chance in everyday life or in science allows a short formulation: small cause – big effect”) (Poincaré, 1912) p. 3. (Newman, 1956), vol. 2, p. 1381

uniqueness, is the Lipschitz condition, limiting the variability of the forces with respect to the positions. Breaking it opens possibilities to have initial positions and velocities that do not determine uniquely the future trajectory. A world, in which there are systems governed by equations admitting nonunique solutions is not deterministic according to our definition. We can either defend determinism in classical mechanics by showing that such pathologies never occur in our world, or agree that classical mechanics admits, at least in some cases, a nondeterministic evolution. Each choice is hard to defend. In fact it is relatively easy to construct examples of more or less realistic mechanical systems for which the uniqueness is not guaranteed. Norton (Norton, 2007) (see also (Norton, 2008)) provided a model of a point particle sliding on a curved surface under the gravitational force, for which the Newton √ 2 equation reduces to ddt2r = r. For the initial conditions r(0) = 0, dr dt |t=0 = 0, the equation has not only an obvious solution r(t) = 0, but, in addition, a one parameter family given by r(t) =

1 144 (t

0, for t ≤ T − T )4 , for t ≥ T

(2)

where T is an arbitrary parameter. For a given T the solution describes the particle staying at rest at r = 0 until T and starting to accelerate at T . Since T is arbitrary we can not predict when the change from the state of rest to the one with a non-zero velocity takes place. The example triggered discussions (Fletcher, 2012; Korolev, 2007a,b; Kosyakov, 2008; Laraudogoitia, 2013; Malament, 2008; Roberts, 2009; Wilson, 2009; Zinkernagel, 2010), raising several problems, in particular its physical relevance in connection with simplifications and idealizations made to construct it. However, they do not seem to be different from ones commonly adopted in descriptions of similar mechanical situations, where the answers given by classical mechanics are treated as perfectly adequate. At this point classical mechanics is not a complete theory of the part of the physical reality it aspires to describe. We are confronted with a necessity to supplement it by additional laws dealing with situations where the Newton’s equation do not posses unique solutions. The explicit assumption of incompleteness of classical mechanics has its history, astonishingly longer than one would expect. Possible consequences of non-uniqueness of solutions attracted attention of Boussinesq who in his Mémoire for the French Academy of Moral and Political Sciences writes: ‘...les phénomènes de mouvement doivent se diviser en deux grandes classes. La première comprendra ceux où les lois mécaniques exprimées par les équations différentielles détermineront à elles seules la suite des états par lesquels passera le système, et où, par conséquent, les forces physico-chimiques ne laisseront aucun rôle disponible à des causes d’une autre nature. Dans la seconde classe se rangeront, au contraire, les mouvements dont les équations admettront des intégrales singulières, et dans lesquels il faudra qu’une cause distincte des forces physico-chimiques intervienne, de temps en temps ou

6 d’une manière continue, sans d’ailleurs apporter aucune part d’action mécanique, mais simplement pour diriger le système a chaque bifurcation d’intégrales qui se présentera.’16 Boussinesq does not introduce any probabilistic ingredient to the reasoning, but definitely, there is a room to go from mere indeterminism to the awaited ‘intrinsic randomness’. To this end, however, we need to postulate an additional law supplementing classical mechanics by attributing probabilities to different solutions of non-Lipschitzian equations17 . It is hard to see how to discover (or even look for) such a law, and how to check its validity. What we find interesting is an explicit introduction to the theory a second kind of motion. It is strikingly similar to what we encounter in quantum mechanics, where to explain all observed phenomena one has to introduce two kinds of kinematics of a perfectly deterministic Schrödinger evolution and indeterministic state reductions during measurements. Similarity consist in the fact, that deterministic (Schrödinger, Newton) equations are not sufficient to describe the full evolution: they have to be completed, for instance by probabilistic description of the results of measurements in quantum mechanics, or by probabilistic choice of non-unique solutions in the Norton’s example. 18 C. Randomness in quantum physics

The chances of proving the existence of ‘intrinsic randomness’ in the world seem to be much higher, when we switch to quantum mechanics. The Born’s interpretation of the wave function implies that we can count only on a probabilistic description of reality, therefore quantum mechanics is inherently probabilistic. It implies that the measurement outcomes (or expectation value of an observable) may have some randomness. However, a priori there are no obvious reasons for leaving the Democritean ground and switch to the Epicurean view. It might be so that quantum mechanics, just as statistical physics, is an incomplete theory admitting deterministic

16

17

18

(Boussinesq, 1878), p. 39. “The movement phenomena should be divided into two major classes. The first one comprises those for which the laws of mechanics expressed as differential equations will determine by themselves the sequence of states through which the system will go and, consequently, the physico-chemical forces will not admit causes of different nature to play a role. On the other hand, to the second class we will assign movements for which the equations will admit singular integrals, and for which one will need a cause distinct from physico-chemical forces to intervene, from time to time, or continuously, without using any mechanical action, but simply to direct the system at each bifurcation point which will appear.” The “singular integrals” mentioned by Boussinesq are the additional trajectories coexisting with “regular” ones when conditions guaranteing uniqueness of solutions are broken. Thus in the Norton’s model, the new law of nature should, in particular, ascribe a probability p(T ) to the event that the point staying at rest at r = 0 starts to move at time T . Similar things seem to happen also in so called "general non-signaling theories" where, in comparison with quantum mechanics, the only physical assumption concerning the behavior of a system is the impossibility of transmitting information with an infinite velocity, see (Tylec and Ku´s, 2015)

hidden variables, values of which were beyond our control. To be precise, one may ask how “intrinsic” this randomness is and if it can be considered as an epistemic one. To illustrate it further, we consider two different examples in the following. 1. Contextuality and randomness

Let us consider a case of a spin-s particle. Now if the particle is measured in the z-direction, there could be 2s + 1 possible outcomes and each appears with certain probability. Say, the outcomes are labeled by {m}, where m ∈ [−s, −s + 1, . . . , s − 1, s] and the corresponding probabilities by {pm }. It means that, with many repetitions, the experimenter will observe an outcome m with the frequency approaching pm , as predicted by the Born’s rule of quantum mechanics. The outcomes contain some randomness as they appear probabilistically. Moreover, these probabilities are indistinguishable from classical probabilities. Therefore, the randomness here could be explained with the help of a deterministic hidden-variable model19 and it is simply a consequence of the ignorance of the hidden-variable(s). But, as we stress in the definition in the Introduction: intrinsic randomness of quantum mechanics does not exclude existence of hidden variable models that can describe, at least partially, outcomes of measurements. Obviously, if the system is in the pure state corresponding to m0 , the outcome of the measurement of z-component of the spin will be deterministic: m0 with certainty. If we measured x-component of the spin, however, the result would be again non-deterministic and described only probabilistically. In fact, this is an instance of the existence of the, so called, non-commuting observables in quantum mechanics that cannot be measured simultaneously with certainty. Uncertainty of measurements of such non-commuting observables is quantitatively bounded from below by generalized Heisenberg Uncertainty Principle (Cohen-Tannoudji et al., 1991; Messiah, 2014). Perhaps the most important consequence of the existence of non-commuting observables is the fact that quantum mechanics is contextual, as demonstrated in the famous KochenSpecker theorem ((Kochen and Specker, 1967), for philosophical discussion see (Bub, 1999; Isham and Butterfield, 1998)). The Kochen–Specker (KS) theorem (Kochen and Specker, 1967), also known as the Bell-Kochen–Specker theorem (Bell, 1966), is a "no go" theorem (Bub, 1999), proved by J.S. Bell in 1966 and by S.B. Kochen and E. Specker in 1967. KS theorem places certain constraints on the permissible types of hidden variable theories, which try to explain the randomness of quantum mechanics as an apparent randomness, resulting from lack of knowledge of hidden variables in

19

Note, here we do not impose any constraint on the hidden variables and these could be even nonlocal. In fact, the quantum theory becomes deterministic if one assumes the hidden variables to be nonlocal (Gudder, 1970).

7 an underlying deterministic model. The version of the theorem proved by Kochen and Specker also gave an explicit example for this constraint in terms of a finite number of state vectors (cf. (Peres, 1995)). The KS theorem deals with single quantum systems and is thus a complement to Bell’s theorem that deals with composite systems. Quoting Wikipedia: "The KS theorem proves that there is a contradiction between two basic assumptions of the hidden variable theories intended to reproduce the results of quantum mechanics: that all hidden variables corresponding to quantum mechanical observables have definite values at any given time, and that the values of those variables are intrinsic and independent of the device used to measure them. The contradiction is caused by the fact that quantum mechanical observables need not be commutative. It turns out to be impossible to simultaneously embed all the commuting subalgebras of the algebra of these observables in one commutative algebra, assumed to represent the classical structure of the hidden variable theory, if the Hilbert space dimension is at least three.20 The Kochen–Specker theorem excludes hidden variable theories that require elements of physical reality to be noncontextual (i.e. independent of the measurement arrangement). As succinctly worded by Isham and Butterfield (Isham and Butterfield, 1998), the Kochen–Specker theorem "asserts the impossibility of assigning values to all physical quantities whilst, at the same time, preserving the functional relations between them." In a more recent approach to contextuality, one proves that non-contextual hidden variable theories lead to probabilities of measurement outcomes that fulfill certain inequalities (Cabello, 2008), similar to Bell’s inequalities for composite systems. More specifically there are Bell-type inequalities for non-contextual theories that are violated by any quantum state. Many of these inequalities between the correlations of compatible measurements are particularly suitable for testing this state-independent violation in an experiment, and indeed violations have been experimentally demonstrated (Bartosik et al., 2009; Kirchmair et al., 2009). Quantifying and characterizing contextuality of different physical theories is particularly elegant in a general graph-theoretic framework (Acín et al., 2015; Acín et al., 2015; Cabello et al., 2014).

20

In fact, it was A. Gleason (Gleason, 1975), who pointed out first that quantum contextuality may exist in dimensions greater than two. It was earlier pointed out by N. Bohr (Bohr, 1935) that EPR-like paradoxes may occur in the quantum systems without the need for an entangled or composite systems. For qubits, i.e. for the especially simple case of two dimensional Hilbert space, one can explicitly construct the non-contextual hidden variable models that describes all measurements (cf. (Scully and Wódkiewicz, 1989; Wódkiewicz, 1995; Wódkiewicz, 1985)). As pointed out in the definition in the Introduction, this does not mean that quantum mechanics of qubits is not intrinsically random – for consistency we consider here all quantum mechanics, including that of qubits, to exhibit intrinsic randomness.

x

y

Alice

Bob

a

b

FIG. 1 Schematic of a two-party Bell-like experiment. The experimenters Alice and Bob are separated and cannot communicate as indicated by the black barrier. The measurements settings and outcomes, of Alice and Bob, are denoted by x, y and a, b respectively. 2. Nonlocality and randomness

It is important to extend the situation beyond the one mentioned above to multi-party systems and local measurements. For example, consider multi-particle system with each particle placed in a separated region. Now, instead of observing the system as a whole, one may get interested to observe only a part of it, i.e. perform local measurements. Given two important facts that QM allows superposition and no quantum system can be absolutely isolated, spatially separated quantum systems can be non-trivially correlated, beyond classical correlations allowed by classical mechanics. In such situation, the information contained in the whole system is certainly larger than that of sum of individual local systems. The information residing in the nonlocal correlations cannot be accessed by observing locally individual particles of the systems. It means that local descriptions cannot completely describe the total state of the system. Therefore, outcomes due to any local observation are bound to incorporate a randomness in the presence of nonlocal correlation, as long as we do not have access to the global system or ignore the nonlocal correlation. In technical terms, the randomness appearing in the local measurement outcomes cannot be understood in terms of deterministic local hidden variable model and a “true” local indeterminacy is present21 . Moreover, randomness on the local level appears even if the global state of the system is pure and completely known – the necessary condition for this is just entanglement of the pure state in question. That is typically referred as “intrinsic” randomness in the literature, and that is the point of view we adopt in this report. Before we move further in discussing quantum randomness in the presence of quantum correlation, let us make a short detour through the history of foundation of quantum mechanics. The possibility of nonlocal correlation, also known as quantum entanglement, was initially put forward with the question if quantum mechanics respects local realism, by Einstein, Podolsky and Rosen (EPR) (Einstein et al., 1935). According

21

Of course one could argue that such randomness appears only to be “intrinsic”, since it is essentially epistemic in nature and arises due to the inaccessibility or ignorance of the information that resides in the nonlocal correlations. In another words, this kind of randomness on the local level is caused by our lack of knowledge of the global state, and further, it can be explained using deterministic nonlocal hidden variable models.

8 to EPR, two main properties any reasonable physical theory should satisfy are realism and locality. The first one states that if a measurement outcome of a physical quantity, pertaining to some system, is predicted with unit probability, then there must exits ‘an element of reality” correspond to this physical quantity having a value equal to the predicted one, at the moment of measurement. In other words, the values of observables, revealed in measurements, are intrinsic properties of the measured system. The second one, locality, demands that elements of reality pertaining to one system cannot be affected by measurement performed on another sufficiently far system. Based on these two essential ingredients, EPR studied the measurement correlations between two entangled particles and concluded that wave function describing the quantum state “does not provide a complete description of physical reality”. Thereby they argued that quantum mechanics is an incomplete but, effective theory and conjectured that a complete theory describing the physical reality is possible. In these discussions, one needs to clearly understand what locality and realism mean. In fact, they could be replaced with no-signaling and determinism, respectively. The no-signaling principle states that infinitely fast communication is impossible. The relativistic limitation of the speed, by the velocity of light, is just a special case of no-signaling principle. If two observers are separated and perform space-like separated measurements (as depicted in Fig. 1), then the principle ascertains that the statistics seen by one observer, when measuring her particle is completely independent of the choice of measurement performed on the space-like separated other particle. Clearly, if it were not the case, one observer could, by changing her measurement choice, make a noticeable change on the other and thereby instantaneously communicate with an infinite speed. Determinism, the other important ingredient, implies that correlations observed in an experiment can be decomposed as mixtures of deterministic ones i.e., occurring in situations where all measurements have deterministic outcomes. A deterministic theory accepts the fact that the apparent random outcomes in an experiment, like in coin tossing, are only consequences of ignorance of the actual state of the system. Therefore, each run of the experiment does have an a priori definite result, but we have only an access to averages. In 1964, Bell showed that all theories that satisfy locality and realism (in the sense of EPR) are incompatible with quantum mechanics (Bell, 1964, 1966). In a simple experiment, mimicking Bell’s scenario, two correlated quantum particles are sent to two spatially separated measuring devices (see Fig. 1), and each device can perform two different measurements with two possible outcomes. The measurement processes are space-like separated and no communication is possible when these are performed. With this configuration a localrealistic model gives bounds on the correlation between the outcomes observed in the two measurement devices, called Bell-inequalities (Bell, 1964). In other words, impossibility of instantaneous communication (no-signaling) between spatially separated systems together with full local determinism

imply that all correlations between measurement results must obey the Bell inequalities. Strikingly, these inequalities are violated with correlated (entangled) quantum particles, and therefore have no explanations in terms of deterministic local hidden variables. In fact, the correlations predicted by the no-signaling and determinism are exactly the same as predicted by EPR model, and they are equivalent. The experimental violations of the Bell-inequalities in 1972 (Freedman and Clauser, 1972), in 1981 (Aspect et al., 1981) and in 1982 (Aspect et al., 1982), along with the recent loop-hole-free Bell-inequalities violations (Giustina et al., 2015; Hensen et al., 2015; Shalm et al., 2015) confirm that any local-realistic theory is unable to predict the correlations observed in quantum mechanics. It immediately implies that either no-signaling or local determinism (or even both!) has to be abandoned. For the most physicists, it is favorable to dump local determinism and save no-signaling. Assuming that the nature respects no-signaling principle, any violation of Bell-inequality implies thus that the outcomes could not be predetermined in advance. Thus, once the no-signaling principle is accepted to be true, the experimental outcomes, due to local measurements, cannot be deterministic and therefore are random. Of course, a valid alternative is to abandon the no-signaling principle and save local determinism, as for instance done in Bohm’s theory (Bohm, 1951). Another crucial assumption is considered for Bell experiments, that is the measurements performed with the local measurement devices have to be chosen “freely”. In other words, the measurement choices cannot, in principle, be predicted in advance. If the free-choice condition is relaxed and the chosen measurements could be predicted in advance, then it is easy to construct a no-signaling, but deterministic theory that leads to Bell-violations. It has been shown in (Hall, 2010; Koh et al., 2012) that one does not have to give up measurement independence completely to violate Bell-inequalities. Even, relaxing free-choice condition to a certain degree, the Bell-inequities could be maximally violated using no-signaling and deterministic model (Hall, 2010). However, in the Bell-like experiment scenarios where the local observers are space-like separated, it is very natural to assume that the choices of the experiments are completely free (this is often referred to free-will assumption). Therefore, the Bell-inequalities violation in the quantum regime, with the no-signaling principle, implies that local measurement outcomes are “intrinsically” random.

III. QUANTUM RANDOMNESS AND PHYSICS

In this section we consider randomness form the point of view of physics or in particular, quantum physics. In doing so, first we briefly introduce quantum measurements, nonlocality and information theoretic measures of randomness. Then we turn to outline, how the quantum feature such as nonlocality can be exploited not only to generate “true” randomness but also to certify, expand and amplify randomness.

9 A. Quantum measurements

According to standard textbook approach, quantum mechanics (QM) is an inherently probabilistic theory (cf. (Cohen-Tannoudji et al., 1991; Messiah, 2014; Wheeler and Zurek, 1983)– the prediction of QM concerning results of measurements are typically probabilistic. Only in very rare instances measurements give deterministic outcomes – this happens when the systems is in an eigenstate of an observable to be measured. Note, that in general, even if we have full information about the quantum mechanical state of the system, the outcome of the measurements is in principle random. The paradigmatic example is provided a d-state system (a qudit), whose space of states is spanned by the states|1i, |2i,..., |di. Suppose that we know the system is in the superposition state Pd |φi = j=1 αj |ji, where αj are complex probability amPd plitudes and j=1 |αj |2 = 1, and we ask whether it is in a state |ii. To find out, we measure an observable Pˆ = |iihi| that projects on the state |ii. The result of such measurement will be one (yes, the system is in the state |ii) with probability Pd |αi |2 and zero with probability 1 − j6=i |αj |2 . We do not want to enter here deeply into the subject of the foundations of QM, but we want to remind the readers the "standard" approach to QM.

1. Postulates of QM

The postulates of QM for simple mechanical systems (single or many particle), as given in (Cohen-Tannoudji et al., 1991), read: • First Postulate. At a fixed time t0 , the state of a physical system is defined by specifying a wave function ψ(x; t0 ), where x represents collection of parameters to specify the state. • Second Postulate. Every measurable physical quantity ˆ this operator is called Q is described by an operator Q; an observable. • Third Postulate. The only possible result of the measurement of a physical quantity Q is one of the eigenˆ values of the corresponding observable Q. • Fourth Postulate (non-degenerate case). When the physical quantity Q is measured on a system in the normalized state ψ, the probability P (qn ) of obtaining the non-degenerate eigenvalue qn of the corresponding obˆ is servable Q Z P (qn ) = dx ϕn (x)ψ(x), ˆ associated where ϕn is the normalized eigenvector of Q with the eigenvalue qn .

• Fifth Postulate (collapse). If the measurement of the physical quantity Q on the system in the state ψ gives the result qn , the state of the system immediately after the measurement is ϕn . • Sixth Postulate (time evolution). The time evolution of the wave function ψ(x; t) is governed by the Schrödinger equation i~

∂ψ ˆ = Hψ, ∂t

ˆ is the observable associated with the total enwhere H ergy of the system. • Seventh Postulate (symmetrization). When a system includes several identical particles, only certain wave functions can describe its physical states (leads to the concept of bosons and fermions). For electrons (which are fermions), the wave function must change sign whenever the coordinates of two electrons are interchanged. For hydrogen atoms (regarded as composite bosons) the wave function must not change whenever the coordinates of two bosons are interchanged. 2. Measurement theory

Evidently, the inherent randomness of QM is associated with the measurement processes (Fourth and Fifth Postulates). The quantum measurement theory has been a subject of intensive studies and long debate, see e.g., (Wheeler and Zurek, 1983). In particular the question of the physical meaning of the wave function collapse has been partially solved only in the last 30 years by analyzing interactions of the measured system with the environment (reservoir), describing the measuring apparatus (see seminal works of Zurek (Zurek, 2003, 2009)) In the abstract formulation in the early days of QM, one has considered von Neumann measurements (Neumann, 1955), ˆ has (possidefined in the following way. Let the observable Q ˆ bly degenerated) eigenvalues qn and let En denote projectors on the corresponding invariant subspaces (one dimensional for non-degenerate eigenvalues, k-dimensional for k-fold degenerated eigenvalues). Since the invariant subspace are orthogˆn E ˆm = δnm E ˆn , where δmn is the Kronecker onal, we have E delta. If Pˆψ denote the projector, which describes a state of a systems, the measurement outcome correspond to the eigenvalue qn of the observable will appear with probability pn = ˆn ), where Tr(.) denotes the matrix trace operation. Tr(Pˆψ E Also, after the measurement, the systems is found in the state ˆn Pˆψ E ˆn /pn with probability pn . E In the contemporary quantum measurement theory the measurements are generalized beyond the von Neumann projective ones. To define the, so called, positive-operator valued measures (POVM), one still considers von Neumann measurements, but on a system plus a additional ancilla system (Peres,

10 1995). In the simplest case, POVMs are defined by a set of Hermitian positive semidefinite operators {Fi } on a Hilbert space H that sum to the identity operator, K X

Fi = I H .

i=1

This is a generalization of the decomposition of a (finitedimensional) Hilbert space by a set of orthogonal projectors, {Ei }, defined for an orthogonal basis {|φi i} by Ei = |φi i hφi | , hence, N X

Ei = IH ,

Ei Ej = δij Ei

i=1

An important difference is that the elements of POVM are not necessarily orthogonal, with the consequence that the number K of elements in the POVM can be larger than the dimension N of the Hilbert space they act in. The post-measurement state depends on the way the system plus ancilla are measured. For instance, consider the case where the ancilla is initially a pure state |0iB . We entangle the ancilla with the system, taking |ψiA |0iB →

X

Mi |ψiA |iiB ,

i

and perform a projective measurement on the ancilla in the {|iiB } basis. The operators of the resulting POVM are given by Fi = Mi† Mi . Since the Mi are not required to be positive, there are an infinite number of solutions to this equation. This means that there are infinitely many different experimental apparatus giving the same probabilities for the outcomes. Since the postmeasurement state of the system (expressed now as a density matrix)

ρi =

Mi ρMi† tr(Mi ρMi† )

depends on the Mi , in general it cannot be inferred from the POVM alone. If we accept the quantum mechanics and its inherent randomness, then it is possible in principle to implement measurements of an observable on copies of a state that is not an eigenstate of this observable, to generate a set of perfect

random numbers. Early experiments and commercial devices attempted to mimic a perfect coin with probability 1/2 of getting head and tail. To this aim quantum two-level systems were used, for instance single photons of two orthogonal circular polarizations. If such photons are transmitted through a linear polarizer of arbitrary direction then they pass (do not pass) with probability 1/2. In practice, the generated numbers are never perfect, and randomness extraction is required to generate good random output. The challenges of sufficiently approximating the ideal two-level scenario, and the complexity of detectors for single quantum systems, have motivated the development of other randomness generation strategies. In particular, continuous-varible techniques are now several orders of magnitude faster, and allow for randomness extraction based on known predictability bounds. See Section IV. It is worth mentioning that the Heisenberg uncertainty relation (Heisenberg, 1927) and its generalized version, i.e., the Robertson-Scrödinger relation (Robertson, 1929; Schrödinger, 1930; Wheeler and Zurek, 1983), are often mentioned in the context of quantum measurements, signify how precisely two non-commuting observables can be measured on a quantum state. Quantitatively, for a given state ρ and observables X and Y , it gives a lower bound on the uncertainty when they are measures simultaneously, as δX 2 δY 2 ≥

1 |Trρ [X, Y ] |2 , 4

(3)

where δX 2 = TrρX 2 −(TrρX)2 is the variance and [X, Y ] = XY − Y X is the commutator. A non-vanishing δX represents a randomness in the measurement process and that may arise from non-commutivity (misalignment in the eigenbases) between state and observable, or even it may appear due to classical uncertainty present in the state (i.e., not due to quantum superposition). In fact the (3) does allow to have either δX or δY vanishing, but not simultaneously for a given state ρ and [X, Y ] 6= 0. However, when δX vanishes, it is nontrivial to infer on δY and vice versa. To overcome this drawback, the uncertainty relation is extended to sum-uncertainty relations, both in terms of variance (Maccone and Pati, 2014) and entropic quantities (Beckner, 1975; Białynicki-Birula and Mycielski, 1975; Deutsch, 1983; Maassen and Uffink, 1988). We refer to (Coles et al., 2015) for an excellent review on this subject. The entropic uncertainty relation was also considered in the presence of quantum memory (Berta et al., 2010). It has been shown that, in the presence of quantum memory, any two observable can simultaneously be measured with arbitrary precision. Therefore the randomness, appearing in the measurements, can be compensated by the side information stored in the quantum memory. As we mentioned in the previous section, Heisenberg uncertainty relations are closely related to the contextuality of quantum mechanics at the level of single systems. Non-commuting observables are indeed responsible for the fact that there exist no non-contextual hidden variable theories that can explain all results of quantum mechanical measurements on a given systems. The inherent randomness considered in this work is steam-

11 ing out the Born’s rule in quantum mechanics, irrespective of the fact if there are more than one observable being simultaneously measured or not. Furthermore the existence of nonlocal correlations (and quantum correlations) in the quantum domain, give rise to possibility of, in a sense, a new form of randomness. In the following we consider such randomness and its connection to nonlocal correlation. Before we do so, we shall discuss nonlocal correlations in more detail. B. Nonlocality 1. Two-party nonlocality

Let us now turn to important ingredient of quantum randomness, that is nonlocality, characterized by Bellinequalities (Bell, 1964; Brunner et al., 2014). In the traditional scenario a Bell nonlocality test relies on two spatially separated observers, say Alice and Bob, who perform spacelike measurements on a bipartite system possibly produced by a common source. For a schematic see Fig. 1. Suppose Alice’s measurement choices are x ∈ X = {1, . . . , MA } and Bob’s choices y ∈ Y = {1, . . . , MB } and the corresponding outcomes a ∈ A = {1, . . . , mA } and b ∈ B = {1, . . . , mB } respectively. After repeating many times, Alice and Bob communicate their measurement settings and outcomes to each other and estimate the joint probability p(a, b|x, y) = p(A = a, B = b|X = x, Y = y) where X, Y are the random variables that govern the inputs and A, B are the random variables that govern the outputs. The outcomes are considered to be correlated, for some x, y, a, b, if p(a, b|x, y) 6= p(a|x)p(b|y).

(4)

Observing such correlations is not surprising as there are many classical sources and natural processes that lead to correlated statistics. These can be modeled with the help of another random variable Λ with the outcomes λ, which has a causal influence on both the measurement outcomes and is inaccessible to the observers or ignored. In a local hiddenvariable model, considering all possible causes Λ, the joint probability can then be expressed as p(a, b|x, y, λ) = p(λ)p(a|x, λ)p(b|y, λ).

(5)

One, thereby, could explain any observed correlation in accordance with the fact that Alice’s outcomes solely depends on her local measurement settings x, on the common cause λ, and are independent of Bob’s measurement settings. Similarly Bob’s outcomes are independent of Alice’s choices. This assumption – the no-signaling condition – is crucial – it is required by the theory of relativity, where nonlocal causal influence between space-like separated observers is forbidden. Therefore, any joint probability, under the local hiddenvariable model, becomes Z p(a, b|x, y) = dλp(λ)p(a|x, λ)p(b|y, λ), (6) Λ

with the implicit assumption that the measurement settings x and y could be chosen independently of λ, i.e., p(λ|x, y) = p(λ). Note that so far we have not assumed anything about the nature of the local measurements, whether they are deterministic or not. In a deterministic local hidden-variable model, Alice’s outcomes are completely determined by the choice x and the λ. In other words, for an outcome a, given input x and hidden cause λ, the probability p(a|x, λ) is either 1 or 0 and so as for Bob’s outcomes. Importantly, the deterministic local hidden-variable model has been shown to be fully equivalent to the local hidden-variable model (Fine, 1982). Consequently, the observed correlations that admit a join probability distribution as in (6), can have an explanation based on a deterministic local hidden-variable model. In 1964, Bell showed that any local hidden-variable model is bound to respect a set of linear inequalities, which are commonly know as Bell-inequalities. In terms of joint probabilities they can be expressed as X xy αab p(a, b|x, y) ≤ SL , (7) a,b,x,y xy αab

where are some coefficients and SL is the classical bound. Any violation of Bell-inequalities (7) implies a presence of correlations that cannot be explained by a local hidden-variable model, and therefore have a nonlocal character. Remarkably, there indeed exists correlations violating Bell-inequalities that could be observed with certain choices local measurements on a quantum system, and hence do not admit a deterministic local hidden-variable model. To understand it better, let us consider an example of the most studied two-party Bell-inequalities, also known as CHSH-inequalities, introduced in (Clauser et al., 1969). Assume the simplest scenario (as in Fig. 1) in which Alice and Bob both choose one of two local measurements x, y ∈ {0, 1} and obtain one of two measurement outcomes a, b ∈ {−1, 1}. Let the expectation values of the local measurements are P hax by i = a,b a · b · p(a, b|x, y), then the CHSH-inequality reads: ICHSH = ha0 b0 i + ha0 b1 i + ha1 b0 i − ha1 b1 i ≤ 2.

(8)

One can maximize ICHSH using local deterministic strategy and to do so one needs to achieve the highest possible values of ha0 b0 i, ha0 b1 i, ha1 b0 i and the lowest possible value of ha1 b1 i. By choosing p(1, 1|0, 0) = p(1, 1|0, 1) = p(1, 1|1, 0) = 1, the first three expectation values can be maximized. However, in such situation the p(1, 1|1, 1) = 1 and ICHSH could be saturated to 2. Thus the inequality is respected. However, it can be violated in a quantum setting. For example, considering a quantum state |Ψ+ i = √12 (|00i + |11i) and measurement choices A0 = σz , A1 = σx , B0 = √1 (σz + σx ), B1 = √1 (σz − σx ) one could check that for the 2 2 quantum expectation values haα bβ i = hΨ+ |Aα ⊗Bβ |Ψ+ i we √ get ICHSH = 2 2. Here σz and σx are the Pauli spin matrices and |0i and |1i are two eigenvectors of σz . Therefore the joint probability distribution p(a, b|x, y) cannot be explained in terms of local deterministic model.

12

x1

xk

xn

a1

ak

an

p(a1,...,ak,...,an|x1,...,xk,...,xn) FIG. 2 Schematic representation of device independent approach. In this approach several users could access to uncharacterized black-boxes (shown as squares) possibly prepared by an adversary. The users are allowed to choose inputs (x1 , . . . , xk , . . . , xn ) for the boxes and acquire outputs (a1 , . . . , ak , . . . , an ) as results. The joint probability with which the outputs appear is p (a1 , . . . , ak , . . . , an |x1 , . . . , xk , . . . , xn ). 2. Multi-party nonlocality and device independent approach

Bell-type inequalities can be also constructed in multi-party scenario. Their violation signifies nonlocal correlations distributed over many parties. A detailed account may be found in (Brunner et al., 2014). Here we introduce the concept of nonlocality using the contemporary language of device independent approach (DIA) (Brunner et al., 2014). Recent successful hacking attacks on quantum cryptographic devices (Lydersen et al., 2010) triggered this novel approach to quantum information theory in which protocols are defined independently of the inner working of the devices used in the implementation. That leads to avalanches of works in the field of device independent quantum information processing and technology (Brunner, 2014; Pironio et al., 2015). The idea of DIA is schematically given in Fig. 2. We consider here the following scenario, usually referred to as the (n, m, d) scenario. Suppose n spatially separated parties A1 , . . . , An . Each of them possesses a black box with m measurement choices (or observables) and d measurement outcomes. Now, in each round of the experiment every party is allowed perform one choice of measurement and acquires one outcome. The accessible information, after the measurements, is contained in a set of (md)n conditional probabilities p(a1 , . . . , an |x1 , . . . , xn ) of obtaining outputs a1 , a2 , . . . , an , provided observables x1 , x2 , . . . , xn were measured. The set of all such probability distributions forms a convex set; in fact, it is a polytope in the probability manifold. From the physical point of view (causality, special relativity) the probabilities must fulfill the no-signaling condition, i.e., the choice of measurement by the k-th party, cannot be instantaneously signaled to the others. Mathematically it means that for any k = 1, . . . , n, the following condition P ak p(a1 , . . . , ak , . . . , an |x1 , . . . , xk , . . . , xn ) = p(a1 , . . . , ak−1 , ak+1 . . . , an |x1 , . . . , xk−1 , xk+1 . . . , xn ), is fulfilled.

The local correlations are defined via the concept of a local hidden variable λ with the associated probability qλ . The correlations that the parties are able to establish in such case are of the form

p(a1 , . . . , an |x1 , . . . , xn ) X = qλ D(a1 |x1 , λ) . . . D(an |xn , λ),

(9)

λ

where D(ak |xk , λ) are deterministic probabilities, i.e., for any λ, D(ak |xk , λ) equals one for some outcome, and zero for all others. What is important in this expression is that measurements of different parties are independent, so that the probability is a product of terms corresponding to different parties. In this n-party scenario the local hidden variable model bounds the joint probabilities to follow the Bell-inequalities, given as X ,...,xn p(a1 , . . . , an |x1 , . . . , xn ) αax11,...,a n a1 ,...,an ,x1 ,...,xn

≤ SLn ,

(10)

,...,xn where αax11,...,a are some coefficients and SLn is the classical n bound. The probabilities that follow local (classical) correlations form a convex set that is also a polytope, denoted P (cf. Fig. Qn 3). Its extremal points (or vertices) are given by i=1 D(ai |xi , λ) with fixed λ. The Bell theorem states that the quantum-mechanical probabilities, which also form a convex set Q, may stick out of the classical polytope (Bell, 1964; Fine, 1982). The quantum probabilities are given by the trace formula for the set of local measurements

p(a1 , . . . , an |x1 , . . . , xn ) = Tr(ρ ⊗ni=1 Maxii ),

(11)

where ρ is some n-partite state and Maxii denote the measurement operators (POVMs) for any choice of the measurement xi and party i. As we do not impose any constraint on the local dimension, we can always choose the measurements to be projective, i.e., the measurement operators additionally satisfy Max0i Maxii = δa0i ,ai Maxii . i This approach towards the Bell inequalities is explained in Fig. 3. Without loss of generality, we consider the simplest (2, 2, 2) scenario consisting of two parties, each measuring a pair of two-outcome observables – it can be straightforwardly extended to a multi-party setting. Any hyperplane in the space of probabilities that separates the classical polytope from the rest determines a Bell inequality: everything that is above the upper horizontal dashed line is obviously nonlocal. But the most useful are the tight Bell inequalities corresponding to the facets of the classical polytope, i.e. its walls of maximal dimensions (lower horizontal dashed line). In general (n, m, d) scenarios, the complexity of characterizing the corresponding classical polytope is enormous. It is fairly easy to see that, even for (n, 2, 2), the number of its

13

FIG. 3 Schematic representation of different sets of correlations: classical (grey area) and quantum (the area bounded by the thick line). Clearly, the former is the subset of the latter and, as has been shown by Bell (Bell, 1964), they are not equal – there are quantum correlations that do not fall into the grey area. The black dots represent the vertices of the classical polytope – deterministic classical correlations – satisfying deterministic local hidden variable models. The dashed lines represent Bell inequalities. In particular, the black dashed line is tight and it corresponds to the facet of the classical set.

P

vertices (extremal points) is equal to 22n , hence it grows exponentially with n. Nevertheless, a considerable effort has been made in recent time to characterize multi-party nonlocality (Brunner et al., 2014; Liang et al., 2015; Rosicka et al., 2016; Tura et al., 2014a, 2015, 2014b). Among the many other device independent applications, the nonlocality appears to be a valuable resource in random number generation, certification, expansion and amplification, which we outline in the following subsections. In fact, it has been shown that Bell nonlocal correlation is a genuine resource, in the framework of a resource theory, where the allowed operations are restricted to device independent local operations (Gallego et al., 2012; Vicente, 2014). C. Randomness: information theoretic approach

Before turning to the quantum protocols involving randomness, we discuss in this section randomness from the information theory standpoint. It is worth mentioning the role of randomness in various applications, beyond its fundamental implications. In fact the randomness is a resource in many different areas – for a good overview see Refs. (Menezes et al., 1996; Motwani and Raghavan, 1995). Random numbers play important role in cryptographic applications, in numerical simulations of complex physical, chemical, economical, social and biological systems, not to mention gambling. That is why, much efforts were put forward to (1) develop good, reliable sources of random numbers, and (2) to design reliable certification tests for a given source of random numbers. In general, there exists three types of random number generators (RNG). They are “true” RNGs, pseudo-RNGs and the quantum-RNGs. The true RNGs are based on some physical processes that are hard to predict, like noise in electrical circuits, thermal or atmospheric noises, radioactive decays etc. The pseudo-RNGs rely on the output of a deterministic function with a shorter random seed possibly generated by a true RNG. Finally, quantum RNGs use genuine quantum features to generate random bits. We consider here a finite sample space and denote it by the

set Ω. The notions of ideal and weak random strings describe distributions over Ω with certain properties. When a distribution is uniform over Ω, we say that it has ideal randomness. A uniform distribution over n-bit strings is denoted by Un . The uniform distributions are very natural to work with. However, when we are working with physical systems, the processes or measurements are usually biased. Then the bit strings resulting from such sources are not uniform. A string with nonuniform distribution, due to some bias (could be unknown), is referred to have weak randomness and the sources of such strings are termed as weak sources. Consider the random variables denoted by the letters (X, Y, . . .). Their values will be denoted by (x, y, . . .). The probability of a random variable X with a value x is denoted as p(X = x) and when the random variable in question is clear we use the shorthand notation p(x). Here we briefly introduce the operational approach to define randomness of a random variable. In general, the degree of randomness or bias of a source is unknown and it is insufficient to define a weak source by a random variable X with a probability distribution P (X). Instead one needs to model the weak randomness by a random variable with an unknown probability distribution. In another words, one need to characterize a set of probability distributions with desired properties. If we suppose that the probability distribution P (X) of the variable X comes from a set Ω, then the degree of randomness is determined by the properties of the set, or more specifically, by the least random probability distribution(s) in the set. The types of weak randomness differ with the types of distribution P (X) on Ω and the set Ω itself – they are determined by the allowed distributions motivated by a physical source. There are many ways to classify the weak random sources, and an interested reader may go through Ref. (Pivoluska and Plesch, 2014). Here we shall consider two types of weak random sources, Santha-Vazirani (SV) and Min-Entropy (ME) sources, which will be sufficient for our later discussions. A Santha-Vazirani (SV) source (Santha and Vazirani, 1986) is defined as a sequence of binary random variables (X1 , X2 , . . . , Xn ), such that 1 1 − ≤p(xi = 1|x1 , . . . , xi−1 ) ≤ + , 2 2 ∀i ∈ N, ∀x1 , . . . , xi−1 ∈ {0, 1},

(12)

where the conditional probability p(xi = 1|x1 , . . . , xi−1 ) is the probability of the value xi = 1 conditioned on the values x1 , . . . , xi−1 . The 0 ≤ ≤ 21 represents bias of the source. For fixed and n, the SV-source represents a set of probability distributions over n-bit strings. If a random string satisfies (12), then we say that it is -free. For = 0 the string is perfectly random – uniformly distributed sequence of bits Un . For = 12 , nothing can be inferred about the string and it can be even deterministic. Note that in SV sources the bias can not only change for each bit Xi , but it also can depend on the previously generated bits. It requires that each produced bit must have some amount of randomness, when 6= 12 , and even be conditioned on the previous one.

14 In order the generalize it further one considers block source (Chor and Goldreich, 1988), where the randomness is not guaranteed in every single bit, but rather for a block of nbits. Here, in general, the randomness is quantified by the min-entropy, which is defined as: H∞ (Y ) = miny [−log2 (p(Y = y))],

(13)

for an n-bit random variable Y . For a block source, the randomness is guaranteed by the most probable n-bit string appearing in the outcome of the variable – simply by guessing the most probable element – provided that the probability is less than one. A block (n, k) source can now be modeled, for n-bit random variables (X1 , X2 , ..., Xn ), such that H∞ (Xi |Xi−1 = xi−1 , ..., X1 = x1 ) ≥ k,

(14)

∀i ∈ N, ∀x1 , ..., xi−1 ∈ {0, 1}n . These block-sources are generalizations of SV-sources; the latter are recovered with n = 1 and = 2−H∞ (X) − 12 . The block sources can be further generalized to sources of randomness of finite output size, where no internal structure is given, e.g., guaranteed randomness in every bit (SV-sources) or every block of certain size (block sources). The randomness is only guaranteed by its overall min-entropy. Such sources are termed as the min-entropy sources (Chor and Goldreich, 1988) and are defined, for an n-bit random variable X, such that H∞ (X) ≥ k.

(15)

Therefore, a min-entropy source represents a set of probability distributions where the randomness is upper-bounded by the probability of the most probable element, measured by minentropy. Let us now briefly outline the randomness extraction (RE), as it is one of the most common operations that is applied in the post-processing of weak random sources. The randomness extractors are the algorithms that produce nearly perfect (ideal) randomness useful for potential applications. The aim of RE is to convert randomness of a random string from a weak source into a possibly shorter string of bits that is close to a perfectly random one. The closeness is defined as follows. The random variables X and Y over a same domain Ω are ε-close, if: 1X ∆(X, Y ) = |p(X = x) − p(Y = x)| ≤ ε. 2

(16)

x∈Ω

With respect to RE, the weak sources can be divided into two classes – extractable sources and non-extractable sources. Only from extractable sources a perfectly random string can be extracted by a deterministic procedure. Though there exist many non-trivial extractable sources (see for example (Kamp et al., 2011)), most of the natural sources, defined by entropies, are non-extractable and in such cases nondeterministic (stochastic) randomness extractors are necessary.

Deterministic extraction fails for the random strings from SV-sources, but it is possible to post-process them with a help of an additional random string. As shown in (Vazirani, 1987), for any and two mutually independent -free strings from SV-sources, it is possible to efficiently extract a single almost perfect bit (0 → 0). For two n-bit independent strings X = (X1 , ..., Xn ) and Y = (Y1 , ...Yn ), the post-processing function, Ex, has been defined as Ex(X, Y ) = (X1 · Y1 ) ⊕ (X2 · Y2 ) ⊕ · · · ⊕ (Xn · Yn ), (17) where ⊕ denotes the sum modulo 2. The function Ex is the inner product between the n-bit strings X and Y modulo 2. Note, randomness extraction of SV-sources are sometime referred as randomness amplification as two -free strings from SV-sources are converted to one-bit string of 0 -free and with 0 < . Deterministic extraction is also impossible for the minentropy sources. Nevertheless, an extraction might be possible with the help of seeded extractor in which an extra resource of uniformly distributed string, called the seed, is exploited. A function, Ex : {0, 1}n × {0, 1}r 7→ {0, 1}m is seeded (k, ε)-extractor, for every string from block (n, k)-source of random variable X, if ∆(Ex(X, Ur ), Um ) ≤ ε.

(18)

Here Ur (Um ) is the uniformly distributed r-bit (m-bit) string. In fact, for a variable X, min-entropy gives the upper bound on the number of extractable perfectly random bits (Shaltiel, 2002). Randomness extraction is well developed area of research in classical information theory. There are many randomness extraction techniques using multiple strings (Barak et al., 2010; Dodis et al., 2004; Gabizon and Shaltiel, 2008; Nisan and Ta-Shma, 1999; Raz, 2005; Shaltiel, 2002), such as universal hashing extractor, Hadamard extractor, DEOR extractor, BPP extractor etc., useful for different post-processing.

D. Nonlocality, random number generation and certification

Here we link the new form of randomness, i.e the presence of nonlocality (in terms of Bell-violation) in the quantum regime, to random number generation and certification. To do so, we outline how nonlocal correlations can be used to generate new types of random numbers, what has been experimentally demonstrated in (Pironio et al., 2010). Consider the Bell-experiment scenario (Fig. (1)), as explained before. Two space-like separated observers perform different measurements, labeled as x and y, on two quantum particles in their possession and get measurement outcomes a and b, respectively. With many repetitions they can estimate the joint probability, p(a, b|x, y), for the outcomes a and b with the measurement choices x and y. With the joint probabilities the

15 observers could check if the Bell-inequalities are respected. If a violation is observed the outcomes are guaranteed to be random. The generation of these random numbers is independent of working principles of the measurement devices. Hence, this is a device independent random number generation. In fact, there is a quantitative relation between the amount of Bell-inequalities violation and the observed randomness. Therefore, these random numbers could be (a) certifiable, (b) private, and (c) device independent (Colbeck, 2007; Pironio et al., 2010). The resulting string of random bits, obtained by N uses of measurement device, would be made up of N pair of outcomes, (a1 , b1 , ..., aN , bN ), and their randomness could be guaranteed by the violation of Bell-inequalities. There is however an important point to be noted. A priori, the observers do not know whether the measurement devices violate Bell-inequalities or not. To confirm they need to execute statistical tests, but such tests cannot be carried out in predetermined way. Of course, if the measurement settings are known in advance, then an external agent could prepare the devices that are completely deterministic and the Bellinequalities violations could be achieved even in the absence of nonlocal correlations. Apparently there is a contradiction between the aim of making random number generator and the requirement of random choices to test the nonlocal nature of the devices. However, it is natural to assume that the observers can make free choices when they are separated. Initially it was speculated that the more particles are nonlocally correlated (in the sense of Bell violation), the stronger would be the observed randomness. However, this intuition is not entirely correct, as shown in (Acín et al., 2012) – a maximum production of random bits could be achieved with a non-maximal CHSH violation. To establish a quantitative relation between the nonlocal correlation and generated randomness, let us assume that the devices follow quantum mechanics. There exist thus a quantum state ρ and measurement operators (POVMs) of each device Max and Mby such that the joint probability distribution P (a, b|x, y) could be expressed, through Born rule, as PQ (a, b|x, y) = Tr(ρMax ⊗ Mby ),

(19)

where the tensor product reflects the fact that the measurements are local, i.e., there are no interactions between the devices, while the measurement takes place. The set of quantum correlations consists of all such probability distributions. Consider a linear combinations of them, X xy αab PQ (a, b|x, y) = S, (20) x,y,a,b xy specified by real coefficients αab . For local hidden-variable xy model, with certain coefficients αab , the Bell-inequalities can be then expressed as

S ≤ SL .

(21)

This bound could be violated (S > SL ) for some quantum states and measurements indicating that the state contains nonlocal correlation.

Let us consider the the measure of randomness quantified by min-entropy. For a d-dimensional probability distribution P (X), describing a random variable X, the min-entropy is defined as H∞ (X) = −log2 [maxx p(x)]. Clearly, for the perfectly deterministic distribution this maximum equals one and the min-entropy is zero. On the other hand, for a perfectly random (uniform) distribution, the entropy acquires the maximum value, log2 d. In the Bell-scenario, the randomness in the outcomes, generated by the pair of measurements x and y, reads H∞ (A, B|x, y) = −log2 cxy , where cxy = maxab PQ (a, b|x, y). For a given observed value S > SL , violating Bell-inequality, one could find a quantum realization, i.e., the quantum states and set of measurements, that minimizes the min-entropy of the outcomes H∞ (A, B|x, y) (Navascués et al., 2008). Thus, for any violation of Bellinequalities, the randomness of a pair of outcomes satisfies H∞ (A, B|x, y) ≥ f (S),

(22)

where f is a convex function and vanishes for the case of no Bell-inequalities violation, S ≤ SL . Hence, the (22) quantitatively states that a violation of Bell-inequalities guarantees some amount randomness. Intuitively, if the joint probabilities admit (21), then for each setting x, y and a hidden cause λ, the outcomes a and b can be deterministically assigned. However, the violation of (21) rules out such possibility. As a consequence, the observed correlation cannot be understood with a deterministic model and the outcomes are fundamentally undetermined at the local level. Although there are many different approaches to generate random numbers, the certification of randomness is highly non-trivial. In general the certification relies on series of statistical tests (Bassham et al., 2010; Marsaglia, 2008) that are designed to check absence of any pattern present in the generated sequence. In fact, no finite set of tests could be considered as complete (Bassham et al., 2010). However, this problem could be solved, in one stroke, if the random sequence shows a Bell-violation, as it certifies a new form of “true” randomness that has no deterministic classical analogue.

E. Nonlocality and randomness expansion

Nonlocal correlations can be also used for the construction of novel types of randomness expansion protocols. In these protocols, a user expands an initial random string, known as seed, into a larger string of random bits. Here, we focus on protocols that achieve this expansion by using randomness certified by a Bell inequality violation. Since the first proposals in Refs. (Colbeck, 2007; Pironio et al., 2010), there have been several different works studying Bell-based randomness expansion protocols, see for instance (Arnon-Friedman et al., 2016; Chung et al., 2014; Colbeck, 2007; Colbeck and Kent, 2011; Coudron and Yuen, 2013; Miller and Shi, 2014, 2016; Pironio et al., 2010; Vazirani and Vidick, 2012). It is not the scope of this section to review the contributions of all these

16 works, which in any case should be interpreted as a representative but non-exhaustive list. However, most of them consider protocols that have the structure described in what follows. Note that the description aims at providing the main intuitive steps in a general randomness expansion protocols and technicalities are deliberately omitted (for details see the references above). The general scenario consists of a user who is able to run a Bell test. He thus has access to n ≥ 2 devices where he can implement m local measurements of d outputs. For simplicity, we restrict the description in what follows to protocols involving only two devices, which are also more practical form an implementation point of view. The initial seed is used to randomly choose the local measurement settings in the Bell experiment. The choice of settings does not need to be uniformly random. In fact, in many situations, there is a combination of settings in the Bell test that produces more randomness than the rest. It is then convenient to bias the choice of measurement towards these settings so that (i) the amount of random bits consumed from the seed, denoted by Ns , is minimize and (ii) the amount of randomness produced during the Bell test is maximized. The choice of settings is then used to perform the Bell test. After N repetitions of the Bell test, the user acquires enough statistics to have a proper estimation of the non-locality of the generated data. If not enough confidence about a Bell violation is obtained in this process, the protocol is aborted or more data are generated. From the observed Bell violation, it is possible to bound the amount of randomness in the generated bits. This is often done by means of the so-called min-entropy, H∞ . In general, for a random variable X, the min-entropy is expressed in bits and is equal to H∞ = − log2 maxx P (X = x). The observed Bell violation is used to establish a lower bound on the min-entrop of the generated measurement outputs. Usually, after this process, the user concludes that with high confidence the NG ≤ N generated bits have an entropy at least equal to R ≤ Ng , that is H∞ ≥ R. This type of bounds is useful to run the last step in the protocol: the final randomness distillation using a randomness extractor (Nisan and Ta-Shma, 1999; Trevisan, 2001). This consists of classical post-processing of the measurement outcomes, in general assisted by some extra Ne random bits from the seed, which map the Ng bits with entropy at least R to R bits with the same entropy, that is, R random bits. Putting all the things together, the final expansion of the protocols is given by the ration R/(Ns + Ne ). Every protocol comes with a security proof, which guarantees that the final list of R bits is unpredictable to any possible observer, or eavesdropper, who could share correlated quantum information with the devices used in the Bell test. Security proofs are also considered in the case of eavesdroppers who can go beyond the quantum formalism, yet without violating the no-signalling principle. All the works mentioned above representes important advances in the design of randomness expansion protocols. At the moment, it is

for instance known that (i) any Bell violation is enough to run a randomness expansion protocol (Miller and Shi, 2016) and (ii) in the previous simplest configuration presented here, there exist protocol attaining an exponential randomness expansion (Vazirani and Vidick, 2012). More complicated variants, where devices are nested, even attain an unbounded expansion (Chung et al., 2014; Coudron and Yuen, 2013). Before concluding, it is worth mentioning that another interesting scenario consists of the case in which a trusted source of public randomness is available. Even if public, this trusted randomness can safely be used to run the Bell test and does not need to be taken into account in the final rate.

F. Nonlocality and randomness amplification

Here we discuss the usefulness of nonlocal correlation for randomness amplification, a task related but in a way complementary to randomness expansion. While in randomness expansion one assumes the existence of an initial list of perfect random bits and the goal is to generate a longer list, in randomness amplification the user has access to a list of imperfect randomness and the goal is to generate a source of higher, ideally arbitrarily good, quality. As above, the goal is to solve this information task by exploiting Bell violating correlations. SV (x,t)

t

x x1

x2

x3

x4 n runs

a1

x

a2

a3

a

a Test

No Abort

a4

Yes

Extractor

Output

FIG. 4 (Color online.) Scheme of randomness amplification using four devices, as in (Brandão et al., 2016). The devices are shielded from each other as indicated with black barriers. The local measurement choices, in each run, are governed by the part of the SV-source, x, and corresponding output forms random bits a. After n-runs Belltest is performed (denoted by – Test). If the test violates Bell inequalities, then the outputs and rest of the initial SV-source t are fed into an extractor (denoted by – Extractor) in order to obtain final outputs. If the test doesn’t violate Bell inequalities, the protocol is aborted.

Randomness amplification based on non-locality was introduced in (Colbeck and Renner, 2012). There, the initial source of imperfect randomness consisted of a SV source. Recall that the amplification of SV sources is impossible by classical

17 means. A protocol was constructed based on the two-party chained Bell-inequalities that was able to map an SV source with parameter < 0.058 into a new source with arbitrarily close to zero. This result is only valid in an asymptotic regime in which the user implements the chained Bell inequality with an infinite number of measurements. Soon after, a more complicated protocol attained full randomness amplification (Gallego et al., 2013), that is, it was able to map SV sources of arbitrarily weak randomness, < 1/2, to arbitrarily good sources of randomness, → 0. The final result was again asymptotic, in the sense that to attain full randomness amplification the user now requires an infinite number of devices. Randomness amplification protocols have been studied by several other works, see for instance (Bouda et al., 2014; Brandão et al., 2016; Chung et al., 2014; Coudron and Yuen, 2013; Grudka et al., 2014; Mironowicz et al., 2015; Wojewódka et al., 2016). As above, the scope of this section is not to provide a complete description of all the works studying the problem of randomness amplification, but rather to provide a general framework that encompasses most of them. In fact, randomness amplification protocols (see e.g., Fig. 4) have a structure similar to randomness expansion ones. The starting point of a protocol consists of a source of imperfect randomness. This is often modelled by an SV source, although some works consider a weaker source of randomness, known as min-entropy source, in which the user only knows a lower bound on the min-entropy of the symbols generated by the source (Bouda et al., 2014; Chung et al., 2014). The bits of imperfect randomness generated by the user are used to perform N repetitions of the Bell test. If the observed Bell violation is large enough, with enough statistical confidence, bits defining the final source are constructed from the measurement outputs, possibly assisted by new random bits from the imperfect source. Note that contrarily to the previous case, the extraction process cannot be assisted with a seed of perfect random numbers, as this seed could be trivially be used to produce the final source. As in the case of expansion protocols, any protocol should be accompanied by a security proof showing that the final bits are unpredictable to any observer sharing a system correlated with the devices in the user’s hands.

IV. QUANTUM RANDOMNESS AND TECHNOLOGY

Random numbers have been a part of human technology since ancient times. If Julius Caesar indeed said “Alea iacta est” (“the die is cast,”) when he crossed the Rubicon, he referred to a technology that had already been in use for thousands of years. Modern uses for random numbers include cryptography, computer simulations, dispute resolution, and gaming. The importance of random numbers in politics, social science and medicine should also not be underestimated; randomized polling and randomized trials are essential methodology in these areas. A major challenge for any modern randomness technology

is quantification of the degree to which the output could be predicted or controlled by an adversary. A common misconception is that the output of a random number generator can be tested for randomness, for example using statistical tests such as Diehard/Dieharder (Brown, 2004; Marsaglia, 2008), NIST SP800-22 (Rukhin et al., 2010), or TestU01 (L’Ecuyer and Simard, 2007). While it is true that failing these tests indicates the presence of patterns in the output, and thus a degree of predictability, passing the tests does not indicate randomness. This becomes clear if you imagine a device that on its first run outputs a truly random sequence, perhaps from ideal measurements of radioactive decay, and on subsequent runs replays this same sequence from a recording it kept in memory. Any of these identical output sequences will pass the statistical tests, but only the first one is random; the others are completely predictable. We can summarize this situation with the words of John von Neumann: “there is no such thing as a random number – there are only methods to produce random numbers” (von Neumann, 1951). How can we know that a process does indeed produce random numbers ? In light of the difficulties in determining the predictability of the apparent randomness seen in thermal fluctuations and other classical phenomena, using the intrinsic randomness of quantum processes is very attractive. One approach, described in earlier sections, is to use deviceindependent methods. In principle, device-independent randomness protocols can be implemented with any technology capable of a strong Bell-inequality violation, including ions (Pironio et al., 2010), photons (Giustina et al., 2015; Shalm et al., 2015), nitrogen-vacancy centers (Hensen et al., 2015), neutral atoms (Rosenfeld et al., 2011) and superconducting qubits (Jerger et al., 2016). Device-independent randomness expansion based on Bell inequality violations was first demonstrated using a pair of Yb+ ions held in spatially-separated traps (Pironio et al., 2010). In this protocol, each ion is made to emit a photon which, due to the availability of multiple decay channels with orthogonal photon polarizations, emerges from the trap entangled with the internal state of the ion. When the two photons are combined on a beam-splitter, the Hong-Ou-Mandel effect causes a coalescence of the two photons into a single output channel, except in the case that the photons are in a polarization-antisymmetric Bell state. Detection of a photon pair, one at each beam-splitter output, thus accomplishes a projective measurement onto this antisymmetric Bell state, and this projection in turn causes an entanglement swapping that leaves the ions entangled. Their internal states can then be detected with high efficiency using fluorescence readout. This experiment strongly resembles a loophole-free Bell test, with the exception that the spatial separation of about one meter is too short to achieve space-like separation. Due to the low probability that both photons were collected and registered on the detectors, the experiment had a very low success rate, but this does not reduce the degree of Bell inequality violation or the quality of the randomness produced. The experiment generated 42 random bits in about one month of continuous

18 running, or 1.6 × 10−5 bits/s. A second experiment, in principle similar but using very different technologies, was performed with entangled photons and high-efficiency detectors (Christensen et al., 2013) to achieve a randomness extraction rate of 0.4 bits/s. While further improvements in speed can be expected in the near future (National Institutes of Standards and Technology, 2011), at present device-independent techniques are quite slow, and nearly all applications must still use traditional quantum randomness techniques. It is also worth noting that device-independent experiments consume a large quantity of random bits in choosing the measurement settings in the Bell test. Pironio et al. used publicly available randomness sources drawn from radioactive decay, atmospheric noise, and remote network activity. Christensen et al. used photon-arrival-time random number generators to choose the measurement settings. Although it has been argued that no additional physical randomness is needed in Bell tests (Pironio, 2015), there does not appear to be agreement on this point. At least in practice if not in principle, it seems likely that there will be a need for traditional quantum randomness technology also in device-independent protocols. If one does not stick to the device-independent approach, it is in fact fairly easy to obtain signals from quantum processes, and devices to harness the intrinsic randomness of quantum mechanics have existed since the 1950s. This began with devices to observe the timing of nuclear decay (Isida and Ikeda, 1956), followed by a long list of quantum physical processes including electron shot noise in semiconductors, splitting of photons on beam-splitters, timing of photon arrivals, vacuum fluctuations, laser phase diffusion, amplified spontaneous emission, Raman scattering, atomic spin diffusion, and others. See (Herrero-Collantes and Garcia-Escartin, 2016) for a thorough review. While measurements on many physical processes can give signals that contain some intrinsic randomness, any real measurement will also be contaminated by other signal sources, which might be predictable or of unknown origin. For example, one could make a simple random number generator by counting the number of electrons that pass through a Zener diode in a given amount of time. Although electron shot noise will make an intrinsically-random contribution, there will also be an apparently-random contribution from thermal fluctuations (Johnson-Nyquist noise), and a quite non-random contribution due to technical noises from the environment. If the physical understanding of the device permits a description in terms of the conditional min-entropy (see Section III.C) H∞ (Xi |hi ) ≥ k, ∀i ∈ N, ∀hi

(23)

where Xi is the i’th output string and hi is the “history” of the the device at that moment, including all fluctuating quantities not ascribable to intrinsic randomness, then randomness extraction techniques can be used to produce arbitrarily-good output bits from this source. Establishing this min-entropy level can be an important challenge, however.

The prevalence of optical technologies in recent work on quantum random number generators is in part in response to this challenge. The high coherence and relative purity of optical phenomena allows experimental systems to closely approximate idealized quantum measurement scenarios. For example, fluorescence detection of the state of a single trapped atom is reasonably close to an ideal von Neumann projective measurement, with fidelity errors at the part-per-thousand level (Myerson et al., 2008). Some statistical characterizations can also be carried out directly using the known statistical properties of quantum systems. For example, in linear optical systems shot noise can be distinguished from other noise sources based purely on scaling considerations, and provides a very direct calibration of the quantum versus thermal and technical contributions, without need for detailed modeling of the devices used. Considering an optical power measurement, the photocurrent I1 that passes in unit time will obey var(I1 ) = A + BhI1 i + ChI1 i2

(24)

where A is the “electronic noise” contribution, typically of thermal origin, ChI1 i2 is the technical noise contribution, and BhI1 i is the shot-noise contribution. Measuring var(I1 ) as a function of hI1 i then provides a direct quantification of the noise contributed by each of these distinct sources. This methodology has been used to estimate entropies in continuous-wave phase diffusion random number generators (Xu et al., 2012). To date, the fastest quantum random number generators are based on laser phase diffusion (Abellán et al., 2014; Jofre et al., 2011; Xu et al., 2012; Yuan et al., 2014), with the record at the time of writing being 68 Gbits/second (Nie et al., 2015). These devices, illustrated in Fig. 5 work entirely with macroscopic optical signals (the output of lasers), which greatly enhances their speed and signal-to-noise ratios. It is perhaps surprising that intrinsic randomness can be observed in the macroscopic regime, but in fact laser phase diffusion (and before it maser phase diffusion) was one of the first predicted quantum-optical signals, described by Schawlow and Townes in 1958 (Schawlow and Townes, 1958). Because stimulated emission is always accompanied by spontaneous emission, the light inside a laser cavity experiences random phase-kicks due to spontaneous emission. The laser itself has no phase-restoring mechanism; its governing equations are phase-invariant, and the phase diffuses in a random walk. As the kicks from spontaneous emission accumulate, the phase distribution rapidly approaches a uniform distribution on [0, 2π), making the laser field a macroscopic variable with one degree of freedom fully randomized by intrinsic randomness. The phase diffusion accumulated in a given time can be detected simply by introducing an optical delay and interfering earlier output with later output in an unbalanced Mach-Zehnder interferometer. It is worth noting that the phase distribution is fully insensitive to technical and thermal contributions; it is irrelevant if the environment or an adversary introduces an additional

19

a)! laser

𝜙!

interference! p!

V0!

detector

uMZI!

phase diffusion!

b)!

V! d!

detection! digitization!

c)!

d)! 966 mV! ~ 10 mV!

phase signal!

noises! V0±8σnoise! d=0!

(d)! 200 mV!

500 ps!

d=1!

FIG. 5 (Color online.) Laser phase-diffusion quantum random number generator (LPD-QRNG). a) schematic diagram showing components of a LPD-QRNG using a pulsed laser and single-bit digitization. A laser, driven with pulses of injection current, produces optical pulses with very similar wave-forms and with relative phases randomized due to phase diffusion between the pulses. The pulses enter a single-mode fiber unbalanced Mach-Zehnder interferometer (uMZI), which produces interference between subsequent pulses, converting the random relative phase into a strong variation in the power reaching the detector, a linear photodiode. A comparator produces a digital output in function of the detector signal. b) Statistics of the pulse shapes produced by the laser, obtained by blocking one arm of the interferometer and recording on an oscilloscope. Main image shows the distribution of the pulse shapes warmer colors show higher probability density. Strong “relaxation oscillations” are seen, but are highly reproducible; all traces show the same behavior. Side image shows histogram taken in the region labeled in orange, showing a narrow peak indicating very small variation in power from one pulse to the next. c) Same acquisition strategy, but using both arms of the interferometer and thus producing interference. The variation due to the random phase ∆φ is orders of magnitude stronger that the noise, and the minimum values approach zero power, indicating high interference visibility. The histogram shows the classic “arcsine” distribution expected for the cosine of a random phase. d) illustration of the digitization process. Curve and points show expected and observed frequencies for the input voltage, approximating an arcsine distribution. The finite width of the peaks is a result of convolving the ideal arcsine distribution with the noise distribution, of order 10 mV. The comparator splits assigns a value d = 0 or d = 1 in function of the input voltage. The probability of a noise-produced error can be bounded by considering the effect of noise on digitization, giving an experimentally-guaranteed min-entropy for the output bits.

phase shift if the phase, a cyclic variable, is already described by a uniform distribution on the circle. Considerable effort has gone into determining the minentropy due to intrinsic randomness of laser phase-diffusion random number generators (Mitchell et al., 2015), especially in the context of Bell tests (Abellán et al., 2015). To date, laser phase diffusion random number generators have been used to choose the settings for all loophole-free Bell tests (Giustina et al., 2015; Hensen et al., 2015; Shalm et al., 2015). Here we outline the modeling and measurement considerations used to bound the min-entropy of the output of these devices. Considering an interferometer with two paths, short (S) and long (L) with relative delay τ , fed by the laser output E(t) = |E(t)| exp[iφ(t)], the instantaneous power that reaches the detector is p pI (t) = pS (t) + pL (t) + 2V pS (t)pL (t) cos ∆φ(t), (25) where pS (t) ≡ 41 |E(t)|2 , pL (t) ≡ 41 |E(t − τ )|2 , ∆φ(t) = φ(t) − φ(t − τ ) and V is the interference visibility. Assuming τ gives sufficient time for a full phase diffusion, ∆φ(t) is uniformly distributed on [0, 2π) due to intrinsic quantum randomness. The contributions of pS (t) and pL (t), however, may reflect technical or thermal fluctuations, and constitute a contamination of the measurable signal pI (t). The process of

detection converts this to a voltage V (t), and in doing so adds other technical and thermal noises. Also, the necessarily finite speed of the detection system implies that V (t) is a function of not only of pI (t), but also to a lesser extent of prior values pI (t0 ), t0 < t. This “hangover,” which is predictable based on the previous values, must be accounted for so as to not overestimate the entropy in pI (t). Digitization is the conversion from the continuous signal V to a digital value d. Considering only the simplest case of binary digitization, we have 0 V (ti ) < V0 di = (26) 1 V (ti ) ≥ V0 where V0 is the threshold voltage, itself a random variable influenced by a combination of thermal and technical noises. We can now express the distribution of d in function of the total noise Vnoise r 2 1 Vnoise P (d = 1|Vnoise ) = arcsin + (27) π 2 2∆V √ where 2∆V ∝ 4V pS pL is the peak-to-peak range of the signal due to the random ∆φ. This derives from the “arcsin” distribution that describes the cumulative distribution function of the cosine of a uniformly-random phase.

20 The noise contributions can all be measured in ways that conservatively estimate their variation; for example interrupting one or the other path of the interferometer we can measure the distribution of pS and pL , and comparing digitizer input to output we can upper bound the variation in V0 . With the measured distributions in hand, we can assign probabilities to Vnoise and thus to the min-entropy of d. For example, if the total noise Vnoise is described by a normal distribution with zero mean and width σnoise = 10 mV, and ∆V = 0.5 V, a probability P (d|Vnoise ) > P (d = 1|8σnoise ) ≈ 12 + 0.0511 will occur as often as Vnoise exceeds 8σnoise , which is to say with probability ≈ 6 × 10−16 . Since P (d|Vnoise ) ≤ 21 + 0.0511 implies a single-bit min entropy H∞ (d|Vnoise ) > 0.86, a randomness extraction based on this min-entropy can then be applied to give fully-random output strings. It is worth emphasizing that the characterizations used to guarantee randomness of this kind of source are not measurements of the digital output of the source, which as mentioned already can never demonstrate randomness. Rather, they are arguments based on physical principles, backed by measurements of the internal workings of the device. To summarize the argument, the trusted random variable ∆φ(t) is known to be fully randomized by the intrinsic quantum randomness of spontaneous emission. This statement relies on general principles that concern laser physics, such as Einstein’s A and B coefficient argument linking spontaneous emission to stimulated emission and the fact that lasers have no preferred phase, due to the time-translation invariance of physical law. The next step of the guarantee follows from a model of the interference process, Eq. (25), whose simplicity mirrors the simplicity of the experimental situation, in which single-mode devices (fibres) are used to ensure a low-dimensional field characterized only by the time variable. Finally there is an experimental characterization of the noises and a simple computation to bound their effects on the distribution of outputs. The computation can and should be performed with worst-case assumptions, assuming for example that all noise contributions are maximally correlated, unless the contrary has been experimentally demonstrated.

V. QUANTUM RANDOMNESS AND FUTURE

Randomness is a fascinating concept and we are now living a fascinating moment to study the problem of randomness certification and generation using quantum effects. The present review has tried to provide an overview of the problem, trying to cover many of the implications and directions emerging in the study of this problem. From a philosophical and fundamental perspective, the recent results have significantly improved our understanding of what can and cannot be said about randomness in nature using quantum physics. While the presence of randomness cannot be proven without making some assumptions about the systems, theses assumptions are constantly weakened and it is an interesting open research problem to identify the weakest set

of assumption sufficient to certify the presence of randomness. From a theoretical physics perspective, the recent results have provided a much better understanding of the relation between non-locality and randomness in quantum theory. Still, the exact relation between these two fundamental concepts is not fully understood. For instance, small amount of nonlocality, or even entanglement, sometimes suffice to certify the presence of maximal randomness in the measurement outputs of a Bell experiment (Acín et al., 2012). The relation between non-locality and randomness can also be studied in the larger framework of non-signaling theories, that is theories only limited by the no-signaling principle, which can go beyond quantum physics (Popescu and Rohrlich, 1994). For instance it is known that in these theories maximal randomness certification is impossible, while it is in quantum physics (de la Torre et al., 2015). From a more applied perspective, quantum protocols for randomness generation follow different approaches and require different assumptions. Until very recently, all quantum protocol required a precise knowledge of the devices used in the protocol and certified the presence of randomness by means of standard statistical tests. The resulting protocols are cheap, feasible to implement in practice, including the development of commercial products, and lead to reasonably high randomness generation rates. Device-independent solutions provide a completely different approach, in which no modeling of the devices is needed and the certification comes from a Bell inequality violation. Their implementation is however more challenging and only few much slower experimental realizations have until now been reported. Due to the importance and need of random numbers in our information society, we expect important advances in all these approaches, resulting in a large variety of quantum empowered solutions for randomness generation. VI. ACKNOWLEDGEMENTS

We acknowledge financial support from the John Templeton Foundation, the European Commission (FETPRO QUIC, STREP EQuaM and RAQUEL), the European Research Council (AdG OSYRIS, AdG IRQUAT, StG AQUMET, CoG QITBOX and PoC ERIDIAN), the AXA Chair in Quantum Information Science, the Spanish MINECO (Grants No. FIS2008-01236, No. FIS2013-40627-P, No. FIS201346768-P FOQUS, FIS2014-62181-EXP, FIS2015-68039-P, and Severo Ochoa Excellence Grant SEV-2015-0522) with the support of FEDER funds, the Generalitat de Catalunya (Grants No. 2014-SGR-874, No. 2014-SGR-875, No. 2014SGR-966 and No. 2014-SGR-1295), and Fundació Privada Cellex. REFERENCES Abellán, C, W. Amaya, M. Jofre, M. Curty, A. Acín, J. Capmany, V. Pruneri, and M. W. Mitchell (2014), “Ultra-fast quantum ran-

21 domness generation by accelerated phase diffusion in a pulsed laser diode,” Opt. Express 22 (2), 1645–1654. Abellán, Carlos, Waldimar Amaya, Daniel Mitrani, Valerio Pruneri, and Morgan W. Mitchell (2015), “Generation of fresh and pure random numbers for loophole-free Bell tests,” Phys. Rev. Lett. 115, 250403. Acín, Antonio, Tobias Fritz, Anthony Leverrier, and Ana Belén Sainz (2015), “A combinatorial approach to nonlocality and contextuality,” Communications in Mathematical Physics 334 (2), 533–628. Acín, Antonio, Serge Massar, and Stefano Pironio (2012), “Randomness versus nonlocality and entanglement,” Phys. Rev. Lett. 108, 100402. Acín, Antonio, Runyao Duan, David E. Roberson, Ana Belén Sainz, and Andreas Winter (2015), “A new property of the lovász number and duality relations between graph parameters,” arXiv:quantph/9803055 . Arnol’d, V I (1973), Ordinary differential equations (MIT Press, Cambridge, Massachusetts). Arnol’d, V I (1989), Mathematical methods of classical mechanics (Springer, New York). Arnon-Friedman, R, R. Renner, and T. Vidick (2016), “Simple and tight device-independent security proofs,” ArXiv e-prints arXiv:1607.01797 [quant-ph]. Aspect, Alain, Philippe Grangier, and Gérard Roger (1981), “Experimental tests of realistic local theories via bell’s theorem,” Phys. Rev. Lett. 47, 460–463. Aspect, Alain, Philippe Grangier, and Gérard Roger (1982), “Experimental realization of einstein-podolsky-rosen-bohm Gedankenexperiment : A new violation of bell’s inequalities,” Phys. Rev. Lett. 49, 91–94. Barak, B, G. Kindler, R. Shaltiel, B. Sudakov, and A. Wigderson (2010), “Simulating independence: New constructions of condensers, ramsey graphs, dispersers, and extractors,” J. ACM 57 (4), 20:1–20:52. Bartosik, H, J. Klepp, C. Schmitzer, S. Sponar, A. Cabello, H. Rauch, and Y. Hasegawa (2009), “Experimental test of quantum contextuality in neutron interferometry,” Phys. Rev. Lett. 103, 040403. Bassham, III, Lawrence E, Andrew L. Rukhin, Juan Soto, James R. Nechvatal, Miles E. Smid, Elaine B. Barker, Stefan D. Leigh, Mark Levenson, Mark Vangel, David L. Banks, Nathanael Alan Heckert, James F. Dray, and San Vo (2010), SP 800-22 Rev. 1a. A Statistical Test Suite for Random and Pseudorandom Number Generators for Cryptographic Applications, Tech. Rep. (Gaithersburg, MD, United States). Beckner, William (1975), “Inequalities in fourier analysis,” Annals of Mathematics 102 (1), 159–182. Bell, J S (1964), “On the Einstein-Podolsky-Rosen paradox,” Physics 1, 195. Bell, J S (1966), “On the problem of hidden variables in quantum mechanics,” Rev. Mod. Phys. 38, 447–452. Berta, Mario, Matthias Christandl, Roger Colbeck, Joseph M. Renes, and Renato Renner (2010), “The uncertainty principle in the presence of quantum memory,” Nat Phys 6, 659–662. Białynicki-Birula, Iwo, and Jerzy Mycielski (1975), “Uncertainty relations for information entropy in wave mechanics,” Communications in Mathematical Physics 44 (2), 129–132. Bohm, David (1951), Quantum Theory (Prentice-Hall, Inc., New York). Bohr, N (1935), “Can quantum-mechanical description of physical reality be considered complete?” Phys. Rev. 48, 696–702. Bouda, Jan, Marcin Pawłowski, Matej Pivoluska, and Martin Plesch (2014), “Device-independent randomness extraction from an arbitrarily weak min-entropy source,” Phys. Rev. A 90, 032313.

Boussinesq, M J (1878), Conciliation du véritable déterminisme mécanique avec l’existence de la vie et de la liberté morale (Mémoire de M. J. Boussinesq précédé d’un rapport á l’Académie des Sciences Morales et Politiques, par M. Paul Janet, Paris). Brandão, Fernando G S L, Ravishankar Ramanathan, Andrzej Grudka, Karol Horodecki, Michał Horodecki, Paweł Horodecki, Tomasz Szarek, and Hanna Wojewodka (2016), “Realistic noisetolerant randomness amplification using finite number of devices,” Nat Commun 7, 10.1038/ncomms11345. Brassard, Gilles, and Paul Raymond Robichaud (2013), “Can Free Will Emerge from Determinism in Quantum Theory?” in Is Science Compatible with Free Will? Exploring Free Will and Consciousness in the Light of Quantum Physics and Neuroscience, Chap. 4 (Springer, New York) pp. 41–61. Bricmont, J (1995), “Science of chaos or chaos in science?” Annals of the New York Academy of Sciences 775 (1), 131–175. Brown, Robert G (2004), “Dieharder: A random number test suite,” http://www.phy.duke.edu/∼rgb/General/dieharder.php . Brunner, Nicolas (2014), “Device-independent quantum information processing,” in Research in Optical Sciences (Optical Society of America) p. QW3A.2. Brunner, Nicolas, Daniel Cavalcanti, Stefano Pironio, Valerio Scarani, and Stephanie Wehner (2014), “Bell nonlocality,” Rev. Mod. Phys. 86, 419–478. Bub, Jeffrey (1999), Interpreting the Quantum World (Cambridge University Press). Cabello, Adán (2008), “Experimentally testable state-independent quantum contextuality,” Phys. Rev. Lett. 101, 210401. Cabello, Adán, Simone Severini, and Andreas Winter (2014), “Graph-theoretic approach to quantum correlations,” Phys. Rev. Lett. 112, 040401. Chor, Benny, and Oded Goldreich (1988), “Unbiased bits from sources of weak randomness and probabilistic communication complexity,” SIAM Journal on Computing 17 (2), 230–261. Christensen, B G, K. T. McCusker, J. B. Altepeter, B. Calkins, T. Gerrits, A. E. Lita, A. Miller, L. K. Shalm, Y. Zhang, S. W. Nam, N. Brunner, C. C. W. Lim, N. Gisin, and P. G. Kwiat (2013), “Detection-loophole-free test of quantum nonlocality, and applications,” Phys. Rev. Lett. 111, 130406. Chung, Kai-Min, Yaoyun Shi, and Xiaodi Wu (2014), “Physical randomness extractors: Generating random numbers with minimal assumptions,” arXiv:1402.4797. Cicero, Marcus Tullius (1933), De Natura Deorum, English translation by H. Rackham (Harvard University Press, Cambridge, Massachusetts). Clauser, John F, Michael A. Horne, Abner Shimony, and Richard A. Holt (1969), “Proposed experiment to test local hidden-variable theories,” Phys. Rev. Lett. 23, 880–884. Cohen-Tannoudji, Claude, Bernard Diu, and Frank Laloe (1991), Quantum Mechanics, Vol. 1 (Wiley). Colbeck, R (2007), Quantum and Relativistic Protocols for Secure Multiparty Computation, PhD Thesis (University of Cambridge). Colbeck, Roger, and Adrian Kent (2011), “Private randomness expansion with untrusted devices,” Journal of Physics A: Mathematical and Theoretical 44 (9), 095305. Colbeck, Roger, and Renato Renner (2012), “Free randomness can be amplified,” Nature Physics 8, 450. Coles, Patrick J, Mario Berta, Marco Tomamichel, and Stephanie Wehner (2015), “Entropic uncertainty relations and their applications,” arXiv:1511.04857 . Coudron, M, and H. Yuen (2013), “Infinite Randomness Expansion and Amplification with a Constant Number of Devices,” ArXiv e-prints arXiv:1310.6755 [quant-ph].

22 Deutsch, David (1983), “Uncertainty in quantum measurements,” Phys. Rev. Lett. 50, 631–633. Diels, H (1906), Die fragmente der Vorsokratiker griechisch und deutsch (Weidmannsche Buchhandlung). Dodis, Yevgeniy, Ariel Elbaz, Roberto Oliveira, and Ran Raz (2004), “Approximation, randomization, and combinatorial optimization. algorithms and techniques: 7th international workshop on approximation algorithms for combinatorial optimization problems, APPROX 2004, and 8th international workshop on randomization and computation, RANDOM 2004, Cambridge, MA, USA, August 22-24, 2004. Proceedings,” Chap. Improved Randomness Extraction from Two Independent Sources (Springer Berlin Heidelberg, Berlin, Heidelberg) pp. 334–344. Einstein, A, B. Podolsky, and N. Rosen (1935), “Can quantummechanical description of physical reality be considered complete?” Phys. Rev. 47, 777–780. Fine, Arthur (1982), “Hidden variables, joint probability, and the bell inequalities,” Phys. Rev. Lett. 48, 291–295. Fletcher, Samuel Craig (2012), “What counts as a newtonian system? the view from norton’s dome,” European Journal for Philosophy of Science 2 (3), 275–297. Freedman, Stuart J, and John F. Clauser (1972), “Experimental test of local hidden-variable theories,” Phys. Rev. Lett. 28, 938–941. Freeman, K (1948), Ancilla to the pre-Socratic philosophers (Forgotten Books, Cambridge, Massachusetts). Gabizon, Ariel, and Ronen Shaltiel (2008), “Approximation, randomization and combinatorial optimization. algorithms and techniques: 11th international workshop, approx 2008, and 12th international workshop, random 2008, boston, ma, usa, august 25-27, 2008. proceedings,” Chap. Increasing the Output Length of ZeroError Dispersers (Springer Berlin Heidelberg, Berlin, Heidelberg) pp. 430–443. Gallego, R, L. E. Würflinger, A. Ací n, and M. Navascués (2012), “An operational framework for nonlocality,” Phys. Rev. Lett. 109, 070401. Gallego, Rodrigo, Lluis Masanes, Gonzalo De La Torre, Chirag Dhara, Leandro Aolita, and Antonio Acín (2013), “Full randomness from arbitrarily deterministic events,” Nat Commun 4, 10.1038/ncomms3654. Gisin, Nicolas (2013), “Are There Quantum Effects Coming from Outside Space–Time? Nonlocality, Free Will and “No ManyWorlds”,” in Is Science Compatible with Free Will? Exploring Free Will and Consciousness in the Light of Quantum Physics and Neuroscience, Chap. 3 (Springer, New York) pp. 23–40. Giustina, Marissa, Marijn A. M. Versteegh, Sören Wengerowsky, Johannes Handsteiner, Armin Hochrainer, Kevin Phelan, Fabian Steinlechner, Johannes Kofler, Jan-Åke Larsson, Carlos Abellán, Waldimar Amaya, Valerio Pruneri, Morgan W. Mitchell, Jörn Beyer, Thomas Gerrits, Adriana E. Lita, Lynden K. Shalm, Sae Woo Nam, Thomas Scheidl, Rupert Ursin, Bernhard Wittmann, and Anton Zeilinger (2015), “Significant-loopholefree test of bell’s theorem with entangled photons,” Phys. Rev. Lett. 115, 250401. Gleason, Andrew M (1975), “The logico-algebraic approach to quantum mechanics: Volume i: Historical evolution,” Chap. Measures on the Closed Subspaces of a Hilbert Space (Springer Netherlands, Dordrecht) pp. 123–133. Gleick, James (2008), Chaos: Making a New Science (Penguin Books). Grudka, Andrzej, Karol Horodecki, Michał Horodecki, Paweł Horodecki, Marcin Pawłowski, and Ravishankar Ramanathan (2014), “Free randomness amplification using bipartite chain correlations,” Phys. Rev. A 90, 032322.

Gudder, Stanley P (1970), “On Hidden-Variable Theories,” Journal of Mathematical Physics 11 (2), 431–436. Hall, Michael J W (2010), “Local deterministic model of singlet state correlations based on relaxing measurement independence,” Phys. Rev. Lett. 105, 250404. Halmos, Paul R (2013), Lectures on Ergodic Theory (Martino Fine Books). Heisenberg, W (1927), “Über den anschaulichen inhalt der quantentheoretischen kinematik und mechanik,” Zeitschrift für Physik 43 (3), 172–198. Hensen, B, H. Bernien, A. E. Dreau, A. Reiserer, N. Kalb, M. S. Blok, J. Ruitenberg, R. F. L. Vermeulen, R. N. Schouten, C. Abellan, W. Amaya, V. Pruneri, M. W. Mitchell, M. Markham, D. J. Twitchen, D. Elkouss, S. Wehner, T. H. Taminiau, and R. Hanson (2015), “Loophole-free bell inequality violation using electron spins separated by 1.3 kilometres,” Nature 526, 682–686. Herrero-Collantes, M, and J. C. Garcia-Escartin (2016), “Quantum Random Number Generators,” ArXiv e-prints arXiv:1604.03304 [quant-ph]. Isham, C J, and J. Butterfield (1998), “A topos perspective on the kochen-specker theorem: I. quantum states as generalized valuations,” arXiv:quant-ph/9803055 . Isida, Masatugu, and Hiroji Ikeda (1956), “Random number generator,” Annals of the Institute of Statistical Mathematics 8 (1), 119–126. Ivancevic, Vladimir G, and Tijana T. Ivancevic (2008), Complex nonlinearity: chaos, phase transitions, topology change, and path integrals (Springer). Jerger, M, Y. Reshitnyk, M. Oppliger, A. Potoˇcnik, M. Mondal, A. Wallraff, K. Goodenough, S. Wehner, K. Juliusson, N. K. Langford, and A. Fedorov (2016), “Contextuality without nonlocality in a superconducting quantum system,” Nature Communications 7, 12930. Jofre, M, M. Curty, F. Steinlechner, G. Anzolin, J. P. Torres, M. W. Mitchell, and V. Pruneri (2011), “True random numbers from amplified quantum vacuum,” Opt. Express 19 (21), 20665–20672. Kamp, Jesse, Anup Rao, Salil Vadhan, and David Zuckerman (2011), “Deterministic extractors for small-space sources,” Journal of Computer and System Sciences 77 (1), 191 – 220, celebrating Karp’s Kyoto Prize. Khinchin, A (2014), Mathematical Foundations of Statistical Mechanics (Martino Fine Books). Kirchmair, G, F. Zahringer, R. Gerritsma, M. Kleinmann, O. Guhne, A. Cabello, R. Blatt, and C. F. Roos (2009), “State-independent experimental test of quantum contextuality,” Nature 460, 494– 497. Kochen, Simon, and E. P. Specker (1967), “The problem of hidden variables in quantum mechanics,” Journal of Mathematics and Mechanics 17, 59–87. Koh, Dax Enshan, Michael J. W. Hall, Setiawan, James E. Pope, Chiara Marletto, Alastair Kay, Valerio Scarani, and Artur Ekert (2012), “Effects of reduced measurement independence on bellbased randomness expansion,” Phys. Rev. Lett. 109, 160404. Korolev, Alexandre (2007a), “Indeterminism, asymptotic reasoning, and time irreversibility in classical physics,” Philosophy of Science 74, 943–956. Korolev, Alexandre (2007b), “The norton-Type lipschitzindeterministic systems and elastic phenomena: Indeterminism as an artefact of infinite idealizations,” (Philosophy of Science Assoc. 21st Biennial Mtg., Pittsburgh, PA). Kosyakov, B P (2008), “Is classical reality completely deterministic?” Foundations of Physics 38 (1), 76–88. Laertius, Diogenes (1925), Lives of Eminent Philosophers, translated by RD Hicks, Vol. 2. Loeb Classical Library, no. 185 (Har-

23 vard University Press, Cambridge, Massachusetts). Landau, L D, and E. M. Lifshitz (1960), Classical mechanics (Pergamon Press, Oxford). Laplace, P S (1814), Essai philosophique sur les probabilités (Courcier, Paris). Laplace, P S (1951), A Philosophical Essay on Probabilities (translated into English from the original French 6th ed. by Truscott, F. W. and Emory, F. L., (Dover Publications, New York, 1951)). Laraudogoitia, Jon Pérez (2013), “On norton’s dome,” Synthese 190 (14), 2925–2941. L’Ecuyer, Pierre, and Richard Simard (2007), “TestU01: A c library for empirical testing of random number generators,” ACM Trans. Math. Softw. 33 (4), 22. Liang, Yeong-Cherng, Denis Rosset, Jean-Daniel Bancal, Gilles Pütz, Tomer Jack Barnea, and Nicolas Gisin (2015), “Family of bell-like inequalities as device-independent witnesses for entanglement depth,” Phys. Rev. Lett. 114, 190401. Lydersen, Lars, Carlos Wiechers, Christoffer Wittmann, Dominique Elser, Johannes Skaar, and Vadim Makarov (2010), “Hacking commercial quantum cryptography systems by tailored bright illumination,” Nat Photon 4, 686–689. Maassen, Hans, and J. B. M. Uffink (1988), “Generalized entropic uncertainty relations,” Phys. Rev. Lett. 60, 1103–1106. Maccone, Lorenzo, and Arun K. Pati (2014), “Stronger uncertainty relations for all incompatible observables,” Phys. Rev. Lett. 113, 260401. Malament, David B (2008), “Norton’s slippery slope,” Philosophy of Science 75 (5), 799–816. Marsaglia, G (2008), The Marsaglia Random Number CDROM including the Diehard Battery of Tests of Randomness. Menezes, Alfred J, Paul C. van Oorschot, and Scott A. Vanstone (1996), Handbook of Applied Cryptography (CRC Press, Boca Raton, FL, USA). Messiah, Albert (2014), Quantum Mechnaics (Dover Publications). Miller, C A, and Y. Shi (2014), “Universal security for randomness expansion from the spot-checking protocol,” ArXiv e-prints arXiv:1411.6608 [quant-ph]. Miller, Carl A, and Yaoyun Shi (2016), “Robust protocols for securely expanding randomness and distributing keys using untrusted quantum devices,” J. ACM 63 (4), 33:1–33:63. Mironowicz, Piotr, Rodrigo Gallego, and Marcin Pawłowski (2015), “Robust amplification of santha-vazirani sources with three devices,” Phys. Rev. A 91, 032317. Mitchell, Morgan W, Carlos Abellan, and Waldimar Amaya (2015), “Strong experimental guarantees in ultrafast quantum random number generation,” Phys. Rev. A 91, 012314. Motwani, Rajeev, and Prabhakar Raghavan (1995), Randomized Algorithms (Cambridge University Press, New York, NY, USA). Myerson, A H, D. J. Szwer, S. C. Webster, D. T. C. Allcock, M. J. Curtis, G. Imreh, J. A. Sherman, D. N. Stacey, A. M. Steane, and D. M. Lucas (2008), “High-fidelity readout of trapped-ion qubits,” Phys. Rev. Lett. 100, 200502. National Institutes of Standards and Technology, (2011), “NIST Randomness Beacon,” http://www.nist.gov/itl/csd/ct/nist_beacon.cfm. Navascués, Miguel, Stefano Pironio, and Antonio Acín (2008), “A convergent hierarchy of semidefinite programs characterizing the set of quantum correlations,” New Journal of Physics 10 (7), 073013. von Neumann, John (1951), “Various techniques used in connection with random digits,” Applied Math Series 12, 36. Neumann, John Von (1955), Mathematical Foundations of Quantum Mechanics (Princeton University Press).

Newman, J R (1956), The world of mathematics (Simon and Schuster, New York). Nie, You-Qi, Leilei Huang, Yang Liu, Frank Payne, Jun Zhang, and Jian-Wei Pan (2015), “The generation of 68 gbps quantum random number by measuring laser phase fluctuations,” Review of Scientific Instruments 86 (6), 063105. Nisan, Noam, and Amnon Ta-Shma (1999), “Extracting randomness: A survey and new constructions,” Journal of Computer and System Sciences 58 (1), 148 – 173. Norton, J D (2007), “Causation, physics, and the constitution of reality: Russell’s republic revisited,” Chap. Causation as folk science (Oxford University Press, Oxford) pp. 11–44. Norton, J D (2008), “The dome: An unexpectedly simple failure of determinism,” Philosophy of Science 75 (5), 786–798. Penrose, O (1979), “Foundations of statistical mechanics,” Reports on Progress in Physics 42 (12), 1937. Peres, A (1995), Quantum Theory: Concepts and Methods (Springer Netherlands). Pironio, S (2015), “Random ’choices’ and the locality loophole,” ArXiv e-prints arXiv:1510.00248 [quant-ph]. Pironio, S, A. Acín, S. Massar, A. Boyer de la Giroday, D. N. Matsukevich, P. Maunz, S. Olmschenk, D. Hayes, L. Luo, T. A. Manning, and C. Monroe (2010), “Random numbers certified by bell’s theorem,” Nature 23, 1021–1024. Pironio, Stefano, Valerio Scarani, and Thomas Vidick (2015), “Focus on device independent quantum information,” New Journal of Physics . Pivoluska, Matej, and Martin Plesch (2014), “Device independent random number generation,” Acta Physica Slovaca 64, 600 – 663. Poincaré, H (1912), Calcul des probabilités (Gauthier-Villars, Paris). Popescu, Sandu, and Daniel Rohrlich (1994), “Quantum nonlocality as an axiom,” Foundations of Physics 24 (3), 379–385. Popper, K R (1982), The Open Universe. An Argument for Indeterminism (Routledge, London and New York). Raz, Ran (2005), “Extractors with weak random seeds,” in Proceedings of the Thirty-seventh Annual ACM Symposium on Theory of Computing, STOC ’05 (ACM, New York, NY, USA) pp. 11–20. Roberts, B W (2009), “Wilson’s case against the dome: Not necessary, not sufficient,” Unpublished manuscript. Robertson, H P (1929), “The uncertainty principle,” Phys. Rev. 34, 163–164. Rosenfeld, W, J. Hofmann, N. Ortegel, M. Krug, F. Henkel, Ch. Kurtsiefer, M. Weber, and H. Weinfurter (2011), “Towards highfidelity interference of photons emitted by two remotely trapped rb-87 atoms,” Optics and Spectroscopy 111 (4), 535–539. Rosicka, M, R Ramanathan, P Gnaci´nski, K Horodecki, M Horodecki, P Horodecki, and S Severini (2016), “Linear game non-contextuality and bell inequalities—a graph-theoretic approach,” New Journal of Physics 18 (4), 045020. Rukhin, Andrew, Juan Soto, James Nechvatal, Miles Smid, Elaine Barker, Stefan Leigh, Mark Levenson, Mark Vangel, David Banks, Alan Heckert, James Dray, and San Vo (2010), A Statistical Test Suite for Random and Pseudorandom Number Generators for Cryptographic Applications, Tech. Rep. 800-22 (National Institute of Standards and Technology). Santha, Miklos, and Umesh V. Vazirani (1986), “Generating quasirandom sequences from semi-random sources,” Journal of Computer and System Sciences 33 (1), 75 – 87. Schawlow, A L, and C. H. Townes (1958), “Infrared and optical masers,” Phys. Rev. 112, 1940–1949. Schrödinger, E (1930), “About Heisenberg Uncertainty Relation,” in Proc. Prussian Acad. Sci., Phys. Math. Section, Vol. XIX, p. 293. Schrödinger, Erwin (1989), Statistical Thermodynamics (Dover Publications).

24 Scully, Marlan O, and K. Wódkiewicz (1989), “Coherence and Quantum Optics VI: Proceedings of the Sixth Rochester Conference on Coherence and Quantum Optics held at the University of Rochester, June 26–28, 1989,” Chap. On the Quantum Malus Law for Photon and Spin Quantum Correlations (Springer US, Boston, MA) pp. 1047–1050. Shalm, Lynden K, Evan Meyer-Scott, Bradley G. Christensen, Peter Bierhorst, Michael A. Wayne, Martin J. Stevens, Thomas Gerrits, Scott Glancy, Deny R. Hamel, Michael S. Allman, Kevin J. Coakley, Shellee D. Dyer, Carson Hodge, Adriana E. Lita, Varun B. Verma, Camilla Lambrocco, Edward Tortorici, Alan L. Migdall, Yanbao Zhang, Daniel R. Kumor, William H. Farr, Francesco Marsili, Matthew D. Shaw, Jeffrey A. Stern, Carlos Abellán, Waldimar Amaya, Valerio Pruneri, Thomas Jennewein, Morgan W. Mitchell, Paul G. Kwiat, Joshua C. Bienfang, Richard P. Mirin, Emanuel Knill, and Sae Woo Nam (2015), “Strong loophole-free test of local realism*,” Phys. Rev. Lett. 115, 250402. Shaltiel, Ronen (2002), “Recent developments in explicit constructions of extractors,” Bulletin of the EATCS 77, 67–95. Smoluchowski, M V (1918), “Über den Begriff des Zufalls und den Ursprung der Wahrscheinlichkeitsgesetze in der Physik,” Naturwissenschaften 6, 253–263. Suarez, Antoine (2013), “Free Will and Nonlocality at Detection as Basic Principles of Quantum Physics,” in Is Science Compatible with Free Will? Exploring Free Will and Consciousness in the Light of Quantum Physics and Neuroscience, Chap. 4 (Springer, New York) pp. 63–79. Suarez, Antoine, and Peter Adams, Eds. (2013), Is Science Compatible with Free Will? Exploring Free Will and Consciousness in the Light of Quantum Physics and Neuroscience (Springer, New York). Tolman, Richard C (2010), The Principles of Statistical Mechanics (Dover Books on Physics) (Dover Publications). de la Torre, Gonzalo, Matty J. Hoban, Chirag Dhara, Giuseppe Prettico, and Antonio Acín (2015), “Maximally nonlocal theories cannot be maximally random,” Phys. Rev. Lett. 114, 160502. Trevisan, L (2001), “Extractors and pseudorandom generators,” J. ACM 48 (4), 860–879. Tura, J, R. Augusiak, A. B. Sainz, T. Vértesi, M. Lewenstein, and A. Acín (2014a), “Detecting nonlocality in many-body quantum states,” Science 344 (6189), 1256–1258. Tura, J, R. Augusiak, A.B. Sainz, B. Lücke, C. Klempt, M. Lewenstein, and A. Acín (2015), “Nonlocality in many-body quantum systems detected with two-body correlators,” Annals of Physics 362, 370 – 423. Tura, J, A. B. Sainz, T. Vértesi, A. Acín, M. Lewenstein, and R. Augusiak (2014b), “Translationally invariant multipartite bell

inequalities involving only two-body correlators,” Journal of Physics A: Mathematical and Theoretical 47 (42), 424024. Tylec, T I, and M. Ku´s (2015), “Non-signaling boxes and quantum logics,” Journal of Physics A: Mathematical and Theoretical 48 (50), 505303. Vazirani, Umesh, and Thomas Vidick (2012), “Certifiable quantum dice: Or, true random number generation secure against quantum adversaries,” in Proceedings of the Forty-fourth Annual ACM Symposium on Theory of Computing, STOC ’12 (ACM, New York, NY, USA) pp. 61–76. Vazirani, Umesh V (1987), “Strong communication complexity or generating quasirandom sequences form two communicating semi-random sources,” Combinatorica 7 (4), 375–392. Vicente, Julio I de (2014), “On nonlocality as a resource theory and nonlocality measures,” Journal of Physics A: Mathematical and Theoretical 47 (42), 424017. Wheeler, John Archibald, and Wojciech Hubert Zurek (1983), Quantum Theory and Measurement (Princeton University Press). Wilson, Mark (2009), “Determinism and the mystery of the missing physics,” The British Journal for the Philosophy of Science 60 (1), 173–193. Wódkiewicz, Krzysztof (1995), “Classical and quantum malus laws,” Phys. Rev. A 51, 2785–2788. Wojewódka, H, F. G. S. L. Brandão, A. Grudka, M. Horodecki, K. Horodecki, Horodecki, M. Pwł owski, and R. Ramanathan (2016), “Amplifying the randomness of weak sources correlated with devices,” ArXiv e-prints arXiv:1601.06455 [quant-ph]. Wódkiewicz, K (1985), “Quantum malu’s law,” Physics Letters A 112 (6), 276–278. Xu, Feihu, Bing Qi, Xiongfeng Ma, He Xu, Haoxuan Zheng, and Hoi-Kwong Lo (2012), “Ultrafast quantum random number generation based on quantum phase fluctuations,” Opt. Express 20 (11), 12366–12377. Yuan, Z L, M. Lucamarini, J. F. Dynes, B. Fröhlich, A. Plews, and A. J. Shields (2014), “Robust random number generation using steady-state emission of gain-switched laser diodes,” Applied Physics Letters 104 (26), 261112, http://dx.doi.org/10.1063/1.4886761. Zinkernagel, Henrik (2010), “EPSA Philosophical Issues in the Sciences: Launch of the European Philosophy of Science Association,” Chap. Causal Fundamentalism in Physics (Springer Netherlands, Dordrecht) pp. 311–322. Zurek, Wojciech Hubert (2003), “Decoherence, einselection, and the quantum origins of the classical,” Rev. Mod. Phys. 75, 715–775. Zurek, Wojciech Hubert (2009), “Quantum darwinism,” Nat Phys 5, 181–188.