arXiv:1803.06824v3 [quant-ph] 31 May 2019 fpitprils-gvnb elnmes xetfrpar- for Except numbers. cases real by ticular given - particles momentum point and of position typically - conditions initial with oee tlati rnil ysinehpesby happens science including by processes. big-bang, - physiological the since all principle determined is in everything i.e. least necessity, that at conclude then to explanation, - tempting scientific covered highly good be only would it the determin- scientific be if Bohmian would Second, discuss ism context. I this VIII in and quantum mechanics recover VII despite to sections as in determinism, way randomness; to classical a try like such physicists something in explanations. and physics scientific quantum philosophers good formulate all many of example, goal For the the as Hence, taken ten times. deterministic. past is and physics classical future that all conclusion at uniquely solutions and completely the determine conditions initial the o cin e ls ociia ausdtrie ythe by determined values critical constants to universal size close and theory get speed action) quantum when well (or accurate and classical more are relativity are of limits predictions validity by whose its of limited that domain is the fact mechanics impressive, Indeed, the truly is understood. because?) power (or explanatory despite physics, whose classical that by theory follow illustrated a must usually is explanations of- This good is structure. all it that Furthermore, claimed world. ten our of explanation ministic 1 ie .. otnsdm 1 n rna olsoso poin of collisions frontal and [1] ticles. dome Norton’s e.g., Like, lsia ehnc sasto yaia equations, dynamical of set a is mechanics Classical nm pno u hsppri needn fthis of independent is paper this but - opinion my In of- is this said, as First, consequences. huge has This hsc sotnpeetda h xml fadeter- a of example the as presented often is Physics neemns nPyis lsia ho n oma Mech Bohmian and Chaos Classical Physics, in Indeterminism 1 eetBhinmcaisa upeetdqatmter,a theory, reality. p quantum physical attribute supplemented no they as phy which mechanics Most to Bohmian numbers reject deterministic. real is with supplement theory theory be t classical supplemented quantum can the to theories that similar quantum way way and a mechanics in me classical equations, classical deterministic alternative of use This equ series numbers. empirically am their is finite-information finite which as mechanics, a numbers”, classical “random than relevant.alternative is more physically numbers not contain infinit real are can’t so-called an numbers space real contain of mathematical numbers volume the real finite all a almost that However, numbers. hs yaia qain oehrwith together equations dynamical these , ti sa oietf nta odtoso lsia dyna classical of conditions initial identify to usual is It .INTRODUCTION I. c ru fApidPyis nvriyo eea 21Geneva 1211 Geneva, of University Physics, Applied of Group n ¯ and h respectively. , r elNmesRal Real? Really Numbers Real Are Dtd ue3 2019) 3, June (Dated: par- t ioa Gisin Nicolas qal uprsa miial qiaetbtindeter- theory. mechanics but classical evidence equivalent alternative of empirically de- ministic body an a same supports the supports equally since ex- mechanics view, the world classical and terministic of evidence empirical power the not planatory that thus claim is It to same VI. correct the section has inde- power, theory explanatory alternative alternative (enormous) the The to theory. applies clas- terministic for equally section evidence mechanics (see empirical sical information huge can of the space with Consequently, amount of IV). numbers finite volume a to finite only any physics contain that limiting is of information favour way finite One in information. argue inaccessible as- to of doesn’t existence it the alternative al- because this sume natural is that more argue practice is I mechanics in III-V classical used sections In make mathematics not finite. the ways do numbers, that nor I notice time space, but information. about of contain- claims amount metaphysical numbers finite any by a given includ- only are ing parameters, conditions all initial equa- but the dynamical mechanics, ing same classical the as keeps tions theory alternative this aibe nsc a orne tdtriitc(in deterministic it render to way a additional such by in supplemented variables be can theory deterministic c,wt rcsl h aesto rdcin,though predictions, of indeterministic set is same theory alternative the this precisely mechan- classical with from ics, different but harmony similar theory, greater other a is spectator. [2]. there any experience then without our valid, box and physics is closed between a paper in this movie If a would world like our be consequences: dreadful has this - opinion 2 fntr,teei oeta n osbefuture. possible one than more is the there and nature, present determinist the of of given negation non-deterministic: to the synonymous merely is indeterministic Here ntescn ato hspprIageta vr in- every that argue I paper this of part second the In ntefis ato hspprIageta hr san- is there that argue I paper this of part first the In vln ocasclmcais u ssonly uses but mechanics, classical to ivalent hnc snndtriitc ept the despite non-deterministic, is chanics yia xsec,wiems physicists most while existence, hysical oevr etrtriooyfrthe for terminology better a Moreover, real mathematical with systems mical fbt r rl adm rps an propose I random. truly are bits of er.Itrsigy ohalternative both Interestingly, heory. db diinlvralsi uha such in variables additional by ed iit tagtowrl supplement straightforwardly sicists gigta oma oiin have positions Bohmian that rguing muto nomto.Iargue I information. of amount e uto nomto,hnethat hence information, of ount ,Switzerland 4, anics. 2 nanutshell, a In . c i.e. ic, laws 2 much the same way as is done by Bohmian mechanics). However, the fact is that almost all classical dynami- In brief, it suffices to assume that all the indeterminism cal systems are not integrable, they are not simple, but that is required at some point in time when, according on the contrary are chaotic. In this paper, for clarity, I to the indeterministic theory, God plays dice, i.e. when consider one typical chaotic dynamical system, but it is potentialities becoming actual, could be hidden as sup- important to realize that all non-simple classical dynam- plementary variables in the initial condition of the equiv- ical systems share the essential features of our example. alent deterministic theory, i.e. God played all dice at In this example, we don’t consider the solution at all the big-bang. This closes the circle: deterministic theo- times, but only at a discrete set of times, let’s say ev- ries are equivalent to indeterministic alternative theories ery microsecond. Furthermore, we assume the system is in which real numbers are replaced by finite-information constrained to remain within the unit interval [0..1], i.e. numbers3, and indeterministic theories can be supple- its coordinate x lies between 0 and 1. Accordingly, its mented by additional hidden variables in such a way that coordinate can be written in binary form as a number the supplemented theories are deterministic. like: In sections VII and VIII the above rule to supplement indeterministic theories is illustrated on the alternative x =0.b1b2b3...bn... (1) classical mechanics theory and on standard quantum theory, leading to standard classical mechanics and to where the bj’s are the bits of x in binary representation Bohmian mechanics, respectively. Admittedly, in these (equivalent to the digits in base 10). The dynamics for two examples, the supplemented deterministic theories each time step of this example is given by the following have, in addition to determinism, some elegance which map: speaks in their favour. However, one may conclude that 1 2x ifx< 2 determinism is too high a price to pay to accept these x 1 (2) →  2x 1 if x 2 supplementary hidden variables. Indeed, indeterminism − ≥ explains nicely, among other things, why probabilistic Such a dynamical map is very simple to represent when tools are so powerful in statistical mechanics [3]. More- the coordinate x is represented in binary form: over, indeterminism opens the future, makes potentiali- ties a real mode of existence and describes the passage of x = 0.b1b2b3...bn... time when potentialities become actual [4, 5]. 0.b2b3...bn... (3) → At each time-step the bits merely get shifted to the left II. CLASSICAL DYNAMICAL SYSTEMS by one place and the initial leading bit b1 drops out. Af- ter n time-steps, the bits shift by n steps to the left. The simplest and thus best known classical dynamical This example of a generic chaotic system is inspired by systems are clocks, harmonic oscillators, two bodies in- the baker’s map [6], though in our example there is a dis- 1 teracting via gravity (e.g. one lonely planet orbiting its continuity at x = 2 . Note however, that the continuous sun) and similar systems. For such systems, the trajecto- baker’s map shares all features of our map essential for ries are ellipses4 (in ordinary or in configuration space), our arguments. including the cases of degenerate ellipses, i.e. circles and Notice that whether the x in eq. (1) lies in the first straight lines5. Such simple dynamical systems are called half of the unit interval or in the second half is entirely integrable. They are characterized by their stability: the determined by the leading bit b1. This has the following solution at any time depends only on the leading digits important consequence: whether the system lies within of the initial condition. More precisely, the solution up the first half or the second half of the unit interval af- to any precision ǫ depends only on the initial condition ter n time-steps is determined by the first bit after n up to a precision ǫ. Hence, for such simple systems, the time-steps, hence it depends on the nth bit of the initial far away digits, let’s say from the billionth digits on, are condition. physically irrelevant, i.e. don’t represent anything phys- Such a chaotic system illustrates, for example, the chal- ical; rhetorically, I sometimes write that these far away lenge of weather predictions. Let’s say that when the digits have no physical existence or are not physically system’s coordinate x lies on the left of the unit interval, real. this represents rainy weather, while an x on the right- hand side represents sunny weather. Then, the weather in a week’s time depends on infinitesimal bits, e.g. the billionth bit, of the initial condition. 3 As explained in sections III and IV, see also Fig. 1, finite infor- The question here is not whether this billionth bit can mation numbers contains all computable numbers, but are not be measured, but rather whether this billionth bit has restricted to them. any physical relevance. Clearly, if the initial condition is 4 Trajectories whose coordinates are sinuses and cosins, functions that every computer “knows” how to calculate efficiently. defined by a real number, then this billionth bit is math- 5 Or the trajectories escape to infinity following parabolas or hy- ematically well defined. Hence, the question is whether perbolas. mathematical real numbers are physically real. 3

III. REAL NUMBERS ARE ARE NOT REALLY paper is that all of physics can be done with only finite- REAL information numbers. Note that these numbers contain all computable numbers, but are not restricted to them: The set of all real numbers is equivalent (isomorphic) they also contain ”numbers” whose far away digits are 7 to the set of real numbers within the unit interval. Hence, undetermined, i.e. not yet determined , as illustrated in all relevant numbers can be written as in (1), with in- Fig. 1. All numbers containing an infinite amount of finitely many bits bj. This way of writing down real information, cannot represent physical entities; specifi- numbers already illustrates the fact that real numbers cally, they cannot be used, and in fact are not used, to 8 do, in general, contain an unlimited amount of (Shan- describe initial conditions , see also [9]. Moreover, in non) information, i.e. infinitely many bits. The only practice, one never uses real numbers, except to prove exceptions are when the series of bits bj terminates, or some general abstract non-constructive existence theo- more precisely when all bits after a finite coordinate m rems. The fact that one doesn’t need real numbers in are nil: bj = 0 for all j >m, or when after a finite practice is quite obvious, as one never accesses an infi- position m the series of bits repeats itself forever, like, nite amount of information. Furthermore, today all pre- e.g., 0.0111011001010101010101010101... which contin- dictions can be - and most of the time are - encoded in ues with an endless repetition of the pattern 01, or more computers, computers that obviously hold at most a fi- generally when there is a finite formula (algorithm) to nite amount of bits, as emphasized in the next section. compute all bits. Consequently, physics is actually done using only finite- Another nice way to illustrate the infinite amount of information numbers and, as we’ll see in section VI, clas- information in typical real numbers is due to Emile Borel, sical physics with finite-information initial condition is a as nicely told by Gregory Chaitin [7]. They emphasize well defined indeterministic alternative theory to classi- that one single real number can contain the answers to all cal mechanics. Admittedly, one may prefer to postulate (binary) questions one can formulate in any human lan- that real numbers are physically significant, as I discuss guage. To see this it suffices to realize that there are only in section VIII. finitely many languages, each with finitely many symbols. Hence, one can binarize this list of symbols (as routinely done in today’s computers) and list all sequences of sym- IV. A FINITE VOLUME IN SPACE CONTAINS bols, first the sequences containing only a single symbol, AT MOST FINITE INFORMATION next those containing two symbols, and so on. This huge list of symbols can then be considered as the bits of a 1 2 Here I present an argument supporting the claim that real number. Let’s leave 2 bits, bnbn, in-between each real numbers cannot represent anything physical. This sequence Sn of symbols: argument is based on the assumption that no finite vol- 1 2 1 2 1 2 1 2 ume of space can contain an infinite amount of informa- 0.S1b1b1S2b2b2S3b3b3...Snbnbn... (4) tion. This is a well accepted result that follows from the holographic principle, known as the Bekenstein bound When the sequence Sn of symbols doesn’t represent a bi- 1 2 [10, 11]. In brief, any storage of a bit of information nary question, we set these two bits to 0 (b b = 00). n n requires some energy and large enough energy densities When they represent a question whose answer is yes, we trigger black holes. However, for the purpose of my argu- set these bits to 01 and if the answer is no we set them to 6 ment, I believe a much simpler reasoning suffice to con- 10. This procedure is not efficient at all, but who cares : vince oneself that every bit of information occupies some since a real number has infinitely many bits, there is no space, hence that information density is limited. Let me need to save space! Hence, one can really code the an- now present this reasoning based on an intuitive assump- swers to all possible (binary) questions in one single real tion. number. This illustrates the absurdly unlimited amount of information that real numbers contain. Real numbers The enormous progress in information storage over the last few decades profoundly impacts our society9. Today, are monsters! In the next section I argue that a finite volume of space can contain no more than a finite amount of informa- tion. Following this reasonable assumption, I argue that 7 This is somewhat reminiscent of Brouwer’s indeterminate num- the so-called real numbers are not really real. More pre- bers [8]. cisely, I argue that the mathematical real numbers are 8 To be clear - I am not making claims pertaining to the nature of not physically real, by which I mean that they do not numbers, or making assumptions regarding the reality or unreal- represent anything physical. Indeed, the thesis of this ity of numbers. My concern is with distinguishing numbers that have physical significance from those that do not. Nor am I at all concerned with the well known and daunting task of accounting for the applicability of math in science. Finally, I am not argu- ing against the use of real numbers as a useful tool for calculus, 6 An already more economical coding would be to ignore all se- simply only finite-information number can represent something quences of symbols that do not represent any question and add physical. after each meaningful question a single bit coding for the answer. 9 Allow me a side remark. The enormous progress in information 4 everyone knows that they hold gigabytes of information V. MATHEMATICAL REAL NUMBERS ARE in their pockets and that companies like Google and agen- PHYSICAL RANDOM NUMBERS cies like the NSA store everything that transits through the internet. Furthermore, everyone knows also that each Some real numbers can be computed up to arbitrary stored bit requires some space. Not much, possibly soon −18 3 precision with a computer, like for example all rational only a few cubic nanometre (10 mm ), but defini- numbers and numbers like √2 and π. Such computable tively some finite volume. Consequently, assuming that numbers contain only finite information, the length in information has always to be encoded in some physical bits of the shortest program that outputs their bits. Note stuff, a finite volume of space cannot contain more than that since there are only countably many programs, the a finite amount of information. At least, this is a very set of real numbers that can be calculated by a computer reasonable assumption. is infinitely smaller than the set of all real numbers. More precisely, the set of computable numbers is countably- Here I would like to explore the deep conceptual con- infinite, like the integers, while the set of real numbers is sequences of this assumption, an assumption easy to for- continuously-infinite, like the mathematical points on a mulate and defend in our information-based society, but line. Consequently, real numbers are uncomputable with an assumption hardly conceivable a century ago. Re- probability one. Or, equivalently, the set of computable member indeed, that the modern concept of information numbers has measure zero among the set of all real num- was formalized by Shannon only in the 1940’s. bers. For more see [7].

Consider a small volume, a cubic centimetre let’s say, containing a marble ball. This small volume can contain but a finite amount of information. Hence, the centre of mass of this marble ball can’t be a real number (and even less 3 real numbers), since real numbers contain - with probability one - an infinite amount of information. Clas- sical physics describes the centre of mass of the ball by 3 real numbers; and this is an extremely efficient descrip- tion. But the assumption that a finite volume of space FIG. 1: The series of bits of typical of real numbers is, after can’t contain more but a finite amount of information some initial finite bit string, indistinguishable from a truly implies that the centre of mass of any object cannot be random sequence of bits, as produce by a physical random identified with mathematical real numbers. Real num- number generator, here illustrated by a quantum random num- bers are useful tools, but are only tools. They do not ber generator (single photons on a beam-splitter followed by represent physical reality. two single-photon detectors). Hence, there are two ways one may think of typical real numbers. First, one may think of them as dynamical quantities: the far away bits don’t have Admittedly, according to today’s physics the above ar- yet any determined value, the values are continuously pro- gument is a bit misleading, since we know that, ulti- duced as time passes. Second, one may think of real numbers mately, the marble ball and its centre of mass should be with all their bits given at ones, as if their random values described by quantum physics, including quantum inde- were produced at some initial time. The second view is the terminacy (often called uncertainty). This is correct, of standard view, while the first one is the view advocated here. course. But let’s continue with classical physics because, Note that computable numbers correspond to pseudo-random first, it remains extremely useful today, and, secondly, number generators where all the information lies in the finite it is often presented as the archetype of deterministic seed. The finite information numbers used here include both the computational numbers and the dynamical real numbers theories. The main point of this paper is that classical whose far away bits are not yet determined. physics is deterministic only if one attributes to the tool of real numbers physical significance. As soon as one real- izes that the mathematical real numbers are “not really The above simple observation has the following impor- real”, i.e. have no physical significance, then one con- tant consequence: after the first bits, the next bits of cludes that classical physics is not deterministic, as we almost all real numbers are random: they don’t follow elaborate in section VI. Actually, things are even worse, any structure. These bits are as random as the outcome as we explain in the next section. of quantum measurements (on half a singlet, let’s say), i.e. they are as random as possible10 [7]. Actually, one

storage and the relatively poor progress in energy storage ex- plains why the science fiction of half a century ago completely 10 Some caution is due here, as not all bits of all real numbers are missed the “internet revolution”. In the science fiction of those random, as illustrated by the following example. Define a num- days no one could hold an encyclopedia in his pocket, but every- ber with all bits at even positions identical to the corresponding one was flying thanks to small backpacks. bits of the computable number π and all bits at odd positions 5 can’t even name or characterize real numbers, as there ness has nothing to do with technological limitations, it are only countably many names and characterizations. is intrinsic pure randomness. Hence, almost all real numbers are totally outside our The view I am suggesting is that the first bits in the grasp: we can’t say anything about them, except that expression of x are “really real” (e.g., at present, it is re- their digits are random, they have no structure. Indeed, ally rather sunny or rather rainy), while the very far away if the digits of a real number had some structure, this bits are totally random. As time passes they get shifted very structure would allow one to characterise and name to the left, one position at each time-step11. Hence, step that number. by step they acquire some definite value. As time passes Accordingly, to name them “real number” is seriously they have a changing disposition (or propensity) [12] to confusing. A better terminology would be to call them hold their eventual value. This propensity changes at “random numbers”. Unfortunately, Descartes named each time-step, similarly to the quantum probability of them “real” to contrast them with the complex num- physical quantities of quantum systems that also evolve bers, those numbers that include the square root of 1, as time passes. I suggest that this is similar to the Brow- traditionally denote i. Hence: − nian motion of some particle that evolves in-between two Mathematical real numbers are physical random num- sticky plates (that code for the bit values 0 and 1) un- bers. til it eventually sticks either on the left plate or on the 12 I think it can be speculated that had we learned in right plate . Accordingly, the openness of the future school to name such numbers ”random” rather than enters gradually, sometimes on millisecond scales and for ”real”, we would be less inclined to adopt a determin- other systems on scales of millions of years. This is just istic outlook on the basis of the science they figure in. a brief statement of an idea, an elaboration will follow in a future paper. One may object that this view is arbitrary as there is no natural bit number where the transition from deter- VI. NON-DETERMINISTIC CLASSICAL PHYSICS mined to random bits takes place. This is correct, though not important in practice as long as this transition is far away down the bit series. The lack of a natural tran- In this section we return to physics, now that we sition is due to the fact that, in classical physics, there have established that the “mathematical real numbers is no equivalent to the Plank constant of quantum the- are physical random numbers”, i.e. most mathemati- ory. But this is quite natural, as the fact is that when cal real numbers are physically immaterial. We define one looks for this transition in the physical description alternative classical mechanics as the same set of dy- of classical systems, one hits quantum physics. namical equations as standard classical mechanics, but Admittedly, Newton’s equations, as well as Maxwell’s all parameters, notably the initial conditions, are given equations, are deterministic: given initial conditions in by finite-information numbers. Note that for integrable the form of real numbers, all the future and past are dynamical systems, the coordinates at all times given fixed13. But the fact that these equations are mathemat- t by a finite-information parameter are themselves de- ically deterministic doesn’t imply that physics is deter- scribed by finite-information numbers. Hence, all predic- ministic. For example, this is definitively not the case tions about integrable systems remain unchanged. when the initial conditions of chaotic systems are not However, looking back to the example of the typical identifiable with mathematical real numbers, as in our classical chaotic dynamical system of section II, we re- alternative classical mechanics. Consequently, whether call that the leading bit describing the system, i.e. the classical physics is deterministic or not is not a scientific bit that determines whether the system lies of the left question, but depends on the physical significance one half (rainy weather) or on the right half of the unit inter- associates with mathematical real numbers. val (sunny weather) after n time-steps, depends on the As the philosopher Elizabeth Anscombe emphasized nth bit of the initial condition, b . But, if n is large n [14], “the high success of Newtons astronomy was in one enough, this nth bit of the initial condition had, at the way an intellectual disaster: it produced an illusion (...) time corresponding to the initial condition, no physical for this gave the impression that we had here an ideal of significance. Hence, according to our alternative classical scientific explanation; whereas the truth was, it was mere mechanics, chaotic dynamical systems are truly random. obligingness on the part of the solar system, by having had Let me emphasize that they are not merely random for all practical purposes, but that they are truly random, as random as quantum measurement results. This random- 11 The time-steps are used here only as illustration. Time could pass much smoother, with the propensities of all bits varying slowly. determined by the successive outcomes of a given (infinite) se- 12 Note that this is similar to the quantum state of a quantum- quence of “true coin tosses”, e.g. the outcomes of a quantum bit (qubit) in spontaneous localization models described with random number generator. In such an artificial case, every sec- stochastic Schr¨odinger equations [13]. ond bit is predictable, but all others are truly random. 13 Up to exceptional cases, see footnote 1. 6 so peaceful a history in recorded time, to provide such a distributed according to the usual quantum probabili- model”. ties, i.e. ψ(~x) 2, then the statistical distribution of the Bohmian| positions| remain in accordance with quantum probabilities at all times. This is very elegant and, like for VII. SUPPLEMENTARY VARIABLES... the real numbers for classical physics, adding Bohmian positions to the ontology of the theory turns quantum So far we have seen that physics is non-deterministic physics into a deterministic theory. Note that it requires and that this is true both of quantum [15] and alternative also to “trust” real numbers, as the Bohmian positions classical mechanics. In this section we turn to the natural and the quantum state vector use them. question of whether one could add supplementary vari- A point of caution is due here: In case of systems com- ables to quantum and to alternative classical mechanics posed of more than one particle, one should realize that in order to restore determinism. the evolution of any particle, let’s say the first one, de- pends on the entire wave-function. Hence, it depends also That it is possible to do so in the case of classical me- on what happens at the location of the other particles, chanics is well known. It suffices to add the mathematical i.e. each particle is guided in a non-local way. This is real numbers, as is usually done without even mention- necessary because quantum predictions violate the Bell ing that these are supplementary variables. Once these inequality, hence all alternative (or supplemented) the- real numbers are added and postulated to be part of the ories that reproduce quantum predictions must contain ontology of the theory, the so extended theory is deter- some non-local features [15, 19, 20]. But this is proba- ministic. Somehow, all the randomness has been pushed bly why most physicists reject Bohmian mechanics: they back to the (unattainable) initial conditions, as discussed dislike explicit (though unavoidable) non-locality. in the next section. For physicists this may look like a joke: we first ar- gued at length that ”‘real numbers are not really real”’ VIII. ... PUSH RANDOMNESS BACK TO THE just to next introduce them again. But notice what is INITIAL CONDITION achieved by viewing things in this way. The real num- bers are certainly not necessarily part of the ontology of classical physics, it is not the experimental facts that As we saw in the previous section, both alternative force physics to include real numbers in the ontology classical and quantum non-determinism can be turned of classical physics. Hence, at first, classical physics is deterministic by adding supplementary variables. In non-deterministic. However, all non-deterministic theo- both cases, the complemented deterministic theory is ries can be turned deterministic by adding supplemen- rather elegant. In both cases, the original randomness tary variables. In full generality, it suffices to add, for is pushed back to the initial conditions. Indeed, as time example, as supplementary variables all results of all fu- passes, instead of new bits in the series (1) gaining de- ture measurements, while making sure that these sup- termined values, new bits from the initial condition gain plementary variables remain hidden as long as the corre- relevance. Hence, we face a choice: either the fact that at sponding measurements haven’t happen. In fact, that’s present certain things happen and others do not is inter- exactly how classical mechanics is done: postulating that preted as revealing, retroactively, information about long the initial condition of all classical dynamical systems are past initial conditions, or else, we understand the present faithfully described by real numbers is an elegant way of as the result of indeterminate reality, and the future as adding all future results, while making sure that they re- open. If we care about how we experience reality, the main inaccessible for long enough a time. Admittedly, later option is obviously superior. just adding future results and postulating that there are Noteworthy, it is a fact that almost all physicists inaccessible won’t convince any scientist. Adding the real do complement alternative classical mechanics with the numbers to classical physics is much more convincing, be- mathematical real numbers; they do so even without cause it is elegant. But is it truly different? thinking about it. At the same time, almost all physicist And what about quantum physics? Here there is a reject Bohmian mechanics arguing that it is unnecessar- well known way to add supplementary variables in such ily complicated and doesn’t lead to new physics. How- a way as to turn quantum mechanics deterministic. This ever, one may argue that the real numbers accepted in is known as Bohmian mechanics (or the de Broglie-Bohm classical physics are also unnecessarily complicated (re- pilot wave) [16–18]. Essentially, one postulates that all member, they contain an infinite amount of information. particles always have well defined - though inaccessible Isn’t that hugely complex?). Furthermore, one can ask - positions and that, at the end of the day, all mea- which new physics the real numbers produced. surements are positions measurements (position of some pointer, position of some electrons that turn on/off some LEDs, etc). These particle positions, which I name IX. CONCLUSION Bohmian positions, are guided by the solution of the usual Schr¨odinger equation in a clever way such that In our society the concept of information is ubiquitous. if the initial positions are assumed to be statistically Today, it is quite natural to assume that no finite volume 7 of space can hold more but a finite amount of (Shannon) easily complement classical physics, but are reluctant to information, as measured by bits. Consequently, I ar- make the similar move for quantum physics. gue that one should not attribute to real numbers, i.e. In summary, physics with all its predictive and ex- to numbers that contain an infinite amount of informa- planatory powers can well be presented as intrinsically tion, any physical significance. This observation implies non-deterministic. The dominant view according to that there is a simple alternative to standard classical which classical physics is deterministic is due, first, to mechanics, based on finite-information numbers, which a false impression generated by it’s huge success in as- is a non-deterministic theory although it has exactly the tronomy and in the design of clocks and other simple same predictive and explanatory powers. mechanical (integrable) systems, and, second, to a lack At the time of Laplace the concept of information, of appreciation of its implication for (infinite) informa- in particular its quantization in terms of bits, was non- tion density. existing. Hence, it was natural to identify initial con- Finally, an indeterministic world is hospitable to Res ditions of classical dynamical systems with the math- Potentia and to the passage of time [22, 23, 25]. ematical real numbers. But today that we know that “real numbers” contain an infinite amount of informa- tion and, as suggested above, that they would be better called “random numbers”, we should realize that such numbers can’t be the basis for determinism. Acknowledgment Accordingly, both classical and quantum theories can and, I claim, ought to be regarded as non-deterministic. This work profited from stimulating discussions with Of course, one may want to complement these theories Augustin Baas, Cyril Branciard, Barbara Drossel, Florian with supplementary variables in such a way that the com- Fr¨owis, Michael Hall, John Norton, Valerio Scarani and plemented theory is deterministic. Note that this can be Christian W¨uthrich. Financial support by the European done both for quantum and for classical physics, as seen ERC-AG MEC is gratefully acknowledged. in sections VII and VIII; in both cases the supplementary variables are inaccessible. The fact is that most physicists

[1] J.D. Norton, www.pitt.edu/∼jdnorton/Goodies/Dome [13] N. Gisin and I.C. Percival, J. Phys. A, 25, 5677 (1992). [2] Y. Dolev, Relativity, Global Tense and Phenomenology, [14] G.E.M. Anscombe, Inaugural lecture at Cambridge Uni- in Cosmological and Psychological Time, Y. Dolev and versity on Causality and Determination, 1971. M. Roubach Eds., Springer, Boston Studies in the Phi- [15] N. Gisin, Quantum Chance, nonlocality, teleportation losophy of Science (2016). and other quantum marvels, Springer, 2014. [3] B. Drossel, On the relation between the second law of [16] D. Bohm, A Suggested Interpretation of the Quantum thermodynamics and classical and quantum mechanics, Theory in Terms of ”Hidden” Variables, Phys. Rev., 85, arXiv:1408.6358 166 and 180, (1952). [4] J. Norton, Time Really Passes, Journal of Philosophical [17] J.S. Bell, On the Impossible Pilot Wave , Found. Phys. Studies 13, 23-34 (2010). 12, 989-999 (1982). [5] Y. Dolev, Physics’ silence on time, Euro. Jnl. Phil. Sci., [18] D. D¨urr and S. Teufel, Bohmian mechanics: the physics https://doi.org/10.1007/s13194-017-0195-z (2018). and mathematics of quantum theory, Springer Science & [6] https://en.wikipedia.org/wiki/Baker%27s map Business Media (2009). [7] G. Chaitin, The Labyrinth of the Continuum, in Meta [19] J. S. Bell, Speakable and Unspeakable in Quantum Me- Math!, Vintage Books, NY, 2008. chanics: Collected papers on quantum philosophy (Cam- [8] C.J. Posy, Varieties of Indeterminism in the Theory of bridge University Press, Cambridge, 1987). General Choice Sequences, J. Philosophical Logic 5, 91- [20] N. Brunner, D. Cavalcanti, S. Pironio, V. Scarani, S. 132 (1976). Wehner, “Bell nonlocality,” Rev. Mod. Phys. 86, 419 [9] G. Dowek, Real numbers, chaos and the principle of a (2014). bounded density of information, Computer Science - The- [21] N. Gisin, Propensities in a non-deterministic physics, ory and Applications, pp 347-353 (2013). Synthese 98, 287 (1991). [10] R. Bousso, The holographic principle, Rev. Mod. Phys. [22] N. Gisin, Time really Passes, Science can’t deny that, 74, 825 (2002). arxiv/1602.01497, in Time in Physics, eds R. Renner and [11] J.D. Bekenstein, Universal upper bound to entropy-to- S. Stupar, Springer 2017. energy ratio for bounded systems, Phys.Rev. D 23, 287- [23] Admittedly, the non-determinism of our alternative clas- 298 (1981) sical mechanics questions when do potentialities become [12] M. Dorato, Do Dispositions and Propensities have a role actual. Continuously as part of the dynamics [24]? This in the Ontology of Quantum Mechnanics? Some Criti- might shed new light on the old and infamous quantum cal Remarks, in Probabilities, Causes and Propensities in “measurement” problem. Also, questions about time re- Physics, Ed. M. Su´arez, Synthese Library, Springer, pp versal may get interesting answers. If one inverts the map 197-218 (2011). (3), new leading bits must be added at each time-steps, 8

but without losing any bit. Hence, both the future and versity Press, ed. Shan Gao, pp 207-224, 2018. the past are open, but the present contains more infor- [25] Y. Dolev, Motion and Passage - The Old B-Theory and mation about the past than about the future. Phenomenology, in Debates in the Metaphysics of Time, [24] N. Gisin, Collapse. What else?, arXiv:1701.08300, pub- Oaklander L. N, Editor, Bloomsbury, 2014. lished in Collapse of the Wave Function, Cambridge Uni-