Essentia Foundation: Could you briefly summarize your interpretation of Quantum Mechanics (QM), and perhaps relate it to QBism and Relational Quantum Mechanics?

My goal is not to interpret QM—at least not in the usual sense of finding an intuitive explanation for what is ‘really going on in the quantum world.’ Instead, I see QM as one motivation (among many) to drop our intuitive, traditional notion of ‘world’ altogether, and to regard it as only approximately true. My work implements this idea in a particular, mathematically rigorous way without assuming the validity of QM. It then turns out that some aspects of QM appear naturally as predictions.

In this approach, quantum states turn out to play a similar role as they do in QBism: they are nothing but a catalog of probabilities of private future experiences (‘outcomes’) of agents. But there are two essential differences to QBism. First: QBists interprets these probabilities as subjective beliefs (Bayesian betting probabilities), whereas in my approach they are objective but private chances—they say how likely an agent is going to see something later, regardless of its beliefs. Second, QBists place a lot of emphasis on the notion of an agent’s ‘action’: the agent can decide to do something, and the world kicks back. In my approach, the notion of ‘agent’ is a more abstract, structural notion of ‘self,’ and its ability to decide and perform actions is seen as irrelevant for my purpose.

 

Essentia Foundation: You motivate your essentially idealist view with some curious thought experiments. Yet, we now have overwhelming experimental confirmation of some predictions of quantum mechanics—namely, Bell’s and Leggett’s inequalities—that seem to refute the notion of an objective physical world describable with definite quantities. Could you comment on what these experiments may mean or imply, in the context of your views?

I believe that both the (experimentally confirmed) violation of Bell’s inequalities and the curious thought experiments hold the same message: trouble for the ‘container’ view.

By ‘container view’ I mean the following intuitive picture: that there exists an objective, material world that evolves in time according to physical laws, containing us as material agents. Relativistic peculiarities notwithstanding, at every time, this world is in some definite configuration in some sense. We, as agents, can make predictions by, first, modelling the evolution of this world, and then by locating ourselves inside this world (“oh, I’m here on Earth, in the Netherlands, hence modelling the motion of the planets according to mechanical laws, I can predict that I will see Jupiter here on the sky tomorrow!”).

Bell’s theorem tells us that this container view cannot be upheld in all cases, unless we are willing to pay a prize: we must be willing to give up either our idea of locality or our idea of what it means for the world to be in some definite ‘configuration.’ If you opt for the first possibility, then you may be a Bohmian: you say that particles interact faster than light, but in a way that is in principle invisible. If you opt for the second possibility, then you may be an Everettian: the configuration of the world is itself a wavefunction (intuitively interpreted as some kind of ‘parallel-worlds’-entity), but then you have to work hard to relate this to the empirical fact that we always see single definite outcomes that are very well characterized by the laws of probability theory.

An obvious way out—for which I argue—is to give up the container view altogether, at least fundamentally. For some reason, this seems to be psychologically difficult. Hence, if the quantum puzzles were the only ones we encounter in physics or philosophy, then one might well choose to be a Bohmian or an Everettian and enjoy the resulting psychological comfort. But as I argue in my work, there are many more challenges to this container view than just quantum mechanics; for example, the puzzles and thought experiments that I have mentioned in my talk, which arguably render the container view methodologically inadequate. I believe that there is something important about our world and our place in it that we have so far failed to grasp.

 

Essentia Foundation: You have alluded to the many-worlds Everettian view. Indeed, some physicists try to hold on to an objective physical world by postulating a practical infinity of parallel physical universes popping into existence every fraction of a second. Could you comment on why this supposedly saves physical realism, and why you think it doesn’t work?

Let me clarify this: I believe that there are immensely important aspects of ‘physical realism’ that we have to defend at all cost because they constitute the basis of science. These include the strive for logical and mathematical rigor, for stringent experimental test whenever possible, and for Occam’s-razor-like simplicity. In this sense, I am a die-hard ‘realist.’

But the word ‘realism’ is ambiguous and overloaded. In particular, these important principles of science are sometimes unduly identified with a very specific notion of realism: with something close to the ‘container view’ described above; as if the only possibility to be rational would be to hold a naïve view of a world that unfolds like a stage play!

Given the puzzles of quantum mechanics, the many-worlds view that you have mentioned in your question aims at telling a coherent story about the quantum world in terms of this container or stage play view. Does this Everettian interpretation succeed in doing so? Yes, absolutely! But the problem is that you can make every worldview consistent with modern physics if you stretch it far enough. Do you dislike the idea of many worlds and would like to hold on to a single-world classical picture? Then pick Bohmian mechanics! Do you prefer to abandon any notion of randomness altogether? Then pick ‘t Hooft’s superdeterministic cellular-automaton interpretation! Pick whatever you like—the experimental predictions will be identical, and nobody can prove you wrong.

Given this situation, I think that the only reliable way to understand what we can really learn from QM—what it tells us about the world—is to disregard interpretations and look at actual scientific practice. There, we find that quantum states are nothing but our calculational tool to determine probabilities of measurement outcomes—and all we ever see are these outcomes. And it turns out that these probabilities have surprising properties. For example, they violate Bell’s inequalities. The simplest logical conclusion to me is to see this as a hint that the world is, first, fundamentally probabilistic in some sense, and second, that we cannot consistently regard the outcomes of our measurements as predetermined in any way that would deserve this name.

 

Essentia Foundation: You make some effort to explain that the emergent world perceived by a first observer—say, Alice—will converge with that of a second observer—Bob—visible in Alice’s world. In other words, the world within which Bob is perceived by Alice will tend to be consistent with Bob’s own world. This suggests that you, contrary to solipsism, acknowledge the existence of multiple observers conscious of their respective worlds, and that the presence of an observer can be perceived in the emergent world of another observer. Is this correct?

This is largely correct and well-formulated. In particular, it is true that my view does not amount to solipsism. I also completely agree with your statement that “the presence of an observer can be perceived in the emergent world of another observer.”

There is, however, some fine print to this. For example, when you say that I “acknowledge the existence of multiple observers,” one has to guard against the reflex to associate the notion that ‘X exists’ with the mental picture that X has a time and a place where it sits, and a world in which it is embedded. This is exactly what is fundamentally denied in my approach.

For example, suppose that you (Bernardo) meet a friend (Charlie) on Christmas eve. There is an ‘observer state’—an abstract structure B—that describes all of your self (your conscious or unconscious memory, your experiences etc.) at that very moment. There is another observer state C that does so for Charlie’s self. But both B and C are simply abstract structures that do not have a time or place ‘where they sit.’ It may be that B contains the experience of seeing, for example, a calendar that shows ‘Dec 24, 2020,’ and a watch that shows ‘20:00’—together with a mental model of the room that supposedly surrounds your perceived body. But B itself doesn’t have a place or time. And C also doesn’t. The block of biological matter that you perceive in front of you is not identical to C, but simply a (probabilistically accurate!) representation of C, similarly as six eggs in a box are a representation of the number 6 (as an abstract structure), but are not equal to it.

My approach relies on the notion of ‘observer states’—that is, momentary ‘modes of being’ in some sense—not of ‘observers.’ Being in state B means feeling like Bernardo in this room on Christmas Eve at 20:00, and it means being in some other state B’ next that is probabilistically determined by B (and with high probability will still feel like Bernardo in this room on Christmas Eve, with small changes). But there is no fundamental notion of ‘being Bernardo.’ And this has surprising consequences. It may happen that, in the extremely distant subjective future, you will actually have ended up in observer state C—meaning that you will feel like Charlie on Christmas Eve 2020. Then we have a situation like in the movie Back to the Future, where after time travel Marty McFly runs into his older future self. Does this involve one observer or two? It depends on how you count.

In summary: my approach does not amount to solipsism, but it also does not make specific claims about the ‘existence of multiple observers’—it rather describes a counterintuitive ontology where this notion dissolves in some sense, and only momentary notions of ‘observer state’ have fundamental relevance.

 

Essentia Foundation: Fascinating. But doesn’t this mean that there must be some ontic context—inaccessible as it may be—in which Bob’s very existence translates into information present in Alice’s world? If not, in what sense would your views differ from solipsism?

Let me explain one aspect that differentiates my view from solipsism. Instead of Alice and Bob, let me again talk about Bernardo and Charlie on Christmas Eve.

Suppose that you (Bernardo) decide to hit Charlie hard in the face—at the very moment when you are in observer state B and Charlie in C. Now, in my view, this has the consequence that somebody really has to suffer. Namely, it means that if one is in observer state C, then one has a high probability of transitioning into some state C’ next that is associated with an experience of pain. Now, claims of this form would certainly not make sense in any kind of solipsistic worldview (if nobody but me exists, then nobody else will suffer from my actions).

This follows from my theorem on consistency: asymptotically, other observers that you see in your emergent world will see the same emergent world. So, insofar as you can claim to have causal influence on some aspects of your world, this directly translates into causal influence on the first-person experiences of other observers. Note that this is true despite the fact that observer states B and C do not have a common time or place where they would sit.

This is a quite fascinating and counterintuitive form of agency or causality as predicted from my approach: everybody is dreaming their own dream, and there is no notion of ‘external time’ that would even allow us to say that we dream ‘at the same time.’ Nonetheless, our dreams are determined by statistical laws that lead to their correlation, which allows us to meet up.

 

Essentia Foundation:  Clear. Given this, how would you define what matter is, as opposed to how it behaves?

In my view, the notion of ‘matter’ is a particular way to conceptualize certain persistent patterns that permeate an observer’s emergent external world.

 

Essentia Foundation:  One of the most vexing problems in cosmology today is the so-called fine-tuning: the fundamental constants of nature are exquisitely tuned to the exact values they would have to have if complexity—and, therefore, life itself—were to exist. Yet there is no physical reason for why they should have the values they have. What is arguably the best answer to this challenge today—an evolutionary theory of parallel universes—is still uncomfortably speculative and lacking in empirical evidence. However, it would seem that, in your approach, the problem disappears altogether. Could you comment on it?

I have not thought enough about this particular problem (fine-tuning) to give you a definite answer, and I do not know enough about cosmology to give you an answer on the spot. But what I can tell you is that approaches like mine challenge the basic terminology that we are used to apply when thinking about such problems in the first place.

Our usual way of addressing cosmological problems begins with something close to what I have called the ‘container view’ above: with the assumption that there exists a single world that is ‘really actualized’ in contrast to all other conceivable worlds (which are not actualized), and everything that ‘really exists’ is contained in it. This world may be very large, and it may perhaps even deserve the name ‘multiverse’ if we think of it as consisting somehow of compartments that do not interact. Such a worldview invites us to ask, intuitively and automatically: Why has this particular world been actualized and not another one? Could our world be just one among many, and if so, how could we find empirical evidence for or against the existence of those other worlds (or about the other compartments of our hypothesized multiverse)? Can we think of ‘actualization of a world’ as a physical process in some sense?

My approach denies the very metaphysical basis of this view. It views questions like ‘does conceivable world W really exist?’ as similarly empty as ‘how many angels fit on the head of a pin?’ Instead of a one-world or many-worlds theory, it is fundamentally a ‘zero-worlds theory.’ It regards the likelihood of first-person experiences as the primary notion, and the notion of an external world as a derived or emergent concept—a remarkable and important concept (a lot of effort in my work goes into proving that such a concept is in fact predicted by my approach), but not a fundamental one. Hence, instead of asking ‘why does the really actualized world have property X?’ it invites us to ask: ‘What are the typical properties of an observer’s emergent external world, according to the fundamental probability rule that governs its first-person experiences?’ Part of the answer is: such a world will typically appear to have begun in a state of low entropy, predicting a property of our world that standard cosmological thinking would regard as remarkably atypical.

 

Essentia Foundation: What do you regard as the necessary theoretical and experimental next steps to further substantiate and advance your views?

Most, if not all, novel approaches in fundamental physics currently face the problem that they are experimentally extremely hard to test. This includes, for example, the different proposals that we have for a theory of quantum gravity, or approaches aiming for unification like string theory. Any experimental deviation from our established theories supporting such novel approaches would represent a breakthrough worth a Nobel prize.

Since my approach aims at addressing a class of questions that can be regarded as even more fundamental, it is necessarily even further removed from the regime of physics that involves the concrete prediction of quantities like particle cross section or decay probabilities. Thus, I see currently no hope of any experimental test. But this is not due to a problem with my approach, but due to the nature of the questions that it (or any comparable approach) intends to address.

However, I do see the possibility of an indirect test: as I have already indicated in my paper, I think that some aspects of quantum theory can perhaps be understood as generic consequences of approaches like mine. But there is quite some theoretical work left to be done to further substantiate this idea. To do so, I also need to address one particular aspect of my approach that currently renders it necessarily incomplete: it implicitly assumes that observers never lose any information (in some formal sense—see my paper). As I already acknowledge in my paper, dropping this assumption and reformulating my theory accordingly is an important improvement to be implemented.

Finally, and most importantly, I would like to get more into discussions with my colleagues, if only to obtain important criticism and to learn where my views may go wrong. This is not always easy, since my work neither falls into the traditional scope of physics nor into that of philosophy.