How simulation hypothesis can potentially solve one of the toughest problems in neuroscience?

Outline: What if simulation hypothesis is correct? If so, will it have any implications in our daily life? I am going to assume the simulation hypothesis is correct and explore a new way to tackle one of the toughest problems in the field of philosophy and neuroscience dubbed as “the hard problem”.

 

Simulation hypothesis:

 

Let us begin with a simple question. Is there life in this universe? The answer is an obvious yes, even if we assume earth is the only planet in the whole universe that can support life. In that case, one of the following propositions must be true. Proposition 1: Life originated and evolved accidentally in our universe. Proposition 2: Multiverse exists, and our universe is one among the infinite possible other universes, where life originated and evolved accidentally. Proposition 3: Our universe is fine-tuned to support life. Let us look at some facts that may help us zero in on one of the above propositions.

 

Since all the three propositions are based on life, it is important that we understand what life is. Physicist Erwin Schrodinger defined life this way: Living things avoid decay into disorder and equilibrium. This entails that living systems spend energy to create order and thereby avoid decay and disintegration. This is what living things do. But what is life? NASA defines, “Life is a self-sustaining chemical system capable of Darwinian evolution”.

 

Now let us look at the prerequisites for life to exist. Our universe operates according to a set of rules. We try to model the universe using a bunch of equations derived out of the laws of physics. Consider for instance the equation for gravitation proposed by Newton. This equation involves three variables and a constant known as the gravitational constant-G. One can imagine constants as simple numbers that set the scale for the equation. There are about 19 such constants in physics that we use to model the universe. An interesting aspect about these constants is that they are not anymore reducible or derivable from other equations and are independent of each other. Their values can be determined only by performing experiments. This means that they are fundamental and are reflecting one or more physical behaviors of the universe. These 19 constants define the physics of the universe.

 

Let us get back to proposition 1. For this proposition to be true, we got to be extremely lucky that all the 19-odd parameters happened to be within a very narrow range of numbers that allow life to form, sustain and evolve. For example, if the fine-structure constant is a tiny bit different, our universe will not even contain galaxies and stars, and of course life! The probability for all those 19 free parameters to have the right set of numbers that can form life seems to be extremely low. This implies that proposition 1 is highly unlikely.

 

Moving on to the next proposition – the multiverse. According to this hypothesis, perhaps these constants are set to the right values in our universe, while there exist infinite other universes with other possible values for the constants that are not conducive to create life. This means that there are enough universes that somewhere the values for the constants had to be exactly right for life, and luckily that is where we find ourselves. This is also known as the anthropic principle. This proposition sounds more probable. But there are a few concerns. Does this fit the notion of Occam's razor? Is multiverse the simplest explanation possible for the underlying problem? In the multiverse thesis, we have essentially multiplied the problem with infinity to increase the likelihood for the creation of life, which seems to be beyond necessity. Another problem is falsifiability. We cannot conduct an experiment to prove or disprove the possible existence of other universes.

 

Now we are left with the final proposition – the fine-tuned universe – which implies that there is only one universe and someone or something had manipulated the values of the fundamental constants right from the big bang itself in order to create life. The next big question is who is that someone or what is that something that did all this. We have three options here as well. First, the most wide-spread hypothesis called “God”. In fact, this is the simplest of the three hypotheses, yet somewhat difficult to accept for various reasons. What else we have? The second option is a conjecture known as “biocentrism” proposed and popularized by Robert Lanza, according to which the universe originated and evolved on its own with a strong purpose in mind - creating intelligent biological entities. This implies that the universe must have fine-tuned itself to produce a conducive landscape for life. This hypothesis forces one to treat the mystical ability of the universe to fine-tune itself as a black box and sheds no light on how this could be possible. There is one last option – we all live in a simulation run inside a giant computer. This is the now famous “simulation hypothesis”.

 

One should be crazy to buy this idea. Some even say this is the mother of all conspiracy theories. However, if we deeply investigate all the available options, simulation hypothesis seems relatively more convincing than the rest. Again, the problem with these ideas is falsifiability. At present, we do not have a reliable way to test their validity.

 

However, in 2003, philosopher Nick Bostrom proposed a trilemma that he called "the simulation argument". Bostrom's trilemma argues that one of three unlikely-seeming propositions is almost certainly true:

 

1. "The fraction of human-level civilizations that reach a posthuman stage (that is, one capable of running high-fidelity ancestor simulations) is very close to zero", or
2. "The fraction of posthuman civilizations that are interested in running simulations of their evolutionary history, or variations thereof, is very close to zero", or
3. "The fraction of all people with our kind of experiences that are living in a simulation is very close to one."

 

The trilemma points out that a technologically mature "posthuman" civilization would have enormous computing power; if even a tiny percentage of them were to run "ancestor simulations" (that is, "high-fidelity" simulations of ancestral life that would be indistinguishable from reality to the simulated ancestor), the total number of simulated ancestors, or "Sims", in the universe (or multiverse, if it exists) would greatly exceed the total number of actual ancestors.

 

Bostrom goes on to use a type of anthropic reasoning to claim that, if the third proposition is the one of those three that is true, and almost all people live in simulations, then humans are almost certainly living in a simulation.

 

The upshot is, fine-tuned universe seems to be more reasonable to provide an explanation for the existence of life in this universe, while simulation hypothesis explains how such a fine-tuning could have been created.

 

Moreover, I personally find this proposition very useful over the other ideas as it has the potential to answer some of difficult problems in science and philosophy.  

 

We have reasons to believe that the universe is fine-tuned to produce life. Likewise, there is a possibility that we might live in a simulation fine-tuned to produce life.  But it is important to note that there's no direct evidence that our universe is a simulation. Perhaps, the advances in theoretical and experimental physics in the near future can help us explore the real nature of our existence. In the meantime, we can keep our imagination wild and ask these two profound questions: what if simulation hypothesis is correct? If so, will it have any implications in our daily life? I am going to assume the simulation hypothesis is correct and explore a new way to tackle one of the toughest problems in the field of philosophy and neuroscience dubbed as “the hard problem”.

 

The hard problem of consciousness

 

What is consciousness? It is the subjective experience of the happenings in our life. It is somewhat analogous to an inner movie capturing what I feel from the first-person point of view. The hard problem is the problem of explaining why there is subjective experience, and why any physical state is conscious rather than non-conscious. This is the problem of “experience” and precisely about exploring the answer to the question “how there is experience generated out of physical processes”. Right now, no one has a convincing answer to this question and possibly no one ever will. This is why David Chalmers, the most prominent philosopher of our time, named it as the hard problem to emphasize on the fact that no theories will ever be able to explain conscious experience in terms of physical processes. David Chalmers says, “The hard problem is hard precisely because it is not the problem about the performances of functions. The problem persists even when the performance of all the relevant functions are explained”.

 

Conversely, the easy problems are those that deal with explaining the function and dynamics of consciousness such as correlating a specific conscious experience with a specific part of the brain or the so-called neural correlates. The easy problems (though not so easy from a practical standpoint) are easy in the sense that they are tractable to the established methods of science, however, that is not the case with the hard problem, which is seemingly intractable to the methods of science.

 

Lately, this problem became a major focus of research in contemporary philosophy of mind, and there is a considerable body of theoretical and empirical research in psychology, neuroscience, and even quantum physics.

 

This problem of consciousness challenges the physicalism viewpoint that there is "nothing over and above" the physical. Essentially physicalism denies and ignores the existence of consciousness altogether.

Do we have any real evolutionary advantage of being conscious? Scientific studies are inconclusive about this point. Consciousness may or may not provide any kind of evolutionary advantage. If consciousness does not have any primary function for the survival within the simulation, what is the need of this feature in the simulation? At the outset, consciousness seems like an unnecessary feature in the simulation. In fact, consciousness seems to be the only non-physical feature in the whole of universe, regardless of whether it is a simulated one or otherwise. On the other hand, if simulation is the basis of our existence, then it must have something to do with consciousness and be able to explain the nature of it. Does the intractability of the hard problem indicate anything deeper? Perhaps we can answer why we have consciousness if only we concede the simulation hypothesis, and that is exactly what I have attempted here.

 

A possible solution for the hard problem

 

Before jumping to the solution part, let me formulate the hard problem in a rigorous manner. Consider two physical systems S1 and S2. Assume that, from all physicality point of view, S1 and S2 are one and the same. This means that one cannot distinguish S1 and S2 by any physical method and both systems will be equivalent from the external perspective. According to the hard problem, S1 can be a conscious system (i.e. having some subjective experience), while S2 can be fully unconscious. How can two physically equivalent systems have different conscious states? Currently we do not have answer for this question. But what becomes clear is that a physical system can possess “internal state” that could be independent of its “external state” governed by the physical laws. The problem statement is - If physics cannot decide a system’s internal state, then what decides it?

 

Having formulated the problem statement, now let us turn to the simulation hypothesis to solve this problem. If we assume simulation hypothesis is correct, we can do some guess work on what could be the best possible computational architecture that can suit for such a giant scale of simulation. It is highly likely that the posthumans choose a de-centralized and distributed architecture to improve the computational efficiency and performance. In fact, de-centralization is inevitable in a large-scale organization. There are numerous advantages in this kind of architecture. For example, it is possible to parallelize the computational process which in turn improves the performance. Moreover, there will not be a single point of failure and hence fixing the issues can be a lot easier. As a matter of fact, we – humans – are constantly moving towards more distributed computational approaches these days as opposed to centralized ones. Hence it is not unreasonable to assume that the simulations created by the posthumans will follow a decentralized architecture.

 

In this architecture, there will be multiple self-simulating units (SSUs) that can interact with the other SSUs. For instance, if we are tasked to simulate a car, we may split the simulations into multiple smaller units, one for simulating the combustion, one for simulating the gearing mechanism, and so on. We will have a code written to integrate all of these smaller sim units to create the sim of the whole car. Like any distributed computational systems, the SSUs are expected to have a certain degree of autonomy and driven by a local vision aligned with the global vision of the simulation (maybe producing more intelligent lives?) just like how different states function within a country. Each SSU can self-simulate and evolve in time with a shared vision. The internet of these countless number of SSUs would essentially generate a grand unified simulation called the universe.

 

If the purpose of fine-tuning of the physical constants was meant to create life, it makes sense to define a single SSU as the smallest entity within the simulation that can digitally represent a smallest scale of life. In other words, a single SSU is an entity in the simulation that can hold one unit of life. If we can somehow derive a mathematical condition to create an SSU or understand when the simulators bifurcate a simulation unit into one or more SSUs, we can gain insight into the mathematical conditions necessary to create life in the context of simulation hypothesis. This provides a novel way to define life itself and study its attributes in a different light. Perhaps this approach can help us understand why we are unable to create life from non-life ingredients in the labs, one reason could be that bifurcation of SSUs is completely controlled from outside the simulation environment by the simulators. This approach may also shed light on another fundamental problem in science – knowing the origin of life. These will be the focus of my next article.

 

Moreover, it is quite possible that the posthumans might use or have used a powerful artificial intelligence (AI) of some sort to efficiently orchestrate the whole simulation that lets the SSUs self-learn over time, while contribute to the overall simulation more effectively.

 

Interestingly, this way of looking at our existence opens a hitherto unknown avenue to solve the hard problem. The phenomenon what we call consciousness could just emerge from the interactions taking place among the numerous SSUs. When a SSU relays a signal about its current state to a neighbor SSU mediated by an AI system, this may produce a consciousness-like sensation or a subjective experience within the SSU that relays the signal and also within the SSUs that receive it. So, consciousness could be the result of inter-SSU communications.

 

Does this mean any kind of interactions between the SSUs would produce a subjective experience? Hang on here. I have addressed this question a bit later with the help of another established theory.

 

Essentially when a SSU sends a signal to another SSU, the signal crosses the physical boundary of the sender SSU and travels inside a pure computational environment for a while before landing inside the physical boundary of the receiver SSU. During the journey from physical to computation to physical, the signal might attain a pure “informational state” when it crosses the boundary of the sender SSU until it touches the boundary of the receiver SSU.  I am not sure if this transition from physical to informational is the source of consciousness. However, this implies that the SSU interactions and the outcome of this interaction (that is supposed to create consciousness) cannot just be studied with the help of physical theories. We may have to formulate a new hybrid theory that incorporates both - the physical aspects and the informational aspects – in order to solve the hard problem.

 

For instance, when you see a red object, the sensation of the “redness” could be the result of how the SSU that generated the color information communicates the same with other SSUs within the simulation environment. The simulation may govern how strongly a given SSU must feel the sensation of redness by using a hybrid logic that incorporates the informational and physical aspects together. For the sake of simplicity, we can say that the physical laws are strictly applicable within each SSUs, while informational laws govern what happens between the SSUs. Perhaps this explains why internal states can be independent of their physical states. Hence we may have to know the informational laws that govern the inter-SSU space in order to understand consciousness fully. Maybe in the near future, we might be able to model consciousness by using the combination of physical laws and informational laws derived from the simulation hypothesis.

 

Let us see how this proposition aligns with one of the most successful theories of consciousness – the integrated information theory or simply IIT, which was proposed by Giulio Tononi at the University of Wisconsin in Madison.

 

At the core, IIT claims to quantify something called “integrated information” of any physical system that can process information (a neural system is a subset of physical system) denoted by the symbol Φ (a Greek letter pronounced as ‘Phi’). According to IIT, a physical system with greater Φ is inherently more conscious and vice versa. Essentially Φ quantifies the amount of “intrinsically irreducible interconnectedness” of the parts of the information processing system. In other words, Φ corresponds to the capacity of the physical system to process information in an integrated way. Essentially IIT provides a set of mathematical conditions and any systems that obey these conditions must be conscious, says IIT.

 

Though IIT provides a mathematical condition for consciousness, it suffers from a major problem. Essentially it does not address the hard problem in first place. This drawback stems from the fact that IIT readily accepts the view that the consciousness is a fundamental feature of reality, and develops a mathematical framework to model consciousness. In fact, IIT directly takes off from the intrinsic perspective – from experience itself and its phenomenal properties, and from there it goes on to postulate the existence of a physical world of elements in a state having properties that can account for those of experience itself. Therefore, IIT is of no help when it comes to solving the hard problem and it does not explain why a certain type of information processing gives rise to consciousness.

 

However, when we mix up the SSU-approach and IIT, it seems to offer a comprehensive route to tackle the hard problem. SSU-approach explains why an informationally-founded method along with physical theories is required to solve the hard problem, while IIT provides the necessary mathematical conditions that the informational-physical interaction must obey to generate consciousness.

 

This addresses my previous unanswered question – what kind of interactions taking place between the SSUs would produce consciousness? If we invoke IIT, the answer is that those interactions that satisfy the mathematical conditions imposed by IIT would produce consciousness. This means that not all interactions between the SSUs would produce consciousness.

 

An interesting similarity between SSU approach and IIT is that both agree on the point that a system can be conscious even if it is silent from all extrinsic perspective. In SSU this phenomenon can explained in terms of “null signals”. When an SSU is completely inactive, still it must relay its null state to other SSUs and hence generate consciousness out of this state of nothingness. IIT uses complex mathematics to show this possibility. We can infer from these notions that the internal state of a system cannot be accurately determined by physics. For instance, take two persons. One is in deep state of coma – no neuron firing - and the other one is in deep state of meditation – again, no neuron firing. From all physical and functional perspective, both are the same. However, according to SSU and IIT, these two persons can have completely different internal states.

 

At the outset, SSU seems to be in alignment with IIT. This is a promising first step. But, have we solved the hard problem? One can still ask why the SSUs have to “feel” like something when it interacts with other SSUs and we are back to square one again. The SSU hypothesis only helps understand why we need much more than physics to comprehend consciousness. But the underlying problem remains unsolved.

 

It is clear that the relay of information from one SSU to the other has to happen only after the information got generated in the SSU. So, it seems like the SSU has to deal with the same information twice – first, when it performed the simulation and produced the information, and second, when it interacted with other SSUs to communicate about the information. This activity can leave an imprint on the SSU when the interaction with other SSUs become stronger (or when the interaction satisfies the conditions imposed by IIT), which may form an inner voice for the SSU that generated the information. This is one way to use SSU hypothesis to explain why there is conscious experience.

 

Though SSU provides a unique way to explain consciousness especially when combined with IIT, it does not offer in its current state a satisfactory answer to the hard problem. Much has to be done and incorporated to make this new idea a fully matured scientific theory to be able to solve the hard problem.