The Other Simulation Hypothesis

Posted on April 27, 2021 by James Cross

The simulation hypothesis is mostly associated with Nick Bostrom and his paper Are You Living in a Computer Simulation? Bostrom argues that we likely are living in a simulation and Elon Musk agrees with him. Frankly I think it is unlikely we are living in a simulation in the way Bostrom’s means it, but at any rate, it is impossible to prove or know and, as far as I can tell, would make no practical difference. In the end, if reality is a simulation, then being in a simulation or not being in one becomes for all practical purposes the same. There is a different way from Bostrom’s that we might be living in a simulation. This way could account for the occasional unreality of things most of us sometimes experience. It could account in a deeper way for why Bostrom might have thought about arguing we are living in a simulation.

Xerxes D. Arsiwalla, a physicist in Spain, was the lead author on a paper Are Brains Computers, Emulators or Simulators? In the paper, he draws a contrast between the brain as a computer vs the brain as a simulator. If the brain is a computer, he argues that “all cognitive processes can be described by algorithms running on a universal Turing machine”. This implies that consciousness is computational. On the other hand, if consciousness is non-computational, then it would be based on what he terms “non-classical logic”. He goes on to state:

Machines implementing non-classical logic might be better suited for simulation rather than computation (a la Turing). It is thus reasonable to pit simulation as an alternative to computation and ask whether the brain, rather than computing, is simulating a model of the world in order to make predictions and guide behavior. If so, this suggests a hardware supporting dynamics more akin to a quantum many-body field theory.

The paper goes on to discuss the limitations of computationalist view. He cites the Turing Halting problem and the Penrose tiling problem which can’t be solved by computation. Then he provides a “third example of a non-computable problem is the collapse of the wave-function or the measurement problem in quantum physics, which evades an algorithmic description”. Not mentioned here is another class of problem. This would be a type of problem that might be solved computationally but one that requires so much computer resources that it cannot be solved in any given amount of time.

An emulator “can be defined as any machine that can be used to specify dynamical states transitions of another system”. Computers can do emulations; however, a computer emulation would be subject to the limits of computation. Emulators can also be what the paper terms “dynamical systems-based simulations” which are not computational. The difference between the two is:

The difference of say computing an explicit solution of a differential equation in order to determine the trajectory of a system in phase space versus mechanistically mimicking the given vector field of the equation within which an entity denoting the system is simply allowed to evolve thereby reconstructing its trajectory in phase space. The former involves explicit computational operations, whereas the latter simply mimics the dynamics of the system being simulated on a customized hardware. For complex problems involving a large number of variables and/or model uncertainly, the cost of inference by computation may scale very fast, whereas simulations generating outcomes of models or counterfactual models may be far more efficient.

We finally reach the key argument of the paper. Brains are not computers. They are simulators.

Beyond this example of the motor system, if the brain is indeed tasked with estimating the dynamics of a complex world filled with uncertainties, including hidden psychological states of other agents… then in order to act and achieve its goals, relying on pure computational inference would arguably be extremely costly and slow, whereas implementing simulations of world models as described above, on its cellular and molecular hardware would be a more viable alternative. These simulation engines are customized during the process of learning and development to acquire models of the world. The simulated dynamics of these models lead to predictions as well as counterfactual hypotheses, which can then be passed through feedback control loops to correct for prediction errors. Note that these dynamics-based simulations differ from computer simulations. In the former, no specific function is being computed. Instead, as in control engineering, a model of the process is encoded (or learnt) in the network’s connectivity and is used to generate subsequent state transitions. More complex models require more complex network architectures and multi-scale biophysical dynamics, rather than heavy computational algorithms, which is presumably not what we see the brain to be designed for.

This explains much about the evolutionary origin of consciousness. Compared to actual computers, the brain and nervous systems must make the best with a relatively small amount of energy and a relatively slow computational speed. In simple organisms those limitations may not be fatal. However, the evolution of greater adaptive capability, the integration of more sensory data, and the development of broader repertoire of behaviors would eventually hit a computational barrier. The brain could not compute quickly enough to provide an selection advantage if it relied solely on a computational approach. The evolutionary response would be development of a simulation on top of a computational base. Unsurprisingly , our consciousness feels occasionally exactly like a simulation, although for the most part we think the simulation is real.

Posted in Consciousness, Human Evolution, Information | 47 Comments

Secret Ingredient?

Posted on April 20, 2021 by James Cross

Most great recipes have a secret ingredient. This is the spice that your grandmother leaves out of the recipe when she writes it down. The secret ingredients of Coca-Cola supposedly are only known to two people who are not allowed to travel together, but they have also been written down and stored in a vault. I listened to a cook on the radio swear that a single bay leaf made all the difference in the flavor of a particular recipe.

Does consciousness have a secret ingredient? If you think I’m talking about EM fields, not this time. Actually it may be something more obvious, not really all that secret, but something that actually gives us some insight on the evolution of consciousness, its nature, and how Friston’s theories fit into all of it.

In my comments on my last post, I mentioned there seemed to be a lack of clarity about how Friston’s free energy principle (FEP) related to consciousness. Since it seemed to be a theory as much or more about life itself, what made consciousness something unique in it? He didn’t seem to be claiming all life was conscious but he did seem to be claiming all life followed FEP. So, there must be something more than FEP by itself to explain consciousness. Both Friston and Solms called attention to learning but how do learning and consciousness connect? What is it about learning that needs consciousness or would create it? In a separate comment, I called attention to the yet previous post on electrical low frequency oscillations primarily in the Hydra but also in life in general. The paper discussed in that post does call out Friston as well as Buszáki in discussing low-frequency oscillations in the brain.

Let me speculate some and try to put some of the pieces together:

If the main goal of life is to maintain a stable internal state in the face of a changing external environment (part of FEP per my understanding), then maybe the best way of doing this is with electrical low-frequency oscillations. Think of a spinning top as a rough analogy. (Note: I understand the physics is not the same. This is an analogy.) The spinning top is stable but also in a slow motion of precession. If the top is tipped slightly, the top will adjust to account for change in gravity. There is also a rotational inertia that keeps the top stable when perturbed by outside bumps. Think now of an organism trying to maintain its form by using electrical oscillations to coordinate its different parts. There may be slow changes in form analogous to precession in the top and larger changes in response to external forces, but there is also a general resistance to change. The oscillations provide stability and structure to what might otherwise disintegrate into a dead, motionless mass. Hanson in the paper I referenced earlier conjectures that electrical low-frequency oscillations “may be the ultimate organism-wide electrical information integrators and communicators in all biological systems at all levels of scale, making them critical for maintenance of organism unity and coherent, adaptive behavior”.

As more complex organisms evolve, the same mechanisms are used resulting eventually in the development of brains and nervous systems with highly specialized cells involved in the low-frequency oscillations. These cells not provide for coordination in the body (basic things like heart and respiration rates), but also allow for movement and primitive reflexes in simple animals. According to the view of Simona Ginsburg and Eva Jablonka, as consciousness appears in more complex organisms, it correlates with complex learning.

Let me repeat a question I asked earlier: What is it about learning that needs consciousness or would create it? Or, to put it another way, what do we need to add to oscillating neurons to make complex learning possible?

Let me answer that it is simply memory that is the missing ingredient.

Solms makes his consciousness and feelings argument by saying (debatably) that feelings are always conscious. Let me make a similar statement. Consciousness does not exist when there is no ability to recall or create memories. When we are sound asleep and not dreaming, we cannot create memories. When we are completely anesthetized (assuming no resistance), we cannot create memories. Learning is impossible without an ability to store the results of the learning, and storing representations of sensual impressions, associations, motor actions is exactly what memory is.

Consciousness may be all about recalling memories, matching memories to current experience, and adjusting the internal memories to match the current experience. The matching of internal models with the external world is what FEP is all about. To be clear I am not talking about episodic memory (which does play a role with humans and maybe some other animals) but about more fundamental memory like memory we accumulated when we learned to see, walk, and ride a bicycle. I am also not suggesting that we might not be born with certain memories. Buszáki thinks that our brains come pre-wired with patterns of firing and consciousness is involved with matching and refining these patterns with experience of the real world. It is the same or similar process in either case. We are recalling patterns, matching them with the world, and refining them.

Consciousness exists to recall memories, match them with current experience, adjust them if necessary, and create new ones when required. The prediction and interference engine that is the brain could not do what it does with an ability to recall and store information. For some reason, consciousness is critical, maybe identical, to this process.

Posted in Consciousness, Friston, Memory | 55 Comments

The Hidden Spring

Posted on April 18, 2021 by James Cross

I’ve been following the views of Mark Solms for a while. Much of my original interest was in his views on psychoanalysis and Freud, especially aspects of Freudian theory for which evidence in neuroscience has gradually accumulated. His book The Brain and the Inner World (with Oliver Trumbull) was a great introduction to some of those issues. This interest was for a project which has since fallen by the wayside, perhaps never to be revived. I also quoted Solms quite a bit in previous post I did on learning and consciousness. So it was with particular interest when I found that he had written a more general work, The Hidden Spring, on the topic of consciousness.

<= DIY fountain with “hidden spring” in lower pot. Two pots and pump.

The book actually seems to be three books in one, although there is no tidy division in the actual structure of the book itself.

The first part of the book is somewhat autobiographical and deals with how he became a psychoanalyst and a neuroscientist. These two professions are not usually seen as compatible. Psychoanalysts talk to people on couches about their dreams and pry into the details of their childhood (at least in popular view). Neuroscientists look at EEGs MRIs, neurons, and brain structures. Anyone, but a behaviorist, who actually gives the matter some thought, would likely see some wisdom in making the approaches complementary rather than opposed. The domain of psychoanalysis is subjective experience as reported by the patient. Its fundamental work was the “talking cure“. The domain of the neuroscientist is the brain. It would seem logical that subjective experience, which we presume is generated by the brain, would tell us something about the brain and that activity in the brain might correlate with subjective experience. Mark Solms has spent most of his career at exactly this intersection of brain and experience, pioneering in a field he calls neuropsychoanalysis. He has written several places about how he thinks the “talking cure” actually works to change things in the brain. In this autobiographical part of the book, one of the most interesting things I learned was Solms’ pioneering work on dreams and how he validated to an extent Freud’s understanding of dreams as wish fulfillment: the activity in the brain during dreams is primarily in dopamine circuits, essentially the pleasure circuits of the brain. Freudian observations are scattered throughout the book but are for the most part not essential to the arguments of the other parts of the book.

The second part or aspect of the book is Solms’ view that consciousness originates in the brainstem. The brainstem, as one might suspect, is at the base of the brain, more or less at the top the spine, and below all of big chunks of the brain we see in the typical illustrations of the human brain. The brainstem is regarded as an evolutionary old part of the brain. It has a role in managing basic bodily functions like respiration, heart rate, and importantly the wake and sleep cycles. Solms has noted that damage to even a small number of certain neurons in the brainstem results in an irreversible coma. That observation, along with Solms’ argument that hydrocephalic children are conscious are key points in his argument that the brainstem is source of consciousness. They were initially persuasive to me when I read them a few years ago. Most neuroscientists, along with Freud himself, think consciousness to originate in the cortex. Of course, as humans with enormous cortexes, many of these neuroscientists may have a conscious or unconscious human bias in attributing consciousness to the cortex. We ourselves have large cortexes. Primates have relatively large cortexes. Mammals have relatively large cortexes. Everywhere we look where we see animals who might most likely be conscious we see relatively large cortexes compared to other animals. Naturally we might tend to think consciousness to be something coming from the cortex.

For Solms consciousness fundamentally consists in feelings. Frankly the more I’ve tried to understand this view the less I am persuaded by it. At one point, Solms explains that feelings are always conscious. He then goes on to quote Freud in support of the idea but the quote from Freud is about emotion, not feeling. He then states: “For now, let me be absolutely clear about what I mean by the term ‘feeling’: I mean that aspect of an emotion (or any affect) that you feel“. This is hardly any definition at all. It is saying “feeling” is what you feel. If we are equating, even partially, “emotion” and “feeling” then I can’t see how much that passes before my daily consciousness is emotion or feeling. Like anyone else, I have feelings and emotions but they do not dominate my day to day consciousness. I can’t see how, as I stare at a computer screen now and gather my words for this sentence, much emotion is involved with it. I do have a “feeling” of being like something in the Nagle sense but how that relates to emotion isn’t clear. Even more perplexing is why a Freudian would think that feelings must always be consciousness. It would seem unconscious feelings are persuasive through the Freudian menagerie of complexes and neuroses. Solms himself acknowledges on the same page with his definition that many psychoanalysts disagree with his view that feelings cannot be unconscious.

Let’s acknowledge that defining consciousness isn’t easy. It is something we all know from our experience but that experience consists of a great many things: memories, dreams, random thoughts, sights and sounds, pleasures and pains, and occasionally a good idea or two mingled among the bad ideas. The only commonality is a sense that we are something, know something, or we are experiencing something, although sometimes even the sense of being something is lost to what we are experiencing. If this is what Solms is saying, it improves little on Nagle’s definition and, by itself, gives no support to the brainstem theory. What I seem to be experiencing at any one point is usually a mixture of cognition, sensual impressions, and general feelings about my well-being. Lumping this together as “feeling” doesn’t seem to help to elucidate its nature. Similarly lumping it all together as some sort of cognitive image seems to leave out real emotions and feelings.

Defining consciousness as feeling is related to locating its origin with the brainstem. Solms finds evidence that hydrocephalic children with small, almost non-existent cortexes are conscious, so a brainstem locus for consciousness would be consistent with that evidence. The observation that irreversible coma results from damage to parts of the brainstem would be consistent as well. Locating consciousness in the most primitive part of the brain allows for the possibility that most animals would have some degree of consciousness and would eliminate a requirement for higher cognitive abilities that we associate larger cortexes. In other writings, Solms has compared to the cortex to the RAM of a computer. It has the ability to form the contents of consciousness but consciousness itself is like a flashlight originating in the brainstem that shines on the images in the cortex to make them conscious. The brainstem in effects adds the feeling to the content that is elsewhere in the larger brained animals.

Of course, placing consciousness in the brainstem and placing it in the cortex does not have to be an either/or proposition. If consciousness emerges as a distinct system whenever a certain critical mass (perhaps small in number) of neurons began to operate in a integrated manner with wave-like oscillations, then children with little or no cortex and small brained animals, such as insects and worms, could all be conscious to some degree. Humans with their large cortexes would be conscious as well but their consciousness would arise from a system that arises over a much larger mass of integrated and oscillating neurons. The brainstem in humans would not only be a participant in larger consciousness but also its chief enabler in its role in the wake cycle which, in turn, generates the transmitters that makes possible the faster, complex oscillations of the cortex that are associated with consciousness.

The third part of the book is all about Karl Friston’s free energy theories about the brain. Friston thinks that the brain (also cells and life in general) is constantly trying to minimize free energy. It is in a constant battle to maintain its internal order and homeostasis. To do this it must constantly learn and adapt to its external environment. The key to doing this is to minimize surprise, to keep internal predictions matching the external world. The brain is an inference engine that is constantly trying to reconcile the internal with the external. It can do this either by adjusting the internal model or taking actions in the external world to change it.

The theory is a good bit more complicated than that and buttressed by considerable mathematics. Friston himself, however, acknowledges it is not falsifiable.

Solms seems somewhat enthralled with the theory but I find it difficult to mesh with the rest of his theories. I can’t see, for example, how it relates to brainstem as origin of consciousness or consciousness as feelings. In fact, it seems much more like a cognitive theory with feelings playing primarily the role of gross indicator of congruity between the internal and external worlds. The theory does point to learning as primary explanation and role for consciousness. Yet Solms doesn’t reference Stephen Grossberg’s Adaptive Resonance Theory that explicitly explains that the “processes whereby our brains continue to learn about a changing world in a stable fashion throughout life are proposed to lead to conscious experiences”. Nor does it reference Simona Ginsburg and Eva Jablonka and their argument that complex learning as an evolutionary marker for minimal consciousness. Grossberg’s theory, in fact, seems to have come to most of the same conclusions as Friston about the brain but a number of years earlier and without some of mathematical baggage.

Solms now seems to have embarked on a quest to create artificial consciousness by implementing Friston’s theories. I’m dubious the effort will succeed and he will end up with a device that feels. But then, how will he know if he has succeeded or not?

Posted in Consciousness | 15 Comments