Max Tegmark has a new theory and everyone is thinking it over. I am confident that I will never feel comfortable with this theory. I will not be convinced, I fear. The whole thing seems to be an elaborate waste of time.
There is a problem with that great tool, mathematics. It is the same problem we have with language and logic. We take a situation and divide it into entities, more or less arbitrarily, hoping it is less rather than more arbitrary. We give these entities symbols and relationships so that we can play with them using rules that we are confident of. A description is created of the situation and now we have understanding of it. But wait – the description is only as good as its translation into symbols/relationships and translation back out. In other words, we have to have enough actual contact with the situation to do this translation well and to judge whether the process has been a success. Metaphorically, we have to play with the actual situation, hold it in our hands, look at it from all angles, and get a feel for it. Without that contact we will just do a ‘garbage in, garbage out’ exercise. I am not saying that there is anything wrong with mathematics or that it is not a magnificent tool, but even if the formulas are without error-free it does not mean that they have been interpreted reasonably.
What Tegmark appears to have done is to relate two very prominent unknowns: the why, how and even what of consciousness; and what the meaning/interpretation of quantum mechanics should be in terms of how the classic world relates to the quantum world. That sort of thing is great science when it is convincing. Theories like plate tectonics, DNA structure, atomic structure and evolution by natural selection were great because they tied up a lot of little problems and loose ends in one large solution. They became fundamental platforms for whole branches of science. It is also true that great changes are often done by new-comers or outsiders to the scientific area. They will not have been so indoctrinated that they cannot think outside the box.
But given the need to keep an open mind on new, unusual theories – I find this one just out to lunch. Tegmark’s paper (here) has a somewhat unfortunate idea of what consciousness is. Table 2 lists “Conjectured necessary conditions for consciousness that we explore in this paper. Principle of Information- A conscious system has substantial information storage capacity; Principle of Dynamics- A conscious system has substantial information processing capacity; Principle of Independence- A conscious system has substantial independence from the rest of the world; Principle of Integration- A conscious system cannot consist of nearly independent parts; Principle of Utility- A conscious system records mainly information that is useful for it; Principle of Autonomy- A conscious system has substantial dynamics and independence.” I find these principles unbelievable.
What is a conscious system as opposed to a system that has consciousness? Is there somewhere within our brains a conscious system along-side an unconscious one? Or is a whole person what is referred to?
Does consciousness alone have substantial information capacity? No, some include working memory in consciousness (but others don’t) but that is about the extent of the capacity. It has it’s contents at any moment but that could not be called substantial. Does consciousness alone have substantial information processing capacity? No, it has practically none. Has consciousness alone substantial independence from the rest of the world? No, it does not have much independence (none I would think) from the other processes in the brain. As far as ‘parts’ is concerned – they would have to be listed because it is not clear how it might be divided. I don’t think that consciousness has control over its contents, and those contents are probably useful to our whole bodies rather than just our consciousness. The idea that consciousness has autonomy is plainly not so. If we are talking about consciousness then we are talking about a series of momentary awarenesses/sharings of small packets of information across the brain. It has none of the principles that Tegmark lists. If he is talking about a system which has consciousness as one of its processes then we cannot draw a boundary short of the whole nervous system, in other words approximately the whole person.
But Tegmark is not talking about the whole nervous system. In his section about the integration paradox he is clearly talking about a much smaller process, like consciousness alone.
“This leaves us with an integration paradox: why does the information content of our conscious experience appear to be vastly larger than 37 bits? If Tononi’s information and integration principles from Section I are correct, the integration paradox forces us to draw at least one of the following three conclusions: 1. Our brains use some more clever scheme for encoding our conscious bits of information…2. These conscious bits are much fewer than we might naively have thought from introspection…3. To be relevant for consciousness, the definition of integrated information that we have used must be modified or supplemented by at least one additional principle.”
You can probably guess that it is 3 that he runs with and this leads deep into quantum mechanics, several dead ends and finally the theory. But the whole argument starts with a view of consciousness that is not in keeping with current research. Reading the paper this feels like reading Alice in Wonderland. So here is the theory (the abstract):
“We examine the hypothesis that consciousness can be understood as a state of matter, “perceptronium”, with distinctive information processing abilities. We explore five basic principles that may distinguish conscious matter from other physical systems such as solids, liquids and gases: the information, integration, independence, dynamics and utility principles. If such principles can identify conscious entities, then they can help solve the quantum factorization problem: why do conscious observers like us perceive the particular Hilbert space factorization corresponding to classical space (rather than Fourier space, say), and more generally, why do we perceive the world around us as a dynamic hierarchy of objects that are strongly integrated and relatively independent? Tensor factorization of matrices is found to play a central role, and our technical results include a theorem about Hamiltonian separability (defined using Hilbert-Schmidt superoperators) being maximized in the energy eigenbasis. Our approach generalizes Giulio Tononi’s integrated information framework for neural-network-based consciousness to arbitrary quantum systems, and we find interesting links to error-correcting codes, condensed matter criticality, and the Quantum Darwinism program, as well as an interesting connection between the emergence of consciousness and the emergence of time. ”
As well as the weird definition of consciousness, there is a reliance on the use of computer networks rather than cellular ones, and on the idea that processing is algorithmic. The only tools brought to bear on the question are those of information theory and quantum mechanics. These are, of course, relevant tools, but given the lack of understanding of how the brain works, it is not clear in what ways they may be relevant. They are certainly not the obvious tools to be using to understand consciousness. It seems positively perverse to restrict the problem solving to only those two tools. The metaphor of a new state of matter is somewhat confusing. In gases, liquids, and solids the atoms and molecules do not change, but only their interaction. So the idea is that there are atoms/molecules in the brain that interact slightly differently than ordinary interactions of matter. It sound a bit like vitalism, with the special type of matter that was once used to explain life. Here we have perceptronium, the special type of matter that explains consciousness. This seems a step backwards rather then forwards in our understanding. There is the nagging feeling that information is not the important thing. I feel we have elaborate nervous systems because of action, not because of perception. He seems to ignore what is known about the perception of objects and treats it as a mystery. There is no mention of oscillations in the brain and their function in communicating information. I could go on but won’t because to be truthful it is somewhat boring.
The upshot is that Tegmarks theory does not have the ring of truth about it for me. It doesn’t even sound vaguely truthy. Even further, I find it incomprehensible. It appears a silly exercise going from nowhere to nowhere.
However Tegmark’s mind must be very interesting. He is a theoretical physicist who is well known for his defense of multiverses, his mathematical universe hypothesis (the universe, each multiverse of it, is literally a mathematical structure - not just well described by mathematics but actual just the mathematics all by itself) and now his consciousness is a state of matter theory. All three things I cannot really even conceive of, let alone accept.