I think that there is no such thing as "human mind", since each human has different brain, even if we consider only one particular human, his/her brain is continously changing during the natural aging process and the interaction with its environment. His/her brain can be active in different parts, triggered by its surrounding enviroment that is also hard to describe with deterministic events. It also depends on how much neurotransmitter materials are already in it and how much supply will be available during the production of a mental function. These are just few examples how much data has to be considered during a mental function of a real human brain. Therefore my assumption is the following: only the quantum computation can be considered to be effective enough to simulate such an environment with such amount of connections and neuro transmission actions (also not to forget the pre-training/initialization of the simulated brain).

Maybe a function like "modelling" (understanding something: its structure) might be dificult to describe through an algorithm for the simultaneous verification of two circunstances: 

-Modelling envolves creativity and abstraction with a sort of fuzzy matching between concepts to reach a satisfactory consistency degree dificult to describe in binary terms. Also, if this leads to a discovery, since it is new, an algorithm has nothing to compare it with. It might say its novel and correct, but it probably cannot say its relevant.

-Modelling also envolves decision making, which requires the participation of emotions (according to the somatic marker hypothesis - see Damasio's work). So it brings the problem outside the strict scope of rational behaviour, making it more difficult to mimic.

You might be interested in the work of Peter Naur (Turing award) - his synapse state theory of mental life: 

http://www.naur.com/synapse-state.pdf

There´s a more recent article that he gaved me himself but I think I only have it in paper. But I can try to get it for you somehow, if you are interested.

I am not an expert in this topic. There are probably some Artificial Intelligence issues connected to this question. I am just guessing.

Anyway, I hope I am right, otherwise lots of Research Gate people might be out of work in the future :-)

Good luck.

Attila makes an interesting point in that it would be very hard to capture all of the properties of a specific person's mind and reproduce experience, personality etc.

However, if you take a scientific, reductionist approach then it should certainly be possible to model something that behaves like a human mind. How to do it is very easy to state at the highest level, whilst practical implementation looks infeasible for the forseeable future. That method is simply (?) to model all of the individual physical / chemical processes involved in the whole human brain / nervous system.

Such a machine would then have to have its parameters set to some state that represented an accumulation of knowledge and experience at some stage in a human lifespan. Alan Turing suggested the best way to do this would be to create a machine with the best possible sensors and locomotion systems added to its sensory and computation system, then release it into the world to learn from experience. 

The  only philosophical argument that could be made against this stance is that humans have a metaphysical  element (soul if you will) which cannot be modeled by physics and chemistry. Not an argument that I would like to make as a scientist.

@Fernando,

Thank you for interesting lecture. I am interested learn even more if possible.

Anyway, synapses states analyze could be possible algorithm for brain process. But, what I am asking for, is the algorithm not for the brain as material object, but for the mind as the function of the brain. The question is, if we even will find the algorithm for neural network or Neural Modeling Fields (following Perlovsky idea of Dynamic Logic), still we have no algorithm for our thoughts, our emotions, our motivations etc.

Ignoring technical problems, is that theoretically conceivable to find the algorithm for all these phenomena? 

If not, how we can imagine strong AI without algorithms we can write for Her?

Oh! you want an algorithm for Strong A.I. I am afraid that the only algorithms available are for weak A.I. and are biologically inspired. Should we even call it "Mind"? is a serious question.

Wieslaw, trying to answer your last post question: "Ignoring technical problems is that theoretically conceivable to find the algorithm for all these phenomena" my opinion is that humans will hardly be able to do it. Elaborating on the important metaphysical argument mentioned by James, if you conceive the existence of a God, within Theology you learn that God shapes the lives of those who believe in and somehow try to follow him, which means that in the mind function of these people you have to consider the participation of a divine (not exclusively human) element.

Since by definition (definition of God) the human mind cannot capture God completely, such divine element (or part of it) might be out of reach of a human algorithm, maybe because it defies human logic (if there is one...) or because it has attributes of "infinity" while the algorithm would have to be finite. Sorry if this does not sound very "scientific" but to me Theology is also a science.

I don't mean you cannot understand, I just mean that you probably cannot capture and reduce your understanding completely to the form of an algorithm. But once more, this is just my opinion and I am no expert.

Good luck.

@Attila,

You have wrote: "His/her brain can be active in different parts, triggered by its surrounding environment that is also hard to describe with deterministic events."

Brain processes are basically nonlinear and consequently chaotic and nondeterministic. Otherwise, we will have no free will. Therefore they are well prepared to reflect nondeterministic environment.  But because the same reason possibly they can't be described by an algorithm. Aren't they?

@Wieslaw,

today I only see one field of science that is well prepared to deal with this kind of non-deterministic behaviour, this is the Universal Quantum Computer or we can call it simply: quantum algorithm (see e.g. http://math.nist.gov/quantum/zoo/ ), since if you measure a qubit register then the outcome of the measurement is non-deterministic but it has probability amplitudes to be zero or one.

So theoretically it is possible to simulate this functionality, but it is so hard that in practice I do not see any chance to see such complex quantum algorithm in the next ten years...

I don't want to seem confusing, but the problem is that we really don't know how nature achieved it, and any attempt to make an algorithm now is probably doomed to failure, however there are ways that we can think of to get the information, from neuroscience eventually, and when that is achieved, then, and only then will we be able to evaluate what algorithms are necessary.

Temporary, I don't care if such algorithm is feasible or not. This is philosophical question only. From one hand, if the consciousness is the function of matter such algorithm should be possible, because one is thinking all what matter can produce can be described by algorithm. From the other hand, if we can't describe our qualia by symbolic, formal language, therefor we can't write any algorithm for consciousness, because qualia are the very basic part of consciousness. What is the real solution for this dilemma?

Does the statement "all what matter can produce can be described by algorithm" really prove to be correct? What limits are you placing on "Matter" that this is so. Is a quantum matter? What algorithm describes action at a distance, or entanglement?

I don't really believe in this statement: "all what matter can produce can be described by algorithm". There is probably no algorithm for Big bang, for evolution and for life creation. For that goal Laplace's Demon would be necessary. But computer scientist claim very often that Touring Machine can do everything, what physically exist and what we can think about. This is possibly the question of definition.

Knowing how computer and algorithms work I see no uniqueness of my mind and theirs, including bugs and infinite loops in some threads.

It is a matter of processing power, organisation of memory, evolutionary changes in some algorithms, awareness of the world from sensors.

What computers usually do not have is a deep awareness of themselves to every bit of their chassis. They generally do not move reproduce search for food partners have emotions. To develop social skills you cannot program them. You must have an adaptive algorithm in them linked to their survival and let them live and reproduce themselves. I see no limits, just technical difficulties and cost.

Bear in mind that by giving them algorithms you pass on human intelligence to them.

The problem I have in trying to answer this question is what is meant by human mind?

When we have an experience of dreaming, is this a facet of the human mind? If so, why?

When the human mind is capable of comprehending infinity, and even larger infinities (Cantor), and can visualise boundaries to one infinity (C) then the challenge of a mathematical model become larger.

Lastly, the question is perhaps fraught with a more fundamental failure, for to completely model the human mind, by an algorithm created by the human mind would be to completely know a system from information gathered entirely from within the system ! 

Dear Colin,

The issues covered in your post, reflect the breadth of issues studied by cognitive science. The logic and rules of formal languages, which include mathematics, well describe problems of idealized, abstract models of material phenomena, and subjective thought processes relating to these models and formulated in a symbolic language.

However, for the understanding of how consciousness arises in our minds, it is crucial to understand how you can follow categorization, generalization and idealization perceived objects. These objects, formed through the senses and associative processes, create increasingly more abstract models of the perceived reality. The transition from qualia to concepts, in the general case, are not subject of any logic. In specific cases, when we have the possibility to pre-define the perceived patterns, there are many effective methods of recognition developed. A proposed by you modal logic reveals his power to further their generalizability, assessing the phenomena which can not undergo the formal description in symbolic language.

However, such analyzes are not available to our animal cousins, because they do not use complex language. But we do not refuse them the awareness and the ability to build simpler models of reality. I have written here before, that formalizing the process of transcription qualia to the concepts is prohibited by the curse of combinatorial complexity, so I will not discuss it again.

Our neural networks do that in much easier way, comparing the similarity of new models to the existing patterns previously generated in the course of learning. This is done through direct physical-chemical process with similar complexity as the perceived world. This enables us to reflect the reality in the brain. (See NeuroElectroDynamics, which I already evoked here). The process of categorizing causes compression of information, which enables its further processing in a simple, symbolic form.

The repeatable experiments of D Hawkins show that consciousness can be measured, on a scale of 1 - 1000, consciousness is best measured on a logarithmic scale. 

The work at PEAR lab (Princeton) show how conscious intent can be stored, transported and then released, again, quite interesting science

As in all science, empirical findings provide useful value to any theory.