What's new in
TGD Inspired Theory of Consciousness
Note: Newest contributions are at the top!
I had an "entertaining" discussion with two fellows - l call them A and B -, which taught a lot, I hope also for A and B -, and actually gave a good example of two kinds of learning. Learning by conditioning and learning by discovery. It also led to a possible understanding about what goes wrong in what I would call ultra-skeptic cognitive syndrome.
[This discussion by the way gave me good laughs. A and B - first summarized his academic background by "studied strings" and second was Bachelor in computer science but pretending to be M-theorist. They tried to demonstrate that I am a crackpot. They carried out an "investigation" following the principles of investigations made for witch candidates at middle ages. The victim had two options: she drowns or not in which case she is burned at stake.]
The highly emotional discussion was initiated by a totally non-sense hype about transferring consciousness of C Elegance to computer program (see this). I told that the news was hype and this raised the rage of A and B. The following considerations have very little to do with this article. Note however that I have done some work AI in general and even with with the basic ideas of deep learning. For instance, we had two years ago a collaboration about AI, IIT approach to consciousness, and about a possible connection with remote mental interactions together with Lian Sidorov and Ben Goertzel, who is behind Sophia robot. There two chapters related to this (see this and this). I think that the latter chapter is published in a book by Goertzel. There is also a critical article inspired by Sophia robot about which Ben Goertzel wrote an enthusiastic article and sent to Lian Sidorov and me (this).
1. Learning by conditioning
Returning to learning. The first kind of learning is learning by conditioning, which deep learning algorithms try to mechanize. Second kind of learning is learning by discovery. The latter is impossible for computers because they obey deterministic algorithm and are unable to do anything creative.
Emotions play a strong role in the learning by conditioning in the case of living systems and in the simplest form it is learning of X-good and X-bad type associations helping C elegance to survive in the cruel world. In case of humans this kind of associations can be extremely dangerous as for instance the course of events in USA has shown.
Very large part of our learning is just forming of associations: this is what Pavlov's dogs did. In school we learn to associate to "2×3=" symbol "6". In our youth we learned also algorithms for sum, division, multiplication and division, and even for finding the roots second order polynomial. Often this is called learning of mathematics. Later some mathematically gifted ones however discovered that this is just simple conditioning of an algorithm, and has very little to do with genuine mathematical thinking. The discovery of the algorithm itself would be mathematical thinking. The skill to code for algorithm - usually given - is also an algorithm and it can be also coded in AI.
If we are good enough in getting conditioned we get a studentship in University and learn science. This involves also learning of simple conditionings of type X-good and X-bad. In this learning social feedback from others reinforces learning: who would not like to earn the respect of the others!
For X-bad conditionings X can be homeopathy, water memory, cold fusion, telepathy, remote viewing, non-reductionistic/non-physicalistic world view, quantum theories of consciousness, TOEs other than M-theory, etc... For X-good conditionings X can be physicalism, reductionism, strong AI, superstrings, Witten, etc...
The student learns also to utter simple sentences demonstrating that he has learned the desired conditionings. This is important for career. Proud parents who hear the baby say their first word encourage the child. In the same manner environment reinforces the learning of "correct" opinions by a positive feedback. The discussion with A and B ga a quite a collection of these simple sentences. "I guessed that he is a crank" from A is a good example intended to express the long he life experience and wisdom of the youngster.
These conditionings make it also easy "recognize" whether someone is a crank/crackpot/etc... and even to carry out personal investigations - analogous with witchcraft investigations at middle ages - whether some-one is a crank or not. This is what A and B in their young and foolish arrogance indeed decided to carry out.
2. Learning by Eureka experience
There is also second kind of learning. Learning by discovery. Computers are not able to do this. I mentioned in the discussion what happens when you look certain kind of image consisting of mere random looking spots in plane. After enough staring suddenly a beautiful 3-D patterns emerges. This is a miracle like phenomenon, Eureka experience. Quantum consciousness based explanation is the emergence of quantum coherence in the scale of the neuronal cognitive representation in visual cortex at least. New 3-D mental image emerges from purely 2-D one. One goes outside of the context.
The increase of dimension might provide an important hint about what happens more generally: and this would indeed occur for the dimension of extension of rationals in Eureka quantum jump in TGD based model of what could occur. Physically this would correspond to the increase of the effective Planck constant heff= n×h0, h=6×h0 assignable to the mental image created by the image. n is indeed the dimension of extension of rationals and would increase and also scale of quantum coherence would increase from that of single spot to that for the entire pictures.
This kind of learning by Eureka is probably very common for children: they are said to be genii. Later the increasing dominance on the learning by conditioning often eliminates this mode of learning and the worst outcome is a mainstream scientist who is hard-nosed skeptic. Solving genuine problems is the manner to gain these learning experiences but they come only now and then. Some of them are really big: during my professional career there have been - I would guess about 10 really big experiences of this kind involving discovery of a new principle or totally new physical idea.
3. How to understand what is wrong with vulgar skeptics?
The discussion was very interesting since it led me to ponder why it is so hopeless to explain something extremely simple for skeptics. There is a beautiful connection with a learning based on Eureka experience. Physically this corresponds in TGD to a phase transition increasing scale of quantum coherence and algebraic complexity: more technically effective Planck constant heff increases at some levels. More intelligent mental images become possible and Eureka experience happens as in the situation when chaotic 2-D set of points becomes beautiful 3-D object.
Biological evolution at the level of species is based on this: we humans are more intelligent than banana flies. This evolution occurs at all levels - also at the level of individuals but it is not politically correct to say this aloud. Some of us are in their intellectual evolution at higher level than others, either congenitally or by our own efforts or both. This creates of cause bitter feelings. Intellectual superiority irritates and induces hatred. This is why so many intellectuals spend most of their life in jail.
Take seeing as an example. If person has become blind at adult age, he understands that he is blind and also what it feels to see. Also congenitally blind person believes that he is blind: this because most people in his environment tell that it is possible to see and that he is blind. He does not however feel what it is to see. Suppose now that most of us are blind and then comes some-one and tells that he sees. How many would believe him? They cannot feel what it to see. Very probably they conclude that this fellow is a miserable crank.
Suppose now that certain person - call him MP - has used 4 decades to develop a TOE based on generalization of superstring model made 5 years before the first superstring evolution and explaining also consciousness. MP tries his best to explain his TOE to a couple of skeptics but finds it hopeless. They even arrange "investigation" following the best traditions of witch hunt to demonstrate his crackpotness. And indeed, they conclude that they were correct: all that this person writes is totally incoherent non-sense just as this 2-D set of random points.
These two young fellows are simply intellectually blind since their personal hierarchy of Planck constants does not contain the required higher values. A Eureka experience would be required. MP could of course cheat and tell that he believes in superstrings and give a hint that the is a good friend of Witten. This would help but would only lead to pretended understanding. The fellows would take MP seriously only because MP agrees with Witten and claims to be a friend of Witten but still they would not have a slightest idea what TGD is. They cannot feel what it is to understand TGD.
The only hope is personal intellectual evolution increasing the needed Planck constants in the personal hierarchy. This is possible only if these fellows admit that they are intellectually blind in some respects but if they are young arrogant skeptics they furiously deny this and therefore also the possibility of personal intellectual evolution.
For background see the chapter Conscious Information and Intelligence. See also the article Two manners to learn and what goes wrong with vulgar skeptics?.
In Facebook I received a link to an interesting popular Science News article titled A New Information Engine is Pushing the Boundaries of Thermodynamics. The article told about the progress in generalizing the conventional second law of thermodynamics to take information as an additional parameter.
Carnot engine is the standard practical application. One has two systems A and B, both in thermal equilibrium but with different temperatures TA and TB ≥ TA. By second law one has heat flow Q from A to B the two systems, and Carnot's engine transforms some of this heat to work. Carnot's law gives an upper bound for the efficiency of the engine as η= W/Q ≤ (T2-T1)/T2. The possibility to transform information to work forces to generalize Carnot's law.
Since information is basically conscious information, this generalization is highly interesting from the point of view of quantum theories of consciousness and quantum biology. Certainly the generalization is highly non-trivial. Especially so in standard physics framework, where only entropy is defined at fundamental level and is regarded as ensemble entropy and basically has very little to do with conscious information. Therefore the argumentation is kind of artwork.
1. Maxwell's demon in its original form
Maxwell's demon appears in a thought experiment in which one considers a system consisting of two volumes A and B of gas in thermal equilibrium at same temperature. At the boundary between A and B having a small hole sits a demon checking whether a molecule coming from A has velocity above some threshold: if so it allows the molecule to go to B. Demon monitors also the molecules coming from B and if the velocity is below the threshold it allows the molecule to continue to A. As a consequence, temperature and pressure differences develop between A and B. Pressure difference can do work very much voltage between the cathode and anode of battery. One can indeed add a tube analogous to wire between ends of the entire system and pressure difference causes a flow of mass doing thus work: one has pump.
The result is in conflict with the second law and one can ask what goes wrong. From the Wikipedia article one learns that a lot of arguments have been represented con and pro Maxwell's demon. Biologist might answer immediately. Demon must measure the states of molecules and this requires cognition and memory, which is turn require metabolic energy. When one takes this into account this, paradox should disappear and second law should remain true in a generalized form in which one takes into account the needed metabolic energy.
2. Experimental realization of Maxwell's demon
The popular article describes an experiment actualizing Maxwell's demon carried out by Govind Paneru, Dong Yun Lee, Tsvi Tlusty, and Hyuk Kyu Pak . Below is the abstract of the article Lossless Brownian Information Engine published in Phys Rev Letters (see this).
We report on a lossless information engine that converts nearly all available information from an error-free feedback protocol into mechanical work. Combining high-precision detection at a resolution of 1 nm with ultrafast feedback control, the engine is tuned to extract the maximum work from information on the position of a Brownian particle. We show that the work produced by the engine achieves a bound set by a generalized second law of thermodynamics, demonstrating for the first time the sharpness of this bound. We validate a generalized Jarzynski equality for error-free feedback-controlled information engines.
Unfortunately, the article is behind paywall and I failed to find it in arXiv. The popular article uses notions like "particle trapped by light at room temperature" and photodiode as "light trap" without really defining what these expressions mean. For instance, it is said that the light trap would follow particles moving in definite direction (from A to B in Maxwell's thought experiment). I must admit that I am not at all sure what the precise meaning of this statement is.
3. TGD view about the situation
TGD inspired theory of consciousness can be regarded as a quantum measurement theory based on zero energy ontology (ZEO) and it it is interesting to try to analyze the experiment in this conceptual framework.
3.1 TGD view about the experiment
The natural quantum interpretation is that the photodiode following the photon is performing repeated quantum measurements, which in standard quantum theory do not affect the state of the particle after the first measurement. From the viewpoint of TGD inspired consciousness, which can be regarded as a generalization of quantum measurement theory forced by zero energy ontology (ZEO), the situation could be as follows.
Zero energy ontology (ZEO) forces to consider different arrangement producing energy somewhat like in perpetuum mobile but not breaking the conservation of energy in any obvious manner. The idea pops into my mind occasionally and I reject it every time and will do so again.
In ordinary thermodynamics energy should be fed from environment to increase < E>: how environment would enter into the game now?
ZEO based theory of consciousness (see this) forces to consider also a genuine breaking of the second law.
TGD also predicts what I call Negentropy Maximization Principle (NMP) .
The hierarchy of Planck constants heff/h=n is a further key notion in TGD inspired quantum biology.