Seven marginal hypotheses about the nature of emotions
According to popular belief, technological singularity will arise immediately after the creation of artificial intelligence, entirely superior to the intelligence of man.
The most obvious step on this path is to understand the principles of the structure of an already functioning mind of animals and humans.
Most of the functions of our mind can be reproduced quite well using a computer: a computer can solve problems using an embedded algorithm, analyze, recognize visual images, recognize and generate speech, answer questions, etc. In many ways, it is even worse than humans and animals, but there is progress.
')
However, there is something in which a computer is powerless - it is an ability for emotions. This is not about their external manifestation, but about the internal component. No progress, not the slightest hope for at least some primitive reproduction of them. Generally zero.
Why is this so important? When solving a research task, one must “hit on all fronts”. Any detail that is not even important at first glance can serve as a clue and help to understand the whole picture, to solve a riddle.
According to the formal model of emotions of OSS , the simplest kinds of emotions are joy and grief. In other words, pleasure and suffering.
So, setting a goal: create a device capable of feeling pain and / or pleasure. In this case, it is not at all necessary that this pain or pleasure be manifested in any way for an external observer.
This article contains 7 hypotheses about how this can be done, including the most delusional. It can be considered a brainstorming, so I will be grateful for new views in the comments.
But first, the analysis of myths:
Whenever I discuss this issue, there are people who are deeply convinced that the nature of emotions has long been known - this is just chemistry, namely the hormone dopamine. Such reasoning, spoken with the confidence of Sharikov, plunges me into despondency.
Dopamine is a neurotransmitter. A neurotransmitter is the substance through which nerve impulse is transmitted. Everything! He does nothing more.
Neurotransmitters, including dopamine, only transmit a signal. Launch some mechanism. By themselves, neurotransmitters are not this mechanism, and the flask with dopamine will not feel any emotions. The mechanism can also be triggered in another way, for example, by connecting the electrode to a specific area of ​​the brain .
It is quite suitable for external manifestation, for example, for robotics. You can create a robot that will smile when pleasure takes a positive value greater than n, or cry when pleasure takes a negative value.
But the outward manifestation has nothing to do with the inner feeling of pain or pleasure. The actor may scream in pain, but in fact he will not be hurt.
So, if we remove the external manifestation, what do we have left? And nothing. A value that does not have any objective value to pain or pleasure. Only in your head will remain the association that positive values ​​are pleasure, and negative values ​​are pain. From the point of view of the mechanism there will be no difference between pain and pleasure, and, therefore, they do not exist.
I will give an analogy. You turned on the computer, started the program SETI @ home. How do you know that the program actually does useful calculations, and does not just load the processor and display imaginary progress? How to distinguish a real program that actually performs the calculations from the program that only does the kind of calculations? This can be done, for example, by examining the source code of the program.
A certain version of the program may not at all display progress (it doesn’t show itself at all for an external observer, except to load the processor). However, in this case it is possible to distinguish the present calculations from the imaginary ones.
Something similar with emotions. Knowing their device, you can say whether one or another mechanism is experiencing emotions, or just displays their manifestation (“imaginary progress”, as part of the example above).
The fact is that all basic concepts are not derivable. One example is the mathematical set. Even such an exact science as mathematics uses terms that have no formal definition.
Try to determine the time or space in physics, the number in mathematics, etc.
So, now options like hypothetically, you can make the simplest device that can experience emotions.
The classical theory describes a neuron as a computing device, the functions of which can be recreated on a classical computer (to the extent that it replaces a living neuron with its artificial counterpart).
The mind consists of neurons. Therefore, the desired pain or pleasure is a kind of program (albeit a very complex one). When performing this program, a feeling of pain or pleasure arises.
Then the task will take the form: to find the minimum algorithm, when executed, the performing device will have a feeling of pain. Unless, of course, the nature of pain is data processing.
The search direction has already been set: the creation of the Blue Brain Project model, the Human Brain Project, etc. Of course, the projects are promising and very important. If suddenly nothing happens, you will be able to turn to marginal hypotheses.
Emergence - the presence of any system of special properties that are not inherent in its subsystems and blocks. One tree is not a forest, the accumulation of individual cells is not an organism.
In this case, the task will take the form: to find the minimum number of subsystems necessary to create a system capable of feeling pain or pleasure. The question of the nature of each subsystem remains open (are all subsystems the essence of computers).
Among the founding fathers of this hypothesis is Roger Penrose. He is not some kind of freak, but a quite authoritative scientist.
Since quantum computations are also computations, the problem is entirely analogous to that defined for hypothesis 1. The only difference is that the quantum algorithm on a classical computer cannot be completed. But this does not change anything: if pain is the execution of an algorithm, even if it requires a quantum computer, bring me this algorithm.
It follows from the electromagnetic theory of consciousness.
In principle, nothing is cast from hypothesis 1, only is its variation (except for synapses and axons, electromagnetic interaction is added).
Why does a magnet attract a nail in the distance? That is his property. And it follows from the properties of the electron: the electron is the minimum magnet. You can magnetize a steel object - for this you need to arrange its electrons using a strong magnetic field. There is no deeper explanation.
Perhaps some elementary particle has the ability to feel, but this ability requires special conditions for its manifestation.
Why does the firefly glow? Of course, it emits photons, elementary particles. But this requires certain conditions.
In this case, the task will take the form: to find a condition in which the material body becomes able to feel pain or pleasure.
Each physical body has a length, width, and height. There is a time. They can be called the basic properties of our world.
It can be assumed that the basic property of our world is also the ability to sense. But as for the detection of the passage of time, certain conditions are necessary (for example, the presence of a clock), and for the discovery of such a property of the world as a sensation - certain conditions are needed.
The task will be: to create such conditions under which the ability of the world to sense is revealed.
In some cases, physics goes by the wayside ... You can study how the transistor works, assemble an electromagnetic oscillator. However, if your generator with a piece of wire as an antenna will operate at a frequency of 156.8 MHz (not the worst option), strangers may appear soon and send you to a bad place.
In addition to the physical quantity, there are also state laws. From the point of view of physics, the frequency of 156.8 MHz is unremarkable, and from the point of view of state laws, this frequency is very important in the maritime business.
If we assume that in addition to us in the vast universe there is another intelligent life that is at a higher level of development, then it is quite likely that the secret of the arrangement of emotions (and, consequently, of consciousness) is the most protected. And the brain is no more than a front-end, and the most interesting thing is hidden in another place, like gold in the impregnable Fort Knox.
In this case, the task will look like: how can you transmit a signal above the speed of light? Some kind of effect may be used, about which we do not yet know (or variations with quantum entanglement).
And although this does not negate the fact that the nature of emotions is reproducible, no one wants to play with fire.
It can be assumed that every particle of our world is just a computer simulation. In this case, if consciousness (and emotions) arises within the framework of a computer model, then the study of its structure does not change (that is, for us there is no difference).
If the consciousness (and emotions) are in a non-emulated world, and they are thrown into our world with the help of a certain bridge, then there is a possibility that it is impossible in principle to study them. It is even possible that consciousness is not realizable within the framework of our world.
In this case, the task will take the form: to find a bridge through which information is transmitted between the computer simulation (our world) and the real world. This bridge can be, for example, a random number generator .
For developers it is very easy to imagine the idea of ​​open individualism. On your computer there can be 1 processor (and 1 core), but at the same time there are a dozen windows open and each of them works independently, it seems that all of them work at the same time. Achieved by interrupt mechanism.
By analogy, each person is like an open program, and a kind of universal "I" is like a processor, which in turn "finds" itself in one of the programs.
This is consistent with hypotheses 6, 7, 8 (that is, open individualism is possible within the framework of these hypotheses, but does not necessarily follow from them).
The article is written in the style of brainstorming without the use of substances. Some ideas can be immediately rejected, but for completeness we will not do that. It will be interesting to hear alternative hypotheses that you know.
The most obvious step on this path is to understand the principles of the structure of an already functioning mind of animals and humans.
Most of the functions of our mind can be reproduced quite well using a computer: a computer can solve problems using an embedded algorithm, analyze, recognize visual images, recognize and generate speech, answer questions, etc. In many ways, it is even worse than humans and animals, but there is progress.
')
However, there is something in which a computer is powerless - it is an ability for emotions. This is not about their external manifestation, but about the internal component. No progress, not the slightest hope for at least some primitive reproduction of them. Generally zero.
Why is this so important? When solving a research task, one must “hit on all fronts”. Any detail that is not even important at first glance can serve as a clue and help to understand the whole picture, to solve a riddle.
According to the formal model of emotions of OSS , the simplest kinds of emotions are joy and grief. In other words, pleasure and suffering.
So, setting a goal: create a device capable of feeling pain and / or pleasure. In this case, it is not at all necessary that this pain or pleasure be manifested in any way for an external observer.
This article contains 7 hypotheses about how this can be done, including the most delusional. It can be considered a brainstorming, so I will be grateful for new views in the comments.
But first, the analysis of myths:
Myth 1: Pleasure is the hormone dopamine.
Whenever I discuss this issue, there are people who are deeply convinced that the nature of emotions has long been known - this is just chemistry, namely the hormone dopamine. Such reasoning, spoken with the confidence of Sharikov, plunges me into despondency.
Dopamine is a neurotransmitter. A neurotransmitter is the substance through which nerve impulse is transmitted. Everything! He does nothing more.
Neurotransmitters, including dopamine, only transmit a signal. Launch some mechanism. By themselves, neurotransmitters are not this mechanism, and the flask with dopamine will not feel any emotions. The mechanism can also be triggered in another way, for example, by connecting the electrode to a specific area of ​​the brain .
Myth 2: to reproduce pleasure and suffering, you just need to enter the variable pleasure
It is quite suitable for external manifestation, for example, for robotics. You can create a robot that will smile when pleasure takes a positive value greater than n, or cry when pleasure takes a negative value.
But the outward manifestation has nothing to do with the inner feeling of pain or pleasure. The actor may scream in pain, but in fact he will not be hurt.
So, if we remove the external manifestation, what do we have left? And nothing. A value that does not have any objective value to pain or pleasure. Only in your head will remain the association that positive values ​​are pleasure, and negative values ​​are pain. From the point of view of the mechanism there will be no difference between pain and pleasure, and, therefore, they do not exist.
Myth 3: It is impossible to know whether the mechanism is experiencing pain or pleasure if it does not manifest them.
I will give an analogy. You turned on the computer, started the program SETI @ home. How do you know that the program actually does useful calculations, and does not just load the processor and display imaginary progress? How to distinguish a real program that actually performs the calculations from the program that only does the kind of calculations? This can be done, for example, by examining the source code of the program.
A certain version of the program may not at all display progress (it doesn’t show itself at all for an external observer, except to load the processor). However, in this case it is possible to distinguish the present calculations from the imaginary ones.
Something similar with emotions. Knowing their device, you can say whether one or another mechanism is experiencing emotions, or just displays their manifestation (“imaginary progress”, as part of the example above).
Myth 4: until a formal definition of pain / pleasure is given - it makes no sense to talk about anything
The fact is that all basic concepts are not derivable. One example is the mathematical set. Even such an exact science as mathematics uses terms that have no formal definition.
Try to determine the time or space in physics, the number in mathematics, etc.
So, now options like hypothetically, you can make the simplest device that can experience emotions.
Classical theory: computer program
The classical theory describes a neuron as a computing device, the functions of which can be recreated on a classical computer (to the extent that it replaces a living neuron with its artificial counterpart).
The mind consists of neurons. Therefore, the desired pain or pleasure is a kind of program (albeit a very complex one). When performing this program, a feeling of pain or pleasure arises.
Then the task will take the form: to find the minimum algorithm, when executed, the performing device will have a feeling of pain. Unless, of course, the nature of pain is data processing.
The search direction has already been set: the creation of the Blue Brain Project model, the Human Brain Project, etc. Of course, the projects are promising and very important. If suddenly nothing happens, you will be able to turn to marginal hypotheses.
Hypothesis 1: Emergence
Emergence - the presence of any system of special properties that are not inherent in its subsystems and blocks. One tree is not a forest, the accumulation of individual cells is not an organism.
In this case, the task will take the form: to find the minimum number of subsystems necessary to create a system capable of feeling pain or pleasure. The question of the nature of each subsystem remains open (are all subsystems the essence of computers).
Hypothesis 2: quantum computing
Among the founding fathers of this hypothesis is Roger Penrose. He is not some kind of freak, but a quite authoritative scientist.
Since quantum computations are also computations, the problem is entirely analogous to that defined for hypothesis 1. The only difference is that the quantum algorithm on a classical computer cannot be completed. But this does not change anything: if pain is the execution of an algorithm, even if it requires a quantum computer, bring me this algorithm.
Hypothesis 3: Electromagnetic Field
It follows from the electromagnetic theory of consciousness.
In principle, nothing is cast from hypothesis 1, only is its variation (except for synapses and axons, electromagnetic interaction is added).
Hypothesis 4: the basic property of matter
Why does a magnet attract a nail in the distance? That is his property. And it follows from the properties of the electron: the electron is the minimum magnet. You can magnetize a steel object - for this you need to arrange its electrons using a strong magnetic field. There is no deeper explanation.
Perhaps some elementary particle has the ability to feel, but this ability requires special conditions for its manifestation.
Why does the firefly glow? Of course, it emits photons, elementary particles. But this requires certain conditions.
In this case, the task will take the form: to find a condition in which the material body becomes able to feel pain or pleasure.
Hypothesis 5: the basic property of our world
Each physical body has a length, width, and height. There is a time. They can be called the basic properties of our world.
It can be assumed that the basic property of our world is also the ability to sense. But as for the detection of the passage of time, certain conditions are necessary (for example, the presence of a clock), and for the discovery of such a property of the world as a sensation - certain conditions are needed.
The task will be: to create such conditions under which the ability of the world to sense is revealed.
Hypothesis 6: the brain as a transceiver
In some cases, physics goes by the wayside ... You can study how the transistor works, assemble an electromagnetic oscillator. However, if your generator with a piece of wire as an antenna will operate at a frequency of 156.8 MHz (not the worst option), strangers may appear soon and send you to a bad place.
In addition to the physical quantity, there are also state laws. From the point of view of physics, the frequency of 156.8 MHz is unremarkable, and from the point of view of state laws, this frequency is very important in the maritime business.
If we assume that in addition to us in the vast universe there is another intelligent life that is at a higher level of development, then it is quite likely that the secret of the arrangement of emotions (and, consequently, of consciousness) is the most protected. And the brain is no more than a front-end, and the most interesting thing is hidden in another place, like gold in the impregnable Fort Knox.
In this case, the task will look like: how can you transmit a signal above the speed of light? Some kind of effect may be used, about which we do not yet know (or variations with quantum entanglement).
And although this does not negate the fact that the nature of emotions is reproducible, no one wants to play with fire.
Hypothesis 7: our world - computer simulation
It can be assumed that every particle of our world is just a computer simulation. In this case, if consciousness (and emotions) arises within the framework of a computer model, then the study of its structure does not change (that is, for us there is no difference).
If the consciousness (and emotions) are in a non-emulated world, and they are thrown into our world with the help of a certain bridge, then there is a possibility that it is impossible in principle to study them. It is even possible that consciousness is not realizable within the framework of our world.
In this case, the task will take the form: to find a bridge through which information is transmitted between the computer simulation (our world) and the real world. This bridge can be, for example, a random number generator .
Bonus: open individualism
For developers it is very easy to imagine the idea of ​​open individualism. On your computer there can be 1 processor (and 1 core), but at the same time there are a dozen windows open and each of them works independently, it seems that all of them work at the same time. Achieved by interrupt mechanism.
By analogy, each person is like an open program, and a kind of universal "I" is like a processor, which in turn "finds" itself in one of the programs.
This is consistent with hypotheses 6, 7, 8 (that is, open individualism is possible within the framework of these hypotheses, but does not necessarily follow from them).
Conclusion
The article is written in the style of brainstorming without the use of substances. Some ideas can be immediately rejected, but for completeness we will not do that. It will be interesting to hear alternative hypotheses that you know.
Source: https://habr.com/ru/post/375467/
All Articles