The problem is that brains are not just doing computation, they are also able to give interpretations and meaning to their own high-level information processing. One interesting case of computation is artificial neural networks, which could be interpreted as semi-deterministic information processing systems. Artificial neural networks evolve in a non-deterministic way thanks to self-learning and training from some given rules, which are not always explicitly programmed. These systems are semi-deterministic in the sense that it is not always possible to ensure what the net is learning, nor control the dynamic evolution of its learning process, even if deterministic learning rules have been given.
However, even if all these properties are controlled, it is never known what the network has learned until it is tested and even after testing; it is never possible to be sure about which node or layer encodes one or another statistical property of the data. Actually, it looks more like a domain-global and distributed characteristic than local Christian et al. Therefore, it is not possible to fully determine or predict classically speaking the way how the net will behave.
Neural and artificial neural nets are neither completely indeterminate nor determinate, but semi-determinate. Since artificial neural networks, as for example Hopfield networks Hopfield, , are inspired by biological principles Hebb, ; Gerstner et al. This will be discussed in section five. While some computer and cognitive scientists might not agree with this interpretation of information and computation, it is still admissible to have processing of information without computation and intelligence without a deterministic way of processing of information.
Actually, the brain apparently does it. In fact, the most important features of the brain are the result of unpredictable, nonlinear interactions among billions of cells Ronald and Nicolelis, ; Haladjian and Montemayor, At this point, the usual idea of digital computation in cognitive science and neuroscience should change in favor of a perspective of computation and information processing by analogy with physical systems where inputs, rules and outputs can be interpreted in a physical and global way.
One reason is that this analogy obscures the complex physical properties of the brain. On the one hand, neuroscience and cognitive science use indiscriminately concepts as information, computation and processing of information without understanding their physical counterpart, sometimes based on the assumption of non-hardware dependency of these concepts, other times because of the assumption that the brain encodes and decodes information and how it does so. The most common assumption is to think that activation or spikes in neurons are the only informative state.
While other cells, for example astrocytes Alvarez-maubecin et al.
- Human Brain: Facts, Functions & Anatomy.
- Better Than Conscious?: Decision Making, the Human Mind, and Implications For Institutions.
- Statistical Mechanics for Engineers?
- Brain Facts That Will Blow Your Mind | The Healthy?
- BBC News Navigation;
- This is How Artificial Intelligence Will Influence the Human Mind - OpenMind.
- New evidence suggests that fire may have influenced the evolution of the human mind.
In addition, inactivation and deactivation states could also carry valuable information about dynamical brain states at macro and micro scale. Neurons are never in a static state and their membranes are presenting fluctuations that could still be informative for instance, Sub-threshold oscillations. The distinctive physical brain properties and their dynamical interactions are apparently more important than in digital interpretations, what implies that hardware cannot be ignored at all. According to this point, the analogy between a drum and the brain would be more relevant than the analogy brain-computer.
Drums can respond with different and complex vibration states when they are stimulated, and they can be also understood on computational terms: input hits , rules physical laws, physical constraints such as material, tension, etc. Indeed, the brain has many more similarities with a dynamical system as a drum than with digital computers, which are based on discrete states.
On another hand, computer science is missing valuable information on the attempt of replicating brain capabilities. Sub-emergent properties in the brain may be also important, such as plasticity changes due to the intentional practice of meditation Lutz et al. These characteristics should be understood and incorporated in order to implement the social behavior in new generations of computers, machines and robots. Considering that some of these behaviors are intrinsic to biological organisms, perhaps these behaviors are not reproducible without some intrinsic constituents of information processing of biological organisms Chappell and Sloman, ; Sloman, as for example oscillations or neurotransmitters.
Finally, abstractions and general concepts are really useful in theoretical terms; however, concepts as computation, information, and information processing in the brain do not have evident interpretation. Realizing that these concepts should not be used as an analogy with computers is the only way to lead us to the correct direction: Focusing on differences between brains and computers, and trying to fill the gaps without assumptions. When differences between concepts appear, it becomes necessary to clarify some of them.
That is why a subset of the features of human beings has been identified and some concepts clarified. For example, a better understanding, and definition of information processing in the context of human intelligence, where computation will be a kind of information processing among many other types, including the characteristic one to biological organisms Chappell and Sloman, Probably, new concepts and foundations of information will be also needed, especially to understand the real language of brain cells, as a crucial theoretical starting point.
These foundations should be inherent to minimal constitutive parts of physical theories and as it mentioned above, important hardware requirements, emergent, plasticity and sub-emergent properties should be considered in any attempt to replicate brains features. Thus, a computer-brain metaphor is not useful anymore, at least in the current sense. Nevertheless, it could still be possible to replicate some brains abilities thanks to new formulations of information processing and theoretical frameworks. Intelligence should also be considered as a whole. Intelligence is often understood as the ability to solve problems in an efficient way, thanks to other mechanisms like learning and memory.
It means the maximization of the positive results in a certain solution while minimizing the negative impacts, for instance, waste of time. To do that, other processes, such as learning and memory, are also needed and associated with the definition of intelligence.
In a general sense, learning has been understood as the process to gain new knowledge or improve some behavior, while the memory is the storage of this knowledge. To solve problems efficiently, it is necessary to access a certain memory that was acquired thanks to a specific learning that will modify again the memory of the system.
Darwin’s Unfinished Symphony: How Culture Made the Human Mind – The Evolution Institute
The more intelligent is the system, the more it learns. However, in that framework, it is forgotten that emotions, subjective experiences, and cognition are deeply connected with human intelligence Haladjian and Montemayor, They play a crucial role in learning, in the consolidation of memories, in retrieved memory and human cognition in general Cleeremans, Therefore, as it was stated in section A Sub Set of Human Capabilities, intelligence is better defined as the capability of any system to take advantage of their environment to achieve a goal.
Specifically, human intelligence would be the ability to take advantage of their environment to keep autonomy and reproduction thanks to a balance between rational and emotional information processing. With this last definition, both main features on human thinking, reason and emotion, are merged in one global concept, together with two other features, autonomy and reproduction, that also define, altogether, the potential set of human being properties.
In this context, perception, cognition, learning, and memory are key features of human intelligence considered as a whole and emerged from specific soft properties of brains, such as for example neural plasticity and oscillations. Learning and memory are intrinsically dynamic processes in the brain, changing all the time and conditional to these soft neural properties, while for computers, memory is a very static feature, mainly grounded on symbolic discretization, and in the best case, learning is driven for efficient algorithms which are also statics.
Biologically, the more intelligent the system, the more balance the system has between different inner processes to achieve specific or general goals. For example, a computer is designed to make faster calculus, algorithms, and other kinds of very useful tasks, however, the computer cannot take advantage of anything that it does, in conclusion, computers are not really intelligent. Nevertheless, the last version of AlphaGo zero Silver et al.
Using the intelligence definition stated here, this system is more intelligent than a simple computer.
By analogy, if a lizard is compared with a mouse, the later has a larger repertoire of actions, taking more advantage of their environment, than the lizard. In this sense, mice are more intelligent than lizards. For instance, a person who wins a discussion with his partner at the expense of their relationship is less intelligent than who wins the discussion and keep a good relationship. The crucial point is that emotions are playing an important role in classical processes of natural intelligence such as learning and memory, but they are also playing a crucial role increasing the repertoire of actions and possibilities to achieve biological goals.
Emotions are not just used to improve memory or learning curves; they are also useful to increase the variability and unpredictability of behavior. On the one hand, high level processes needed for moral thinking such as self-reflection, sense of confidence, error detection, understanding context, among others Figure 1B are essential part of consciousness and subjective experience as a whole Gehring et al.
The Coolest Thing in the Universe
Self- reflection and sense of confidence are understood as the ability to report a mistake, like error detection, and grade the confidence of some decisions or action, even before receiving any feedback about the mistake. In fact, some researchers have suggested the intrinsic relation between social complexity associated with these processes and the emergence of consciousness Arsiwalla et al.
On another hand, humans first need to be conscious to take some complex rational decisions, to plan, and to have the intention to do something Baars, ; Tononi and Koch, For example, vegetative patients and minimally conscious patients do not present signals neither planning nor having intentions to do minimal tasks Gosseries et al.
Planning and intentions apparently emerge when minimal signs of consciousness exceed a threshold. In fact, these minimal signs can be interpreted as predictors of recovering in minimally conscious patients Bekinschtein et al. Other works are re-defining the idea of subjective experience until its minimal constitutive part and argue the existence of basic subjective experience even in insects Barron and Klein, It would mean that complex decisions, planning, and have intentions which are needed to moral thoughts are different from consciousness, although they are closely related: Subjective and conscious perceptions are apparently previous to rational intelligence, planning, moral thoughts, and even efficient behaviors.
For example, experiments in the psychology of judgment and behavioral economics have also shown that subjects tend to perform some tasks in a biased manner even if they have been trained, suggesting that logical and rational intelligence appear only after more elaborated information processing Gilovich et al. It is clear that how biology implements high-level intelligence is completely different from how computer science implements it Moravec, The whole set of human intelligence, as the capacity to take advantage of the environment, would only emerge after awareness.
The need to incorporate subjective experience and eventually consciousness to reach complex intelligence implies a complex problem which involves many different processes as awareness, emotions, subjectivity, intentionality, and attention, among others.
Consciousness should be composed by all of these processes like a differentiated and unified whole, but it is not any of them. For example, it could be necessary to be aware to have emotions and subjective experiences, or maybe vice versa, and we will need them to show intentionality, attention and high-level cognitive abilities. It is also necessary to insist and distinguish that these are different processes, for instance, awareness and attention; while it is important understanding all of them as constituent parts of what we describe as consciousness.
Facts about the human brain
For example, at least two main processes have been identified with consciousness: 1 the fact of knowing something or what here will be understood as awareness, i. Here, the notion refers to the idea of self-reference for living machines. Nevertheless, consciousness is not reduced to the possible relationship between awareness and self-reference, it is the whole process of processes interconnected with awareness, self-reference, subjectivity, rational and emotional thoughts, among many others.
Consciousness emerges from all of them as a whole Varela and Goguen, Hence, after consciousness emerges from the interaction between these processes, human intelligence would appear as the group of strategies to take advantage of the environment thanks to the balance of emotional and rational information processing. Four types of cognition and some of their associated tasks can also be defined from awareness and self-reference Shea and Frith, ; Signorelli, ; Figure 2A : 1 Type 0 Cognition corresponds to systems which have neither awareness of their internal or external contents nor self-reference of their internal processes.
One example in humans is motor control.
Motor control is the automatic control that the neural central system has to move some joints and muscles without any necessity of voluntary control or awareness. Many apparently high-level tasks in human can be classified in this category, as for example the extraction of individual word meaning and primary attention sometimes called priming. In other words, it is aware of the elements that the system needs to manipulate and solve particular or general problems, but the system does not monitor this manipulation.http://expo24.online/img/51.php
Explaining consciousness and the other biggest mysteries of your brain
It can be also associated with a holistic kind of information. For example, when subjects answer very quickly to some apparently intuitive questions but their answers are normally wrong Fallacy questions. Type 1 cognition also involves mental imagery, emotions, voluntary attention and most of our subjective capabilities as to be aware of the experience of color or pain, among others. This type of cognition involves the high-level cognitive capabilities defined above and needed for human morality. Some tasks, which are part of this type of cognition, can be: the ability of self-reflection; rational thinking; detection of error even before receive any clue about the mistake; sense of confidence, before and after any decision; complex meanings; voluntary and quick learning, among other interesting features of human thinking.
In other words, the system has self-reference, but it cannot extract meaning either from their manipulation nor their contents. It could be like an automaton, and actually, there is not a biological example of this category.
Related Making the Human Mind
Copyright 2019 - All Right Reserved