Comparing Artificial and Biological Intelligence: Insights and Challenges

Essay, Pages 12 (2975 words)

Views

Apart from the machine are those that relate to abstract conceptualization, which is best described by understanding a language. Dialect is known as a subset of a large collected group of behaviors. Numerous reports have shown that many living beings with higher cognitive ability, apart from human, think and are able of learning and problem-solve. The issue of consciousness is fundamentally related to the connection between the physical brain and mind. A scientific approach to understanding the nature of consciousness by reducing it to interrelating pieces reveals that both brain and mind are the same, and mind is simply the total activity of the brain, which perceived at a greater degree of interpretation.

Don't use plagiarized sources. Get your custom essay on

“ Comparing Artificial and Biological Intelligence: Insights and Challenges ”

Get custom paper

NEW! smart matching with writer

In contrast with this viewpoint, some believe the mind requires a physical structure in which to provide consciousness it goes beyond that structure.

The mind analyses the incoming brain signals to obtain information in different domains of senses and uses them to produce and store our memories. However, an act of cognition is a dynamic process where the desired sensory systems and their processing is organized based on the anticipation from a specific mental task, desire and purpose.

As a result, intelligence is a product of the cooperation of several dynamic mental elements. Psychology and neuroscience are very much influenced by artificial intelligence (AI) and computer science and cognitive activity are rational ways of computation. Nevertheless, so far neither classical computing nor quantum computing and the mechanistic model of computing has been able to gain all the power of biological intelligence and nature’s reckoning.

The explanation of the mind in scientific debates has not been able to keep its speed with the advancements in the fields of physical and computer sciences. Most discussions on consciousness and intelligence have been centered on the emergence of mind from convoluted and parallel computer-like brain processes. Although the quantum theory has provided the groundwork for computer and physical sciences for many years, it is just lately that comprehensive quantum-like processes in the brain have been considered.

We should consider that the underpinning elements on which the integrated circuitry of conventional computers are based at its greatest level largely explained by quantum mechanics. The new views and findings in advancement of the latest AI developments have been inspired because of various obstacles in this field. AI developments and by new study and findings. The methods used to explain the function of the brain have been shaped after the main scientific theories of our time. The conceptualization of the mind as a conscious machine is one of the results of the growth of mechanistic science.

Even though the neural network methodology was successful in modeling many innate brain functions, there are several types of cognition that seem to fall outside the reach of such models. In brief, the model of traditional neural network does not provide an answer to the question of the connection of thoughts’ patterns. The question of exactly how the activity of the neurons in the brain result in specific meanings or lead to specific images remained unanswered. Other questions are how standard computations inspire the description of the mind that can comprehend its surroundings, or how machines can simulate biological computing.

The conventional computation is governed by a binary logic that is distinct from the computations of quantum mechanics. In principle, as the application of quantum mechanics in computation is not limited by any size or scale, conventional computation seems to be inadequate. A traditional computer is not able to reorganize itself in response to inputs and even if they could, they will reach a state that associated with cease in reacting with their environment. In this state, computers will simply transmute data according to their designated program. This means, a conventional computer lacks the ability to choose its inputs. In fact, this is what biological systems are capable to do with no difficulty. Consideration of brain function based on quantum theory appears to be an answer to the question of the binding of thought’s patterns and non-local characteristics of cognition.

As quantum theory shows a non-local characteristics behavior, therefore, brain performance might have a quantum basis. The reorganization process has been recognized as a key process in the brain that allows the brain to define various perspectives. Now, the neuronal firing can be represented as processing within this reorganized reckoning hardware. Such a viewpoint can bring considerable implications in dual signaling systems, which ultimately requires a justification in terms of characterizing a binding field. They may not explain the fundamental of binding problem, but they help to make it easier to recognize the process of brain plasticity and adjustment.

The use of quantum mechanical methodology to the study of consciousness has a long history and the authors of quantum theory were amongst the first researchers to recommend its application in this type of study. More recently, many have proposed certain quantum theoretic models of brain activity, but there is no specific model that has been recognized as the ideal one at this moment. Some considered the connections between quantum theory and intelligence claiming that the brain's processing is coordinated in a hierarchy of language, which it is associative at the bottom side, self-organized in the middle, and has a quantum feature at the top. Neural learning model is associative, and it advances to generate the necessary structures to a higher level by reorganizing the greater levels of the neural formation. Representation of each cognitive agent in this network can be considered an abstract quantum system.

The relationships amongst the cognitive agents are structured by an applicable quantum field. This allows the individual at higher level of conceptual processes to begin cognition or action, leading to achieve an active behavior. One remarkable achievement of the quantum theory models is that they provide a solution to the determinism-free will problem. Based upon quantum theory, a system evolves causally until it can be observed. The act of observation produces a break down in the causal chain, which leads to the idea of a participatory universe. Consciousness offers a break in the rigorous law of causality. It might be reasonable to assume that this type of autonomy is associated with all forms of life. Nevertheless, its effect on the constant processes will depend on the entropy associated with the break in the causal chain.

Quantum theory attempts to describe knowledge in a relative sense [9]. In the world of quantum theory, it is pointless to consider a reality impartially. Information is a gathering of thoughts on the reductions of the wavefunction, produced by measurements using various types of instrumentations [10]. Both scientific and mathematical concepts of uncertainty reveal that quantum theory does not reside in the microworld on its own. Brain activities can not be described entirely by the neuronal firing as we need to recognize their higher-order bindings as swell, such includes thoughts and intellectual notions, as they have an impact on neuronal firings. Describing the brain function by the way of wavefunction has many advantages as it includes variables for the higher-order activities, such as considering abstract concepts.

AI is part of a scientific field that attempts to accomplish intellectual tasks that are generally difficult for advance living beings and computers. AI is becoming increasingly widespread, as many claims of its strong relationship with the function of neurons and biological intelligence. The similarities and differences between AI and artificial biological intelligence have been the focus of many debates. One of the rationales regarding the notion of using the simulated models and elements of nature in AI systems is to support in their advancement.

Today biological systems in the form of Artificial Neural Networks (ANNs) engage in various tasks, in which this may just be due to design decisions, rather than a fundamental resemblance of their underlying mechanisms. Failure of this approach to solve problems critical to intelligence provides a good distinction with Machine Learning (ML), an alternative to AI which is crucial to the development of artificially intelligent machines. To explore the relationship between artificial and biological intelligence we may consider the distinction between the choices of defining chess moves and moving the pieces. As the rules of a chess game are quite simple anyone with basic coding knowledge can write a chess program and by given adequate computational power it can defeat any chess champion by applying the rules.

This approach is known as symbolic AI as the computer reaches a solution by following predefined rules. Unfortunately, not all problems lend themselves to being solved in this way. At the same time, the hard-coded rules of symbolic AI do not bear any resemblance to the nature of biological intelligence. MI is a division of AI in which the machines understand how to solve problems by themselves without being given a set of specific rules on how to solve the task. This usually can be achieved by learning new rules through repeated attempts, and feedback on the amount of success in each attempt. This is the main fundamental resemblance between the current AI and biological Intelligence. We were not born either with the ability to recognize the objects in our surroundings nor were we given rules to identify those objects.

Instead, we see a significant number of objects that were occasionally identified and allow us to learn from the experience. Machine Learning works in quite a similar way, in which computers are shown millions of pictures featuring different objects until they gradually learn what those objects are. Even though this comparison may seem insignificant it is by far the most important assumption guiding the developments of current AIs. While the similarity between artificial and biological intelligence goes much deeper, important differences remain. The machinery that has allowed computers to learn these powerful rules, known as ANNs is directly inspired by neural networks in biological brains.

The rise of intelligence is fundamental in the development of all computational models. Exploring or generalization to solve specific problems that are shown to be a sign of AI has an all or none quality. The performance of living beings depends fundamentally on its normal behavior and biological intelligence that has demonstrated to have the form of gradation [13]. It may be claimed that all living beings are sufficiently intelligent as they survive in their natural environment. Nevertheless, even in mental tasks typically associated with human intelligence, animals may perform quite well. Higher animals, such as apes and dolphins, have quite astonishing intelligence as they can solve rational problems by using a certain type of generalization. These types of performance in theory can be used to define the brain gradation phenomena.

The features that distinguish the human brain apart from the machine are those that connect to intellectual conceptualization and best characterized by language comprehension. Although, no one denies that individuals that are deaf and unable to speak and don’t have a language, are able think. Language is best understood as a subset of a larger collection of behavior. Different studies reveal that higher living beings think and are capable of learning and problem-solving. In many living beings except human the capacity to think is based on discrimination of thought at different levels as they can not use abstract language.

It was not needed to believe that this would have developed similarly for all species if such conceptualization was a result of evolution. Other animals learn various concepts non-verbally, so it is hard for humans to determine their significance. Conceptualization is not a unique feature of human beings and neither having a human brain nor being able to use language is a precondition for cognition. Comprehensive knowledge of neural activity and function will reveal the magnificent abilities of brain’s function other than the one in human being. If the function of the brain is to think and to understand, it should be the job of behavioral neuroscience to provide the actual accounts of that thinking and learning in humans and other living beings.

Various important insights from study of animal intelligence shows the presence of different level of gradations in their cognitive function. It is obvious that many other groups of higher living beings are not as smart as humans; equally, certain living beings appear to be more intelligent than others. Studies on animal intelligence suggest that they present different styles of solving problem. Are the cognitive abilities of a certain group of animals is limited since their style of thinking has fundamental limitations?

The cognitive style of all living beings can be similar and the differences in their cognitive capabilities may come from the variations in the size of their brains. Since the present intelligent machines can not use internal representations, it is proper to assume that their performance can not match that of humans and other living beings. Recursive nature, or part-whole hierarchy of animal behavior is one of the most important perspectives [14]. Such recursive behavior has been viewed all the way up to the earth itself as a living entity. Incidentally, it may be asked whether the earth itself, as a living but unconscious mass, should be viewed such as the unconscious brain.

Memory and learning processes within the brain cortex may be organized adaptively. While there are several ways to achieve this, nesting process among neural networks within the brain cortex can be considered as a key principle in self-organization and adaptation [15]. Nested distribution of neuronal networks offers a means of orchestrating upward or downward regulation of complex neural processes operating both within and between various levels of brain structure. Two types of signaling can be important within a nested arrangement of distributed neuronal networks. One is a fast system that demonstrates a spatiotemporal arrangement of activation among modules of neurons.

These patterns produce neuronal firing and encode correlations that are the signs of the networks’ activity within the cortex. They are equivalent to the action of hormones and cellular chemical interactions that can be seen in certain living organisms. Other types of signaling in the brain have a slow mode that include such processes as protein phosphorylation and synaptic plasticity. The slow mode is closely linked to learning, development (i.e., ontogeny), experience and adaptation to the environment that affects both learning and memory. The ontogenetic view emphasizes chains of interactions that are mediated by the organism and has the potential to produce unexpected outcomes in population interactions and the community structure.

If evolution has led to the emergence of mind, machines with conscious intellect must also emerge. This paper attempted to look at artificial and biological intelligence from several viewpoints. Study of physical and biological aspects of cognition shows that machines are deficient compared to biological systems at integrating intelligence. Intelligent machine falls short on two main aspects as compared to a conscious brain. Primarily, distinct from the brain, intelligent machines are unbale to self-organize themselves in a recursive approach. Moreover, machines function mostly can be described based on classical logic, whereas Nature’s intelligence may very much associate with quantum theory.

It can be expected to have quite similar computational properties from both artificial and biological neurons if sophisticated ANNs can be advanced to a desirable level. Therefore, we should not be concerned regarding the impact of structural differences between biological and artificial neurons on the intelligent activity that arises from new technologies. What might be important is to know how the learning of the desired functions is accomplished. The lack of similarity between artificial and biological intelligence, except that both can understand many of the tasks, is their differences in performance. Criticism relating to lack of biological credibility of the ANN machinery seems to come from a lack of adequate information in neurobiology.

However, still, the difference between the two types of neurons remains very significant. In addition, we should consider that neurons are not the only cells capable of computation and communication. Such abilities expected to be spread beyond the nervous system, which is mostly undetermined. In addition, focus on the accuracy of the representation of biological neurons seems unwise as the properties of neurons have not been fully understood. Despite the profound differences in both structure and function between ANNs and biological intelligence it is necessary to remember that some of the fundamental principles of biological intelligence applied in the ANNs led to their astonishing success.

Since machines only follow commands, it is unrealistic to believe that on account of an increase in its connections and network, it can gain self-awareness. In the challenging quest to build a machine with the consciousness that is self-aware, we may ask why it is hard to make a brain-machine that has consciousness. The response is that the conscious brain is a self-organizing system that reacts to the environment and the quality of its communication with nature, whereas the computer is not able to do that. Biological communities that are the inhabitants of ecological systems show complex interrelationship amongst their components, they are self-organizing without being self-aware.

This suggests that while self-organization is one of the necessary pre-requisites for consciousness, by itself, it is not enough to produce a machine with awareness. On the other hand, based on classical computing principles, AI machines have a fixed universe of discourse, so they are unable to adapt in a flexible manner to a changing universe. Therefore, they lack the ability to match with biological intelligence.

Quantum theory has been able to provide an understanding of certain biological processes not amenable to classical explanation. If machines with consciousness can be created, they would be living machines, which would be an addition to variations of life forms as we know them. Nevertheless, the material world is not causally closed, and consciousness, mater and minds complement each other. Even at the level of the individual basis, current biomedical science that is deeply based on the advancement of machine paradigm is acknowledging the influence of mind on the body.