Artificial Intelligence

The essence of Artificial Intelligence (AI), with its rapid growth and unfolding realms to reach its prime, lies in the production of new kinds of intelligent machines that are capable of responding similarly to human intelligence. Computational technologies with machine learning as the core component of AI, with extensive utilization and transformative impacts, ensure the training of complex data to automate or augment the skills of human beings. Major advances in research enable the harnessing of volumes amounts of data, software utilized for algorithms and predictive models, computing with high-performance as well as the workforce involved, all indicating the crosscutting nature of AI providing significant motivational power for research formulization in a prioritized and organized way. To put it differently, Artificial Intelligence way of thinking, in a way, along with its processes and inherent complexity fosters the profound examination of matters, phenomena and problems through the lenses of more nuanced and versatile understanding [Karaca, Y. (2022)].

The succinctly provided functions of AI can be considered as learning (data learning, reinforcement learning, deep learning, explainable machine learning, and so forth). The reasoning aspect involves knowledge representation, problem-solving and reasoning, fuzzy logic, expert systems, automated reasoning, among many others. The AI models include decision trees, support vector machines, kernel machines, fuzzy models, Artificial Neural Networks (ANNs), Convolutional Neural Networks (CNNs), modeling, classification, and so on. While robotics encompass intelligent robots and human-machine interfaces along with biomimetics smart controls and mechatronics, vision refers to patterns’ recognition, machine perception and automated surveillance besides the other related affordances. Hardware, quantum computing, language, planning with multi-agent systems and applications make up the other important facets of AI.

The concept of inanimate objects’ coming to life and existence as intelligent entities has been a scheme since antiquity, dating back to the Greeks that developed a number of narratives on robots, as well as to Egyptian and Chinese engineers who were able to build automatons.. In other words, during the ancient time of Greeks, classical philosophers attempted to describe human thinking as a system of symbolic conceptualization The time of antiquity has ingrained notion of robots.

At the age of reason, René Descartes was reputedly keen on automata, inspired from the idea that living things were like biological machines functioning in clockwork manner, and thus, mechanical humanoids were imagined during the Renaissance, evoking wonder in clockwork automata expanding in Europe with hydraulic and spring-powered qualities. Besides Descartes, Leibniz also had the idea of machines that were able to carry out tasks requiring intelligence [Fjelland, R. (2020)], [Karaca, Y. (2022)]. These sorts of combinations also inspired Leonardo da Vinci who had been designing plans for a robotic knight with internal operating system through some pulleys and weights. The modern project of creating human-like AI emerged after the World War II as it was discovered that electronic computers did not only serve for number-crunching but they also managed to manipulate the symbols. The first incarnation of AI is contended to have been invented by the British mathematician and computer scientist Alan Turing who is known to be the father of AI. Turing employed the automated electromagnetic machine during the Second World War to break the enigmatic codes.

As one of the prominent names in the field, Marvin Minsky defined AI as the science of making machines do things that necessitate intelligence if conducted by people. Referring to weak AI, this definition also unfolds the historical path where researchers pursued the developing of AI which would be in principle being identical to human intelligence, which is referred to as as strong AI [Fjelland, R. (2020)]. The conducts of outperforming or imitating human cognition became prominent in 1950 when Alan Turing proposed an imitation game (Turing Test) to assess if a computer could fool humans into thinking during communication. 10 years before, in the 1940s, mathematician Turing's work had ended up with the creation of first electromechanical computer named the Bombe. Following Turing’s trial, Princeton University researchers built MADALINE, the first AI neural network implemented for a real-world problem, with a system modeled on the nervous system and brain. As for the first use of the term, John McCarthy used AI as a term for the first time at Dartmouth Conference in 1956. John McCarthy put forth the definition of AI as ‘the science and engineering of making intelligent machines’. Following that date, AI has been through three booms, namely springs:

The first boom (1956-1976) development of neural network software and chat software, proving some mathematical theorems announcing that AI era was about to come, the idea was put forward that robots would surpass humans in a decade or so [Nilsson, N.J. (2009)], [Y. Dong, J. et al. (2020)] 1960s and 1970s are characterized by the development of knowledge-based systems [Buchanan, B.G. (2005)].
The second boom (1976-2006) shows the development of Hopfield neural network and BT training algorithm, which developed the advent of speech translation plan and speech recognition.
The third boom (2006 and onwards) with the AI outbreak that manifested itself with Hinton putting forth the deep learning technology in 2006, and ImageNet Competition made many breakthroughs in image recognition visible in 2012 [Y. Dong, J. et al. (2020)], [Karaca, Y. (2022)].

Coming to our day, the huge investments and interest in AI boomed 2020s with machine learning techniques being successfully implemented to diverse problems including technology, industry, medicine, science, media, educational systems and academia owing to the development of methods reinforced by the application of powerful computer hardware as well as the collection of big datasets.

Regarding the types of AI, there is the categorization of Artificial Narrow Intelligence, Artificial General Intelligence and Artificial Super Intelligence. While artificial narrow intelligence is quite common currently, artificial super intelligence pertains exclusively to a distant future. More specifically, the lowest and most common level of AI refers to ‘artificial narrow intelligence’. This AI uses machine learning algorithms which help it for completing a single task without the need for human beings to intervene (like daily virtual assistants of Apple’s Siri and Amazon’s Alexa, self-driving cars using this technology). Secondly, artificial general intelligence, also referred to as artificial ‘strong’ intelligence is intended to carry out tasks which a human is capable of (i.e. finding solutions to unfamiliar tasks). Despite some ongoing debates on this matter, it is known that no AI exists as of yet that could match up to the vast and complex capabilities of the human mind. Finally, artificial super intelligence is a type of AI that would supersede human intelligence, which is currently only hypothetical, being around in science fiction only for now. Whilst this is currently way beyond the realms of possibility, it is not inconceivable that one day, considering the rapid advancement in AI learning, a super-intelligent system should come into being. The abilities that could define super AI are distinctly human like developing emotions, desires, and / or beliefs.

To recap, the course Artificial intelligence (AI) followed is opposite of the mechanical crafts which in their early times lost themselves with the properties of specific domains. The advent of AI reversed the direction, showing the promise of general mechanisms, in heuristic search, toward deductive reasoning concerning machine learning as well as its counterpart as if these universal concepts would attain every kind of problem-solving task with no need for building the meticulously articulated representations of knowledge [Michie, D. (1979)]. It is possible to extend the definition through the development of computer systems capable of performing tasks that require human intelligence including decision-making, object detection, solving complex problems, to name many more in due course of time.

For the evolutionary pattern of AI, it can be noted that non-biological intelligence has been around for some time, anteceding the nominal establishment of AI. Much of the history is characterized by reasonable considerations for recast the goal of creating intelligence as with designing intelligence. The evolution of AI is marked by the initial understanding of the principles, subsequently using of human intelligence for devising a design resting on the principles which leads to the ultimate building a system in accordance with the design [Spector, L. (2006)]. The problem-solving performance of evolutionary algorithms, on the other hand, has shown significant advances with the human-competitive results achieving different designs in science and engineering, by integrating the achieved results through the employment of systems integrating the insights of biological progresses most of which encompass genetic representation of developmental processes.

As for the goals pertaining to AI, simulating or creating intelligence is split into subproblems constituting particular capabilities researchers expect that intelligent systems manifest within the scope of AI research. Computationalism and functionalism come to the foreground concerning AI, with computationalism being at the position in the philosophy of human mind as an information processing system, whereas thinking is regarded as a form of computing. The relationship between mind and body, being similar or identical to the relationship between software and hardware, underlies the argument of computationalism, which is a philosophical standing inspired by the work of AI researchers and cognitive scientists back in the 1960s [Nilsson, N. (1998)]. In the current landscape, there are discussions revolving around superintelligence (hypothetical agent having the intelligence of surpassing the most gifted human mind’s capacity) and singularity (the improved software’s being better at improving itself and leading to intelligence explosion). These capabilities and notions lead to another angle of discussion, which is transhumanism, namely the position that human beings having the allowance to use technology and enhancing human cognition as well as bodily function by modifying and expanding the capabilities transcending the current biological constraints. Please see...

Themes + News