sign in sign up
<<
Home , Computational science

Journal of Information and Computational ISSN: 1548-7741

A REVIEW STUDY ON FUTURE OF

Amjed Khan Bhatti1, Irfan Jalal Bhat2 1Assistant Professor, Department Of Science and IT, Govt. P.G. College Rajouri. 2Research Scholar, Department Of Computer Application B.U Ajmer Rajasthan India.

ABSTRACT Artificial intelligence (AI), , and neural networks represent incredibly exciting and powerful machine learning-based techniques used to solve many real- world problems. From the early days of , games have been important testbeds for studying how well machines can do sophisticated decision making. In recent years, machine learning has made dramatic advances with artificial agents reaching superhuman performance in challenge domains like Go, Atari, and some variants of poker. As with their predecessors of chess, checkers, and backgammon, these game domains have driven research by providing sophisticated yet well-defined challenges for artificial intelligence practitioners. AI technology has long history which is actively and constantly changing and growing. It focuses on intelligent agents, which contains devices that perceives environment and based on which takes actions in order to maximize goal success chances. In this paper, we will explain the current AI basics and various representative applications of AI. In context of current digitalized world, Artificial Intelligence (AI) is the property of machines, computer programs and to perform the intellectual and creative functions of a person, independently find ways to solve problems, be able to draw conclusions and make decisions. Most artificial intelligence systems have the capability to learn, which allows people to improve their performance over time. The recent research on AI tools, including machine learning, deep learning and predictive analysis intended toward increasing the planning, learning, reasoning, thinking and action taking capability it will explore the future predictions for artificial intelligence and based on which potential solution will be recommended to solve it within next decades. In the present white paper we discuss the current state of Artificial Intelligence (AI) research and its future opportunities. We argue that solving the problem of invariant representations is the key to overcoming the limitations inherent in today’s neural networks and to making progress towards Strong AI. Based on this premise, we describe a research strategy towards the next generation of machine learning beyond the currently dominant deep learning paradigm.

Keywords: Artificial intelligence (AI), machine learning, Artificial neural networks (ANNs), Deep learning.

Volume 10 Issue 7 - 2020 171 www.joics.org Journal of Information and Computational Science ISSN: 1548-7741

INTRODUCTION Throughout human societies, people engage in a wide range of activities with a diversity of other people. These multi-agent interactions are to everything from mundane daily tasks, like commuting to work, to operating the organizations that underpin current life, such as governments and economic markets. With such complex multi-agent inter-actions playing a pivotal role in human lives, it is desirable for artificially intelligent agents to also be capable of cooperating effectively with other agents, particularly humans. Artificial intelligence (AI) is an area of that emphasizes the creation of intelligent machines that work and reacts like humans. Some of the activities with artificial intelligence are designed for include: , Learning, Planning, Problem solving. In this topic we shall discus the following subjects; deep learning, Machine learning, Computer Programming, Medical field. Deep Learning has enabled many practical applications of Machine Learning and by extension the overall field of AI. Deep Learning breaks down tasks in ways that makes all kinds of machine assists seem possible, even likely. Driverless cars, better preventive healthcare, even better movie recommendations, are all here today or on the horizon. AI is the present and the future. With Deep Learning’s help, AI may even get to that science fiction state we’ve so long imagined. Machine learning at its most basic is the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world. So rather than hand-coding software routines with a specific set of instructions to accomplish a particular task, the machine is “trained” using large amounts of data and algorithms that give it the ability to learn how to perform the task. The term intelligence refers to the ability to acquire and apply different skills and knowledge to solve a given problem. In addition, intelligence is also concerned with the use of general mental capability to solve, reason, and learning various situations Intelligence is integrated with various cognitive functions such as; language, attention, planning, memory, perception. The evolution of intelligence can basically is studied about in the last ten years. Intelligence involves both Human and Artificial Intelligence. In this case, critical human intelligence is concerned with solving problems, reasoning and learning. Furthermore, humans have simple complex behaviors which they can easily learn in their entire life.

BRIEFLY ELUCIDATE METHODOLOGICAL BACKGROUND OF AI AS WELL Artificial Intelligence has facilitated us in almost every field of life and has immense scope in future for more productivity and betterment. The origin of artificial intelligence goes back to the advances made by Alan Turing during World War II in the decoding of messages. The term as

Volume 10 Issue 7 - 2020 172 www.joics.org Journal of Information and Computational Science ISSN: 1548-7741

Such was first used in 1950, but it was only in the 1980s when research began to grow with the resolution of algebra equations and analysis of texts in different languages. The definitive takeoff of Artificial intelligence has come in the last decade with the growth of the internet and the power of microprocessors. ”Artificial intelligence may be the most disturbing technology the world has ever seen since the industrial revolution” Paul Daugherty, Accenture’s chief technology officer, recently wrote in an article published by the World Economic Forum. This field is now booming due to the increase in ubiquitous computing, lowcost cloud services, new algorithms and other innovations, adds Daugherty. Developments in Artificial Intelligence go hand in hand with the development of processors that over time have made them start to see these technologies as intellectual, even changing our idea of intellect and forthcoming the perceptions of ’machine’, traditionally unintelligent capacity previously assigned exclusively to man. The AI was introduced to the scientific community in 1950 by the English Alan Turing in his article “Computational Machinery and Intelligence.” Although research on the design and capabilities of computers began some time ago, it was not until Turing’s article appeared that the idea of an intelligent machine captured the attention of scientists. The work of Turing, who died prematurely, was continued in the United States by John Von Neumann during the 1950s. His central contribution was the idea that computers should be designed using the human brain as a model. Von Neumann was the first to anthropomorphize the language and conception of computing when speaking of memory, sensors etc. of computers. He built a series of machines using what in the early fifties was known about the human brain, and designed the first programs stored in the memory of a computer. McCulloch (1950) formulate a radically different position by arguing that the laws governing thought must be sought between the rules that govern information and not between those that govern matter. This idea opened great possibilities for AI. In addition, Minsky (1959) modified his position and argued that imitation of the brain at the cellular level should be abandoned. The basic presuppositions of the theoretical core of the AI were emphasis on recognition of thought that can occur outside the brain. On 1958, Shaw and Simon design the first intelligent program based on their information processing model. This Model of Newell, Shaw and Simon was soon to become the dominant in cognitive psychology. At the end of the 19th century, sufficiently powerful formal logics were Obtained and by the middle of the 20th century, machines capable of make use of such logics and solution algorithms.

THE FUTURE OF ARTIFICIAL INTELLIGENCE Intelligence, while a broad and comprehensive concept, is also a notoriously elusive one. In their comprehensive survey of available definitions of intelligence, Legg and Hutter list and review more than 70 different notions. Extracting the most common features, they define intelligence as follows: Intelligence measures an agent’s capacity to achieve goals in a wide range of environments. The difficulty in grasping what intelligence actually is directly carries over to the attempts of emulating it in machines. The term Artificial Intelligence has been around for many decades and, depending on technological progress at the respective time, it has carried quite

Volume 10 Issue 7 - 2020 173 www.joics.org Journal of Information and Computational Science ISSN: 1548-7741

different connotations. The fact that the marketing departments of large software companies have captured the term AI for their own purposes has not helped to improve clarity either. If we apply Legg and Hutter’s definition strictly, then no Artificial Intelligence exists today. No existing today is able to achieve goals in a range of environments comparable to the one in which humans operate. On the other hand, algorithms can perform certain well defined cognitive tasks - like playing a computer game or recognizing the race of a dog in an image - on (super-) human level. Hence, the AI community distinguishes between Narrow AI and Strong AI. Similar to the definition above, we can define Narrow AI as an agent’s capacity to achieve goals in a (very) limited range of environments. Often it is assumed that the performance of this Narrow AI will be as good as or better than the human performance. From self-driving cars to AlphaGo’s defeat of Go champion Lee Sedol, all the current AI breakthroughs are examples of Narrow AI. A Strong AI – also called General AI or Artifical General Intelligence (AGI) – would be a which fits Legg and Hutter’s definition above, i.e., it would achieve goals in a wide range of environments. Again, it is often assumed that in doing so this agent would be at least as successful as an average human being. In particular, such a Strong AI would pass the famous Turing test by tricking a human conversation partner in an online chat (and taking into account Recent advances in voice synthesis, certainly also on the phone) into believing that it is actually another human rather than a machine.

CONCLUSION In this way, artificial intelligence can achieve great discoveries and advances for humanity due to its multiple possibilities. Most artificial intelligence systems have the capacity to learn, which allows people to improve their performance over time. The adoption of AI outside the technology sector is at an early or experimental stage. The evidence suggests that AI can provide real value to our lives.AI bases its operation on accessing huge amounts of information, processing it, analyzing it and, according to its operation algorithms, executing tasks to solve certain problems. Due to the new computing architectures of the cloud, this technology becomes more affordable for any organization. AI is an extremely powerful and exciting field. It’s only going to become more important and ubiquitous moving forward, and will certainly continue to have very significant impacts on current society. Artificial neural networks (ANNs) and the more complex deep learning technique are some of the most capable AI tools for solving very complex problems, and will continue to be developed and leveraged in the future. While a terminator-like scenario is unlikely any time soon, the progression of artificial intelligence techniques and applications will certainly be very exciting to watch! We aim to reflect the interdisciplinary nature of our work in the background of our research group’s permanent staff - covering expertise in neuroscience, computer science, mathematics, and theoretical . In addition, we are interested in building up collaborations with academic research groups from the mentioned fields. Given our ambition to rethink the

Volume 10 Issue 7 - 2020 174 www.joics.org Journal of Information and Computational Science ISSN: 1548-7741

basics of AI, we also feel that important contributions might come from brilliant junior researchers with limited experience (and bias). We are therefore actively engaging with students and young researchers via internships, university competitions, and similar formats, trying to generate as many fresh ideas as possible for this fascinating and important field of science.

References [1]. N. Bostrom: “Superintelligence - Paths, Dangers, Strategies” (Oxford University Press, 2014) [2]. E. Brynjolfsson, A. McAfee: “The second machine age” (Norton paperback, 2016) [3]. P. Dayan, L.F.Abbott: “Theoretical Neuroscience” (Massachusetts Institute of Technology, 2001) [4]. X.Wang, D. Fouhey, and A. Gupta, “Designing deep networks for surface normal estimation,” in Proceedings of the IEEE Conference on and , 2015, pp. 539–547. [5]. H. Cuay´ahuitl, S. Keizer, and O. Lemon, “Strategic dialogue management via deep ,” arXiv preprint arXiv: 1511.08099, 2015. [6]. B. M. Lake, T. D. Ullman, J. B. Tenenbaum, and S. J. Gershman, “Building machines that learn and think like people,” Behavioral and Brain , vol. 40, 2017. [7]. E. Ohn-Bar and M. M. Trivedi, “Looking at humans in the age of self-driving and highly automated vehicles,” IEEE Transactions on Intelligent Vehicles, vol. 1, no. 1, pp. 90– 104, 2016. [8]. J. Wei, H. Liu, G. Yan, and F. Sun, “Robotic grasping recognition using multi-modal deep extreme learning machine,” Multidimensional Systems and Signal Processing, vol. 28, no. 3, pp. 817–833, 2017. [9]. M. Mathieu, . Couprie, and Y. LeCun, “Deep multi-scale video prediction beyond mean square error,” arXiv preprint arXiv: 1511.05440, 2015. [10]. G. Chen, D. Clarke, M. Giuliani, A. Gaschler, and A. Knoll, “Combining and discrimination for 3d action recognition,” Signal Processing, vol. 110, pp. 67–81, 2015. [11]. J. Wulff and M. J. Black, “Efficient sparse-to-dense optical flow estimation using a learned basis and layers,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 120–130. [12]. J.-. Ruiz-Sarmiento, C. Galindo, and J. Gonzalez-Jimenez, “Scene object recognition for mobile robots through semantic knowledge and probabilistic graphical models,” Expert Systems with Applications, vol. 42, no. 22, pp. 8805–8816, 2015. [13]. L. S. Gottfredson, “The general intelligence factor,” 1998. [14]. Y. Zhang, D. K. Robinson, A. L. Porter, D. Zhu, G. Zhang, and J. Lu, “Technology roadmapping for competitive technical intelligence,” Technological Forecasting and Social Change, vol. 110, pp. 175–186, 2016.

Volume 10 Issue 7 - 2020 175 www.joics.org Journal of Information and Computational Science ISSN: 1548-7741

[15]. M. Turan, Y. Almalioglu, H. Araujo, E. Konukoglu, and M. Sitti, “A non-rigid map fusion-based direct slam method for endoscopic capsule robots,” International journal of intelligent robotics and applications, vol. 1, no. 4, pp. 399–409, 2017. [16]. D. C. Parkes and M. P. Wellman, “Economic reasoning and artificial intelligence,” Science, vol. 349, no. 6245, pp. 267–272, 2015. [17]. T. Y. Wong and N. M. Bressler, “Artificial intelligence with deep learning technology looks into diabetic retinopathy screening,” Jama, vol. 316, no. 22, pp. 2366–2367, 2016. [18]. M. Turan, Y. Almalioglu, E. Konukoglu, and M. Sitti, “A deep learning based 6 degree- of-freedom localization method for endoscopic capsule robots,” arXiv preprint arXiv: 1705.05435, 2017. [19]. M. Turan, Y. Almalioglu, H. Gilbert, A. E. Sari, U. Soylu, and M. Sitti, “Endo-vmfusenet: deep visual-magnetic sensor fusion approach for uncalibrated, unsynchronized and asymmetric endoscopic capsule robot localization data,” arXiv preprint arXiv:1709.06041, 2017. [20]. M. Turan, J. Shabbir, H. Araujo, E. Konukoglu, and M. Sitti, “A deep learning based fusion of rgb camera information and magnetic localization information for endoscopic capsule robots,” International journal of intelligent robotics and applications, vol. 1, no. 4, pp. 442–450, 2017. [21]. D. M. Mart´ınez and J. C. Fern´andez-Rodr´ıguez, “Artificial intelligence applied to project success: a literature review,” IJIMAI, vol. 3, no. 5, pp. 77–84, 2015. [22]. Devlin, Jacob, et al. “Bert: Pre-training of deep bidirectional transformers for language understanding.” arXiv preprint arXiv: 1810.04805 (2018). [23]. T. Poggio, F. Anselmi: “Visual Cortex and Deep Networks: Learning Invariant Representations” (MIT Press,2016) [24]. O. Sacks: “The Man Who Mistook His Wife for a Hat” (Gerald Duckworth & Co Ltd.,1985) [25]. D. Silver et al.: “Mastering the game of Go with deep neural networks and treesearch”(Nature 529, p. 484–489,2016)

Volume 10 Issue 7 - 2020 176 www.joics.org