Table Of Content

Educational Communications and Technology: Issues and Innovations Mark V. Albert Lin Lin Michael J. Spector Lemoyne S. Dunn Editors Bridging Human Intelligence and Artificial Intelligence Educational Communications and Technology: Issues and Innovations Series Editors J. Michael Spector Department of Learning Technologies University of North Texas, Denton, TX, USA M. J. Bishop College of Education, Lehigh University University System of Maryland, Bethlehem, PA, USA Dirk Ifenthaler Learning, Design and Technology University of Mannheim, Mannheim, Baden-Württemberg, Germany Allan Yuen Faculty of Education, Runme Shaw Bldg, Rm 214 University of Hong Kong, Hong Kong, Hong Kong This book series, published collaboratively between the AECT (Association for Educational Communications and Technology) and Springer, represents the best and most cutting edge research in the field of educational communications and technology. The mission of the series is to document scholarship and best practices in the creation, use, and management of technologies for effective teaching and learning in a wide range of settings. The publication goal is the rapid dissemination of the latest and best research and development findings in the broad area of educational information science and technology. As such, the volumes will be representative of the latest research findings and developments in the field. Volumes will be published on a variety of topics, including: • Learning Analytics • Distance Education • Mobile Learning Technologies • Formative Feedback for Complex Learning • Personalized Learning and Instruction • Instructional Design • Virtual tutoring Additionally, the series will publish the bi-annual AECT symposium volumes, the Educational Media and Technology Yearbooks, and the extremely prestigious and well known, Handbook of Research on Educational Communications and Technology. Currently in its 4th volume, this large and well respected Handbook will serve as an anchor for the series and a completely updated version is anticipated to publish once every 5 years. The intended audience for Educational Communications and Technology: Issues and Innovations is researchers, graduate students and professional practitioners working in the general area of educational information science and technology; this includes but is not limited to academics in colleges of education and information studies, educational researchers, instructional designers, media specialists, teachers, technology coordinators and integrators, and training professionals. More information about this series at https://link.springer.com/bookseries/11824 Mark V. Albert • Lin Lin Michael J. Spector • Lemoyne S. Dunn Editors Bridging Human Intelligence and Artificial Intelligence Editors Mark V. Albert Lin Lin Computer Science and Engineering Department of Learning Technologies University of North Texas University of North Texas Denton, TX, USA Denton, TX, USA Michael J. Spector Lemoyne S. Dunn Department of Learning Technologies Department of Learning Technologies University of North Texas University of North Texas Denton, TX, USA Denton, TX, USA ISSN 2625-0004 ISSN 2625-0012 (electronic) Educational Communications and Technology: Issues and Innovations ISBN 978-3-030-84728-9 ISBN 978-3-030-84729-6 (eBook) https://doi.org/10.1007/978-3-030-84729-6 © Association for Educational Communications and Technology 2022 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Foreword Neuroscience and artificial intelligence (AI) have an exciting history with periods where AI and biological intelligence strongly inform one another, interrupted by periods where AI focuses on other issues and views neuroscience as a distraction. In the early 2000s, AI, or more specifically machine learning (ML), was furthest from neuroscience, focusing on statistical learning theory and Bayesian statistics, branches of AI that are powerful in their ability to describe everything but that are, in their generality, hard to link to neuroscience. That changed in 2012 when Alexnet,1 with inspiration from neuroscience, managed to beat the entire competition at recognizing objects in images. This, along with a rebranding as deep learning, has sparked a renewed interest in artificial neural networks, and currently is the domi- nating branch of ML. This is a timely book as we are in a phase where ML/AI and neuroscience are somewhat aligned, opening the doors to a fruitful exchange of ideas. Like many, if not most, neuroscientists, my personal journey into the intersection of biological and artificial intelligence started with an interest in how we think. I quickly got the feeling that we cannot understand one without the other. As such, my entire career happened between studies attempting to understand biological intelligence, with its rich inspiration from animals and brains, and studies attempting to understand artificial intelligence, with its rich framework of mathematics and computational intuition. I used experiments with humans to ask how people deal with uncertainty. I used implementations, for example, in brain machine interfaces, to ask how we can build well-working systems. I used theories to understand how AI systems work. But above all, I tried to keep alive the communication between these communities. I found the biological intelligence and the artificial intelligence communities speak fundamentally different languages, with much of what they can contribute to one another’s progress being lost in translation, highlighting the need for a book like this. 1 Krizhavsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems 25, 1097–1105. v vi Foreword One bridge between these communities is through mutual understanding how AI has been developing. And modern methods from ML have been moving very quickly. An accumulating literature helps us understand why artificial neural networks function so well. And the first part of the book nicely and digestibly gets across in broad strokes what we know about the developments in the ML field itself and how these developments relate to human learning. To me, this conceptual understanding of where ML is has always been a necessary factor to make ML use- ful to the human intelligence community. A second bridge is the use of AI to improve human intelligence. The traditional model of teaching suffers from many shortcomings. As a professor, I was able to observe this firsthand. Students have huge trouble following lectures and learning how to use the concepts at a much later point of time in homework. People are lonely, lacking ways to connect with one another. Tests are slow and time consum- ing. And the COVID pandemic brought on a host of new problems. AI now prom- ises to help with all this. We can use it to scale and democratize education. For example, I am involved with Neuromatch Academy, a nonprofit organization bring- ing AI-powered learning to thousands of students. The role of universities will, arguably, be to incorporate insights from AI to improve its many processes. And that will probably require most instructors to have a level of understanding of artificial as well as human intelligence. Models of the brain can also inform the AI community. Machine learning tech- niques are often used as model of the brain – people argue that because brains solve similar problems to ML systems that they must share properties. For example, many aspects of the visual system of mammals are well described by assuming similar representations between brains and artificial neural networks. In fact, I remember well my first exposure to this idea. I had long looked at the properties of neurons in the visual system. A really elegant paper by Bruno Olshausen showed that ML systems applied to understanding natural scenes can lead to system behaviors similar to neurons in the early visual system.2 Evolution matters for human intelligence, but the ML field is also performing a kind of evolution: we only ever get to see the algorithms that work best. As such, there may be a lot that people interested in human intelligence can learn from artificial intelligence and vice versa. This book is highlighting a range of these similarities. AI is starting to have a profound effect on us humans. AI systems are being built into just about all human endeavors, ranging from the mundane, such as recognizing spoken words, to the socially involved, such as deciding if a criminal is likely to re- offend and should thus remain incarcerated. So we really need to understand AI to build a better world. Where do our biases come from? Are the biases in the algorithm, the data, or their combination? How do AI systems relate to human creativity? When it comes to the application of AI, the sky is not the limit. As AI systems give us new powers, we need to make sure we use them responsibly. 2 Olshausen, B., Field, D. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature 381, 607–609 (1996). https://doi.org/10.1038/381607a0 Foreword vii Human and artificial intelligence have experienced crosspollination since AI systems existed. Both AI and the study of human intelligence have made massive progress over the last couple of decades, with many great advances made through shared concepts, tools, and especially people. However, there is still much work to be done, and this decade in particular is an exciting time of convergence between AI and human intelligence. This book builds that connection in a way that brings communities together for the benefit of all. Konrad Körding Penn Integrated Knowledge Professor University of Pennsylvania Philadelphia, PA, USA, CIFAR Member Toronto, Canada, Neuromatch Cofounder Philadelphia, PA, USA Preface Fig. 1 Poster of the “Human-Technology Frontier” credit to Dr. Xue Yang The concept of this edited book started with the two-and-a-half-day symposium sponsored by the Association for Educational Communications and Technology (AECT) and hosted by the Texas Center for Educational Technology at the University of North Texas in 2018. We are deeply grateful for the support of Dr. Phillip Harris, the executive director of AECT (https://www.aect.org/), and for the support of Dr. Kinshuk, dean of the College of Information at the University of North Texas (https://ci.unt.edu/). ix x Preface The theme of the symposium was “The Human-Technology Frontier: Understanding the Learning of Now to Prepare for the Future” (See Fig. 1 for the symposium poster). The symposium was successful, and was well attended by over 15 distinguished speakers and over 80 participants from multiple institutions, coun- tries, and professional contexts. The participants engaged in discussions on a range of topics such as human and holistic learning as informed by new research in neuroscience, creativity, critical thinking, self-directed learning, artificial intelligence, learning analytics, and measurements. The original idea was to have the symposium presenters contribute chapters to the book. Yet, the book has since evolved to include much younger authors thanks to Dr. Mark Albert, who joined the editors’ team in 2019. Dr. Albert introduced the book idea to the students at the Texas Academy of Math and Sciences (TAMS), who are in fact junior and senior high school students. As a result, we created a mentorship mechanism, where the TAMS high-school students would work with graduate students, who would be mentored by the faculty members to co-author the chapters. Consequently, most chapters of the book are co-authored by a high-school student, a graduate student, and a faculty member. We are very pleased with the outcomes of the book. The book followed the original vision, which would: (1) be multidisciplinary and transdisciplinary; (2) provide a forward-thinking perspective likely to lead to significant and sustained improve- ment in learning; and (3) embrace an integrative approach to designing and imple- menting advanced technologies in learning and instruction. In addition, the process created an apprenticeship and mentorship model which provided the opportunity for the younger students (high-school students) to learn the methods of conducting research, to engage in intellectual dialogues on the interactions between human intelligence and artificial intelligence, and for the younger students to establish scholarship and interdisciplinary inquiries. We would like to acknowledge the hard work by our chapter authors and we look forward to further conversations on this important topic with our readers. Denton, TX, USA Lin Lin Michael J. Spector Mark V. Albert Lemoyne Dunn