Handbook of Human-AI Collaboration
Description
This Open Access book presents the historical evolution of artificial neural networks and the principles that underpin deep learning. It introduces the main concepts of Foundation Models employed in Large Language Models (LLMs) and more generally in Large Whatever Models (LWMs). It addresses the crucial need for explainability in both language and hybrid models, projecting future directions in the field.
The work extends beyond technical dimensions to explore the intricate dynamics of Human-AI Collaboration, from the foundations of human-centered AI methodologies to generalized AI-human intelligence. The book explores challenges of multimodal foundation models in particular when it comes to multimodal perception, generation and embodiment.
Contributors delve into topics such as complex reasoning, planning, argumentation, and applications in education and personal growth. Human-Large Whatever Models Interaction is examined in the context of co-adaptation, co-evolution, and the reciprocal influence between AI and human cognition, emotions, and behaviours. Benchmarking criteria and datasets for evaluation are discussed, providing insights into the evolving landscape of human-AI interaction. The societal impact of foundation models is explored in-depth, considering the dynamics of AI-driven techno-social systems, role distribution in AI-human collaborations, and the long-term implications on society. Ethical and legal aspects encompass conceptual backgrounds, metrics, and regulatory frameworks. The critical roadmap on foundation models addresses diverse stakeholders, including policy and decision-makers, the public sector, researchers, and developers.
As the book unfolds, it illuminates the intricate interplay between society and foundation models, providing a comprehensive overview of the past, present, and potential future trajectories of foundation models in the ever-evolving landscape of artificial intelligence.
More details
Persons
Prof. Mohamed Chetouani is currently a Full Professor in signal processing and machine learning for human-machine interaction. He is the Deputy Director the Institute of Intelligent Systems and Robotics (CNRS UMR 7222), Sorbonne University (formerly Pierre and Marie Curie University). His activities cover social signal processing, social robotics and interactive machine learning with applications in psychiatry, psychology, social neuroscience and education. He was the coordinator of the ANIMATAS H2020 Marie Sklodowska Curie European Training Network (2018-2022). He was the President of the Sorbonne University Ethics Committee from 2019 to 2023. He was involved in several educational activities including organization of summer schools. He is member of the EU Networks of Human-Centered AI (HumanE AI NET) and Robotics (euROBIN). He was General Chair of ACM ICMI 2023 and VIHAR 2021. He is currently serving as Associate Editor of IEEE Transactions on Affective Computing. He co-published several books and proceedings, including the Human-Centered Artificial Intelligence (Springer, 2023). He is in charge of the inclusion of Students with Disabilities for the Faculty of Science and Engineering of Sorbonne University.
Prof. Paul Lukowicz - under preparation (main PI for the project).
Professor Andrzej Nowak is a Full Professor of Psychology at both Florida Atlantic University and the University of Warsaw, where he is also the Director of the Center for Complex Systems and New Technologies, holding a Ph.D. in psychology.He has authored or edited over 15 books and numerous influential publications in top-tier journals in psychology, physics, and interdisciplinary journals. His current research concentrates on artificial intelligence (AI), complex systems, human-computer interactions, social influence, social media, social influence, and self-structure. Dr. Nowak's pioneering work on AI includes developing models of feedback control in artificial neural networks, which has significantly advanced our understanding of how these systems can dynamically change their properties by adjusting to incoming information. He has significantly contributed to understanding complex social systems, group dynamics, and the psychological foundations of human behavior through computational models and AI technologies. His work not only advances the field of psychology but also contributes to interdisciplinary dialogues between psychologists, computer scientists, physicists, and policymakers on how social interactions and AI shape human behavior and social processes.
Content
Section 1: Foundations of Foundation Models.- Section 2: Foundations of Human AI Collaboration.- Section 3: Multimodal Foundation Models.- Section 4: Learning and reasoning with Foundation Models.- Section 5: Interaction with Foundation Models.- Section 6: Society-Large Whatever Models Interaction.- Section 7: Ethical and legal aspects of Foundation Models.- Section 8: Critical roadmap on collaborative foundation models.