3rd KATHY WILKES MEMORIAL CONFERENCE
The Behavioral and Ethical Consequences of
Large Language Models (LLM)
Università di Torino
Palazzo Badini
Dipartimento di Psicologia – Sala Lauree
Via Giuseppe Verdi, 10
26-27 April 2024
Large Language Models – including ChatGPT – are now being widely used – raising a range of challenging ethical, legal and behavioral concerns - which this conference will set out to discuss and address. Some even term this a turning point in the history of mankind.
LLMs can be fun, but also search information, answer the most complex questions, write articles, create musical scores or poems in the style of Nobel prize winners, provide guidance and advice on any question, drawing on all online published material.
What is the moral standing of such LLMs or ChatGPT advice? Should we worry that a ‘black box’ replaces human intervention, or should we be pleased that in many cases it can be quicker and more thorough? What will be the impact on many human decisions such as recruitment, choice of tender projects, performance evaluation, promotions, attribution of credits, financial decision-making and so. Does ‘neutrality’ as a concept operate, and can we counter stochasity from the web which might recycle sexist, racist and even violent terms and concepts.
Are there ethical rules that should be built into these LLM systems? Can we re-design LLMs with built-in ‘ethics’? How do we spot ‘faked’ LLM products masquerading as originals in the creative arts, in academic circles, in journalism? Transparency, digital literacy, specialist training, will surely all help. But how do enforce it and check on all this?
This Conference will confront many of these issues/ It will draw on the European Union AI code, passed in June 2023, to regulate artificial intelligence, and set up a potential model for global policymakers as they grapple with setting guard-rails on the rapidly developing technology.
Programme
FRIDAY 26/4/2024
H 14:00-14.45 Welcome Remarks: Monica Bucciarelli, Anita Avramides, Nada Bruer, Riccardo Viale
Chair: Riccardo Viale
H 14:45-15:45, David Papineau, “The Moral Standing of Artificial Creatures”
H 15:45-16:45, Edoardo Datteri, “Psychological human likeness and people's folk-ontological stances towards robots”
H 16:45-17:30, Tea-Coffee break
H 17:30-18:30, Tvrtko Tadić, “The Science and Practice of Large Language Models: An Industry Perspective
SATURDAY 27/4/2024
Chair: Anita Avramides
H 9:30-10:30,Lukas Meier, "Moral Dilemmas: Large Language Models vs Fuzzy Cognitive Maps"
H 10:30-11:30,Tomislav Furlanis, “LLM’s and narrative misalignment”
H 11:30-12:15, Coffee/Tea
H 12:15-13:15, Mario Rasetti, “Large Language Models: a new paradigm to ‘measure’ behavior”
H 13:15, Conclusion
Speaker’s bio
Monica Bucciarelli is Full Professor of Psychology and President of the Center for Logic, Language, and Cognition at the University of Turin. She is the coordinator of the LRR research group (Learning, Reasoning, Remembering group) at the Department of Psychology of the University of Turin.Her research interests include deontic reasoning, kinematic mental simulation in action perception and action memory, deductive reasoning, and abductive reasoning. She collaborates with the Department of Computer Science at the University of Turin to develop trainings that promote children’s understanding of the principles underlying computer programming and machine learning. She has been invited to speak at national and international conferences, including the Irish Conference on Artificial Intelligence and Cognitive Science, Dublin, Trinity College, 2018. She teaches General psychology and Reasoning psychology. She is a member of editorial boards of national (Giornale Italiano di Psicologia) and international journals (Memory & Cognition, Journal of Cognitive Psychology).
Anita Avramides is Emeritus Fellow at St Hilda's College, Oxford. She was Senior Research Fellow at St Hilda's from 2020-23. Prior to that she was Southover Manor Trust Fellow in Philosophy at St Hilda's College and Reader in Philosophy of Mind at Oxford University. She has served as Vice Principal of St Hilda's and Junior Proctor for Oxford University. She is the author of Meaning and Mind: An Examination of a Gricean Account of Language (MIT:Bradford Books), Other MInds (Routledge) and co-editor of Knowing Other Minds (OUP).
Nada Bruer Ljubišić is the Executive Secretary of the Inter-University Centre Dubrovnik (IUC), an independent international institution for advanced studies, operating in Dubrovnik, Croatia. She is coordinating the implementation of different academic programmes, organised by professors and scientists from around the world. The majority of these programmes are from the fields of humanities and social sciences and address contemporary social and academic issues. The IUC emphasizes and supports inter-disciplinary and cross-national collaboration on global challenges such as peace, universal human values and rights, health, education, and climate change. In addition to east-west cooperation, the IUC encourages new north-south initiatives, trying to contribute in overcoming the deepening of global divisions.
Riccardo Viale is Professor of Behavioral Sciences and Cognitive Economics at the Department of Economics of the University of Milan Bicocca and of Behavioral Economics at LUISS, Rome. President of Behavioral Insights Bicocca and Co- Scientific Director of the Master “Behavioral Design of Public Utilities and Services”. Early in his career he was CNR-Nato Fellow at Balliol College, Oxford University, Research Scholar at Italian Academy of Columbia University, at Max Planck Institute for Human Development, Berlin and at Chinese Academy of Science, Bejing. He was European Research Council reviewer and panel member of Mind and its Complexity Consolidator Grants. He is currently Head of Italy Behavioral Insights Team (Team di Analisi Comportamentale) at Prime Minister Department of Civil Service. Founder and Secretary General of Herbert Simon Society. Editor in Chief of Mind & Society (Nature-Springer). Author of many books and publications including the Handbook on Bounded Rationality (Routledge, 2021), Nudging (The MIT Press, 2022), Cognitive Economics (The MIT Press, forthcoming) and coeditor with Gerd Gigerenzer and Shabnam Mousavi of the Companion to Herbert Simon (Elgar, 2024). His previous empirical research was in category-based induction, conjunction fallacy, vague predicates and deductive reasoning. Recent empirical research is on social norm nudging and behavioral design in public policy and public administration. His theoretical research focuses on the cognitive foundation of the methodology of social sciences, and on the wide embodied dimension of bounded rationality.
David Papineau has held academic posts at the universities of Reading, Macquarie, Birkbeck London, and Cambridge. Since 1990 he has been at King's College London, and for 2015-20 he also spent the second half of each academic year at the Graduate Center of the City University of New York. He has served as President of the British Society for Philosophy of Science, the Mind Association, and the Aristotelian Society. His books include Theory and Meaning (1980), Philosophical Naturalism (1992), Thinking about Consciousness (2002), Philosophical Devices (2012), Knowing the Score (2017) and The Metaphysics of Sensory Experience (2021). He is currently working on causation.
Edoardo Datteri is a Full Professor of Logic and Philosophy of Science at the University of Milano-Bicocca, and Director of the RobotiCSS Lab (Laboratory of Robotics for the Cognitive and Social Sciences). His research is focused on the epistemological and methodological foundations of the cognitive sciences. His work delves into how robots, computer simulations, and bionic systems can contribute to our understanding of animal behavior and cognition. This research touches on crucial issues related to scientific explanation and modeling within the field of cognitive science. Additionally, he is actively involved in overseeing and participating in empirical research on human-robot interaction within the RobotiCSS Lab. This includes investigations into the attribution of cognition to robots and the use of robots as platforms for studying theory of mind in humans. The outcomes of his epistemological research have been published in both philosophical and robotics journals.
Tvrtko Tadić is a Senior Applied Scientist at Microsoft Search, Assistant and Intelligence, where he develops intelligent search solutions for Microsoft Office. He belongs to the Graph Intelligence Sciences team, which leverages the complex network of Office 365 entities (such as documents, messages, users, events, etc.) to enhance search and discovery within Office 365. His team is currently working on
integrating the insights from this entity network with large language models. Tvrtko Tadić has been awarded many prizes and honors, such as the Croatian Power Company Award for High School Students, the Croatian National Institute of Education Honor, the UW Department of Mathematics Academic Merit Award, and the McKibben & Merner Fellowship.
Lukas J. Meier is a Junior Research Fellow at the University of Cambridge and a Fellow at the Carr Center for Human Rights Policy at the John F. Kennedy School of Government. Lukas studied philosophy at the University of Oxford and political science at the University of Göttingen. As part of a team developing an algorithm for ethical decision-making in the clinic, he also spent a year at the Technical University of Munich. His doctoral thesis, completed at the Universities of St Andrews and Heidelberg, linked the topic of brain death to the debate on personal identity. After exploring questions of distributive justice and triage in the COVID-19 pandemic, his current research focuses on the interrelation between brain-computer interfaces, machine intelligence, and consciousness. Lukas teaches in ethics, metaphysics, philosophy of mind, political philosophy, and medical ethics.
Tomislav (Tom) Furlanis is an academic researcher with a PhD in AI ethics, a narrative, board game, designer, and a motivational speaker with two decades of experience in diverse cultural environments. Tom's research expertise lies in the ethical conceptualisation and analysis of symbiotic human-AI systems that facilitate the augmentation of human capabilities, especially human morality. His current research is currently situated in fields of human-robot interaction, brain-computer interfaces and, more recently, digital democracy.
Mario Rasetti, President of the Science Board of CENTAI Institute, is an Emeritus Professor of Theoretical Physics at Politecnico di Torino. He holds Master's degrees in Nuclear Engineering and Mathematics, and a PhD in Theoretical Physics. Prof. Rasetti's career includes affiliations with prestigious institutions such as, the Institute for Advanced Study at Princeton, Yale University, Miami University, and King’s College London.
He is a recipient of several awards including the Volta Medal for outstanding contributions to science, the Majorana Prize for Quantum Field Theory and the IBM Plaque for Quantum Computation.Notably, he served as President of the ISI Foundation and was honored with the title of ‘Grande Ufficiale al merito della Repubblica Italiana’ by the Italian President. With over 350 publications spanning theoretical physics and mathematics, Prof. Rasetti's research encompasses diverse scientific paradigms, from Quantum Mechanics to Complex Systems Science, from Statistical Mechanics to Quantum Computation and Information, from Group Theory and Algebra to Artificial Intelligence. Currently, he focuses on the application of advanced algebraic topology methods to AI, with successful implementations in Neuroscience and Economics