Is Artificial General Intelligence Possible?
Artificial intelligence research has had remarkable success in recent years with machines like ChatGPT and other large language models. This has led to renewed enthusiasm for the idea of Artificial General Intelligence (AGI) — intelligence which is not just tailored to specific tasks, but something more like the intelligence human beings have. Some researchers have said that AGI is on the horizon in the next few decades. However, the question of whether there can be AGI is not just a technological question, but one that involves theoretical or philosophical assumptions about the nature of intelligence, thinking and computation. In this talk I present the outline of an argument that AGI is not possible, so long as we are restricting ourselves to computational artificial intelligence.
Biography
Tim Crane is a Professor of Philosophy at the Central European University (CEU) in Vienna, where he is also Pro-Rector for Foresight and Analysis. Before coming to CEU, he was the Knightbridge Professor of Philosophy at the University of Cambridge, and before that he was Professor of Philosophy at UCL in London. He founded the Institute of Philosophy in the University of London in 2005. He has been the President of the Aristotelian Society, the UK’s oldest philosophical society. In his philosophical work he has investigated questions about the nature of the mind in the most general sense. For example: what does it mean to have a mind? What is it to think? What is it to be conscious? How are thought and consciousness related? Are there any deep and essential differences between human minds and the minds of other animals? What is the relationship between the knowledge which we all have of our own minds and the knowledge that we get from neuroscience and psychology? What does it mean for the mind to be physical or material? His first book, The Mechanical Mind, an introduction to the philosophical questions surrounding Artificial Intelligence, was published by Penguin Books in 1995 (third edition 2016). His other books include Elements of Mind (Oxford University Press 2001), The Objects of Thought (Oxford University Press 2013) and Aspects of Psychologism (Oxford University Press 2014). His most recent book is The Meaning of Belief, an examination of the nature of religious belief, published by Harvard University Press in 2017. Crane’s work has been translated into Arabic, Chinese, Croatian, Estonian, French, German, Hungarian, Italian, Japanese, Korean, Persian, Polish, Portuguese, Romanian, Spanish and Swedish. Crane is the Director of Research of the Cluster of Excellence ‘Knowledge in Crisis’, a major collaboration starting in 2023 between the CEU and the universities of Vienna, Graz and Salzburg, funded by the Austrian Science Fund, the FWF. He is currently working on the nature of belief, the nature of the unconscious, and philosophical questions about AI.
Website: www.timcrane.com and https://people.ceu.edu/tim_crane