Theory of Mind May Have Spontaneously Emerged in Large Language Models

Cyber Policy Center, together with the Program on Democracy and the Internet, for Theory of Mind May Have Spontaneously Emerged in Large Language Models, a conversation with Michal Kosinski moderated by Jeff Hancock, co director of the Stanford Cyber Policy Center. This session is part of the Spring Seminar Series, a series spanning April through June, hosted at the Cyber Policy Center with the Program on Democracy and the Internet. Sessions are in-person and virtual, with in-person attendance offered to Stanford affiliates only. Lunch is provided for in-person attendance.
Theory of mind (ToM), or the ability to impute unobservable mental states to others, is central to human social interactions, communication, empathy, self-consciousness, and morality. We will present results suggesting that ToM-like ability (thus far considered to be uniquely human) may have spontaneously emerged in Large Language Models, as a byproduct of their improving language skills. We will discuss the impact of these findings on our understanding of human and artificial brains and behavior.
About the Speaker:
Michal is a Professor at Stanford University. He studies humans in a digital environment using cutting-edge computational methods, AI and Big Data. He has co-authored Modern Psychometrics (a popular textbook) and published over 90 peer-reviewed papers in leading journals including Nature Scientific Reports, Proceedings of the National Academy of Sciences, Psychological Science, Journal of Personality and Social Psychology, and Machine Learning, that have been cited over 17,000 times. He is among the Top 1% of the Highly Cited Researchers according to Clarivate. His research inspired a cover of The Economist, a 2014 theatre play “Privacy”, multiple TED talks, a video game, and was discussed in thousands of books, press articles, podcasts, and documentaries. Michal was behind the first press article warning against Cambridge Analytica. His research exposed the privacy risks that they have exploited and measured the efficiency of their methods. He holds a doctorate in psychology from the University of Cambridge and master's degrees in psychometrics and in social psychology. He used to work as a post-doctoral scholar at Stanford's Computer Science Department, the Deputy Director of the University of Cambridge Psychometrics Centre, and a researcher at Microsoft Research (Machine Learning Group).

[@xenograd4422]

seems like he is not arguing that a theory of mind has emerged, only that gpt4 has become very good at giving the right responses to prompts people would use their theory of mind to answer. Theory of mind can be interpreted as just this ability, but more commonly I think it is interpreted as implying "understanding".