AI and psychology: Towards a fruitful synergy
This project focuses on studying possible links between psychology and AI (with a particular focus on large language models). We will explore ways in which these research areas can connect and benefit from each other. Given its focus, this is an interdisciplinary project involving experts in computer science and psychology. Specifically, the project will focus on the following directions:
1. Use of large language models (LLMs) to assess health-related variables and automatically understand people's narratives:
1.1. Simulating populations to test the psychometric properties of different questionnaires designed to assess health-related variables (e.g. quality of life, disability or symptom severity). 
1.2. Testing the ability of LLMs to interpret, understand and assess the personal narratives of people with chronic pain. 
2. Implementation of psychology-based tests for the development of trustworthy and responsible AI technology:
2.1. Determining the validity of classic human assessment questionnaires for assessing LLMs (see e.g. ). The aim is to use psychological tools to improve LLMs in the context of developing ethical AI applications.
We are open to hearing other innovative ideas related to AI and psychology from potential candidates.
 Griffin, L., Kleinberg, B., Mozes, M., Mai, K., Do Mar Vau, M., Caldwell, M., & Mavor-Parker, A. (2023). Large language models respond to influence like humans. In Proceedings of the First Workshop on Social Influence in Conversations (SICon 2023) (pp. 15–24). Association for Computational Linguistics. https://10.1093/pm/pnz147
 Nieto, R., Sora, B., Boixadós, M., & Ruiz, G. (2020). Understanding the experience of functional abdominal pain through written narratives by families. Pain Medicine, 21(6), pp. 1093–1105. https://doi.org/10.1093/pm/pnz147
 Hagendorff, T. (2023). Machine psychology: Investigating emergent capabilities and behaviour in large language models using psychological methods. arXiv. https://doi.org/10.48550/arXiv.2303.13988