AI for Adolescent Well-being

Engaging Youth from diverse backgrounds in Co-creating Ethical Chatbots

Principal Investigators: Dr. Caroline Figueroa

Collaborators: João Goncalves (EUR), Freya de Keyzer (EUR), Amy van Grieken (EMC), Danielle Remmerswaal (EUR), Eva Thalassinou (Gro-Up Youth Work), Niko Vegt (EUR, TUD), Michael Klenk (TUD), Crystal Smit (EUR), Kayla Green (EUR)

Researchers: Kathleen Guan and Nic Orchard

Planned period: March 2024-March 2025

Project summary: Our project aims to understand how we can harness the potential of AI, particularly Large Language Models (LLMs), to support the wellbeing of adolescents.

About the project

Can Artificial Intelligence (AI) help youth navigate through life's challenges, especially when it comes to mental well-being? Our project aims to understand how we can harness the potential of AI, particularly Large Language Models (LLMs), to support the wellbeing of adolescents.

However, while AI holds great promise for mental health interventions, the lack of control over these algorithms can lead to biased and harmful content. Moreover, adolescents, particularly those with migration backgrounds and lower socio-economic positions, are often overlooked in the development of AI tools for mental well-being.

In our projects we aim to understand the unique needs and perspectives of adolescents from diverse backgrounds. Using Youth Participatory Action Research, we will ensure that youth’s voices are heard, and their experiences are considered. Together with youth, our second aim is to co-create a prototype LLM assistant that is not only ethically legitimate but also tailored to address the diverse needs of adolescents, ensuring linguistic and social aspects are personalized in the model.

We will use the Erasmian Large Language Model (ELM), an open-source platform that prioritizes privacy and fair labor practices. Unlike commercially available LLMs, such as ChatGPT, ELM fosters community-wide participation and enables participatory development by involving end users directly in model training.

Through this project, we aim to pave the way for inclusive and ethical AI interventions that can be integrated into health systems to reduce the number of youths who need mental and social care, reduce the strain on youth care professions, and reduce health inequities.

 

This project is funded by Healthy Start (Sprinter)

Contact