Thierry Desot

Thierry Desot is a Lecturer in Applied Natural Language Processing at the Creating 010 knowledge center within Rotterdam University of Applied Sciences. His research focuses on the ethical application of AI in healthcare, particularly addressing biases in large language models that could lead to diagnostic inaccuracies affecting minority groups.

• Education: PhD in Spoken Language Understanding (University of Grenoble-Alpes, 2017-2020); Postdoctoral research at Ghent University (2021-2023).

Desot investigates how biases in AI systems like ChatGPT impact healthcare decision-making. His work emphasizes responsible AI and transparency in NLP technologies, with a focus on early disease detection. He led a public lecture on November 28, 2024, discussing these challenges.

Key projects include TheraVatars, exploring virtual humans for healthcare training. His publications highlight NLP's role in mitigating bias and improving diagnostic accuracy.