About the CDT
The UKRI AI Centre of Doctoral Training in Designing Responsible Natural Language Processing is dedicated to addressing the increasing necessity for an ethically grounded approach in AI technologies. As NLP becomes more integral to applications like personal assistants (Siri, Alexa), chatbots (such as ChatGPT), and generative AI, the stakes are higher than ever. Ensuring these systems are responsibly integrated into our daily lives requires a deep understanding of their operational mechanisms and societal impacts.
Doctoral students of the centre will engage in an interdisciplinary educational experience that spans the technical, social, design, and legal domains of AI. The curriculum fosters understanding across these diverse fields, emphasizing collaboration and the development of hybrid expertise. This is your opportunity to contribute to building AI systems that are not only advanced but are also equitable, sustainable, and sensitive to the complex social fabrics they operate within.
Skills Domains
During their studies students will train across five skills domains.
Responsible NLP data and models
Students will develop techniques for building and curating inclusive and diverse datasets, as well as practices leading to fairer and less biased models.
Explainable NLP for users in-the-world
Students will be trained in techniques that enable end-users to examine datasets and models, using NLP-based techniques to provide human-readable explanations of datasets and AI models.
Designing for Human-NLP partnership
Students will use new human-centred methods to create adaptable human-AI partnership systems (e.g., human-in-the-loop, human-NLP teams).
NLP governance and accountability
Responsibly deploying NLP requires understanding the intersect with current (IPR, GDPR, Equalities, Copyright) and emerging (EU AI) laws and governance.
UK, EU and International applicants
Edinburgh. Extraordinary futures await. The University of Edinburgh is one of the world’s top universities, consistently ranked in the worl...