26/05/2026
In the age of AI, data is no longer just a supporting component of research โ it has become the foundation of intelligent systems.
Iโm pleased to share the full recording of our session on โ๐จ๐๐ฒ ๐ผ๐ณ ๐๐ ๐ณ๐ผ๐ฟ ๐๐ฎ๐๐ฎ ๐๐ผ๐น๐น๐ฒ๐ฐ๐๐ถ๐ผ๐ป & ๐๐ฎ๐๐ฎ๐๐ฒ๐ ๐๐ฟ๐ฒ๐ฎ๐๐ถ๐ผ๐ปโ conducted as part of the Research in the AI Era series at the University of Peradeniya.
Together with Dr. Isuru Nawinne, we explored how researchers can leverage AI to improve data collection, annotation, validation, and dataset curation while also addressing the growing concerns around bias, ethics, synthetic data generation, and reliability.
One important insight from the discussion was the need for developing datasets that truly represent our own contexts, communities, and real-world challenges. Many AI systems still fail to accurately represent regions like Sri Lanka simply because our data is underrepresented in global datasets.
The session also covered:
โ
AI-assisted annotation and labeling
โ
Human-in-the-loop validation approaches
โ
Ethical risks in AI-generated datasets
โ
Medical and research data challenges
โ
Synthetic data generation techniques
โ
Building scalable data pipelines for AI research
A heartfelt thank you to the organizers, participants, and everyone who contributed to the discussion.
๐ฅ Full session available on YouTube now : https://www.youtube.com/watch?v=2P6i-Eji0Y0
I hope this discussion will encourage more researchers and students to think critically about the role of data in AI innovation.
