Vanichkina D1
1Sydney Informatics Hub, University Of Sydney, Newtown, Australia
Biography:
Darya Vanichkina PhD SFHEA is the Data Science & AI Group Lead at the Sydney Informatics Hub, a University of Sydney Core Research Facility dedicated to enabling excellence in data and compute-intensive research. SIH is the University's AI for research "shopfront". Darya leads the consultancy and training team driving research outcomes and funding, accelerating projects, and helping academics foster partnerships with industry and government in this space.
https://orcid.org/0000-0002-0406-164X
Abstract:
Generative Artificial Intelligence (AI) tools are becoming adopted by researchers worldwide. As the lead of the Data Science & AI team at the Sydney Informatics Hub (SIH), the University of Sydney's "AI for research shopfront", my team and I have been supporting academics from across all faculties and career levels as they integrate these tools into their work. In the past 12 months, this has included dedicated consultations, ad-hoc questions, casual conversations, emails and every combination thereof with ~100+ research teams.
In this talk, I'll summarise the stories we see time and time again:
1. The double-bind between GenAI's usefulness and research needs for reliability and reproducibility
2. The perils of quick evolution and model sunsetting for academic work
3. The role local models can play – and why researchers don't reach for them
4. How using AI well depends on everything we've championed for a decade
5. Why qualitative researchers need distinct support
6. Ethics, guidelines and cybersecurity playing catch-up at unprecedented speed
7. What predating ethics approvals and consent mean for data used in AI tools
This session will be of interest to eResearch professionals supporting AI adoption at other institutions, research computing providers considering how to ensure advanced infrastructure actually gets utilised, national platforms designing shared workflows and resources – and policymakers considering how to develop frameworks that balance the need for innovation with responsible, ethical use of this technology.
