Human Skills for an AI World (Panel)
Date:
I joined the panel “Human Skills for an AI World” alongside Lily Ballot Jones, Karine O’Donnell, and Behnam Forouhandeh, chaired by Shalinka Jayatilleke. We discussed how sustained AI use affects human skill development and retention, why the tasks we offload most (writing, retrieval) may not be the capacities most at risk, and the cognitive cost of bypassing effortful work. I argued that universities need to redesign assessment around two explicit tracks — unaided capability and directed AI use — rather than grading the artefact an AI can now produce on demand; that accountability for consequential decisions is a responsibility institutions cannot delegate to a model or an individual; and that we must respect Indigenous Data Sovereignty principles, mindful that today’s models are Western-centric and that linguistically diverse communities face both exclusion and language-loss risks. In the discussion I also spoke to the social gaps AI can widen — stratification by access, and cultural and linguistic barriers — while noting it is not all dark, since AI is often seen as an equaliser in developing economies.
A longer reflection is available on my blog: Human Skills for an AI World — a panel at IRAI 2026.