Winning strategy: Setting guardrails for generative AI

Published in The Australian Financial Review

Large language models like OpenAI’s ChatGPT and Google’s Gemini are poised to change how we live and work. Our team’s research into the impact of generative artificial intelligence tools on learning has shown us that these systems often provide excessive assistance, finding they over-help almost 50 per cent of the time, even when prompted otherwise, potentially harming learning.