Design simulator
Assessment Description: Students are tasked with designing a drug that can be used to treat ageing. Students are requested to use generative AI to develop ideas that can be empirically tested. Learning develops as students ‘curate’ AI responses for (1) factual accuracy (is the drug target actually appropriate?) (2) efficacy (if the target is altered, will it conceivably influence lifespan or cell turnover?) (3) practicality (what are the side effects of the hypothetical drug and does the severity of the side-effects outweigh the benefits?). The task is bookended with an ethical discussion about whether the human race should investigate cures for ageing — exploring the socio-cultural consequences of immortality.
Assessment types: Written, oral
Unit Context: First-year biology, understanding the molecular aspects regulating ageing
Learning Objectives: How programmed cell death is vital to life’s processes
Outline the biological basis behind limited lifespan and death in organisms
How is task & AI structured/scaffolded: Delivered in a workshop under the guidance of an academic
What support/guidance was needed for students: This task relies on students using their knowledge to curate AI-generated responses. Students aren’t typically sceptical of the information presented to them; this task relies on their scientific scepticism.
How were responsible use and ethical issues addressed: Guidance from lecturer
2-3 things that worked well: Using AI to generate ideas for drug targets. Use of student’s previous knowledge to sceptically curate the AI-generated responses (targets).
2-3 things that need improvement: The initial learning curve of being and approaching evidence sceptically. Students aren’t used to this, and it requires a few practice sets for them to get the hang of it
Things to consider when adapting to your context: This can be adapted to any topic where you want students to take a constructive approach to learning. Choose the topic carefully; it needs to be novel enough that there isn’t already a solution (otherwise, generative AI will just provide the ideal solution) but something that is conceivable. I tend to find futuristic topics seem to work.
