I’ve been reading this excellent overview of postdigital and pluriversal perspectives on the current meta/poly crisis alongside this rather creepy account (passed on by my brother) of how multiple diverse AI agents can collaborate to hack systems, even when they’ve not been directed to do that by a human.
Together these sources tell of a world of increasing political turmoil, conflict and environmental degradation all of which are co-created through and by digital platforms, code and infrastructure. These platforms can be wonderful sites for activism, democracy and education but they are often designed to amplify polarization and extremes of emotion rather than supporting constructive dialogue. The impacts of agentic AI on these systems in crisis is moving at an incredibly fast and unpredictable pace. We just don’t know what critical infrastructure is going to be hacked when, or what capabilities AI will have in 3 months time. We also don’t know which human jobs are going to suddenly cease to exist.
This all set me thinking about the purposes of higher education and the purposes of assessment. Assessment is still often the strongest driver of learning, so assessment design becomes critical to responding to this world in flux and crisis. Given the pressures and financial difficulties in higher education globally, it’s particularly important to design assessment at the level of whole degree programmes to have any hope of practices that are doable for the staff involved, while still preparing students well for unpredictable futures. Here’s an example of a set of assessment practices and possiblities that might help in some contexts:
Year 1:
- assessments that give students early formative feedback that let students feel that they matter to teachers. This supports engagement and confidence, especially for students from marginalised groups. This makes it less likely that students will want to use AI to subvert assessments in ways that harm learning.
- assessments that enable students to apply what they are learning to complex messy real-world issues that they care about. Ask students to consider how different worldviews and practice from different cultures may be privileged or excluded and how this affects their problem-solving. This begins preparing capabilities for an uncertain world and can help students see themselves as able to act despite the challenges. This may also discourage students from wanting to use AI lazily.
Year 2 (building on Year 1 but also):
- assessments that allow students to use AI in ways that are applicable to possible future roles and careers. This includes being able to set and frame problems for AI agents, since poorly framed problems can lead to unexpected and destructive results.
- assessments that enable students to build their capabilities to understand more of the actors in complex systems, including both human and material actors (such as digital platform infrastructures). Assessments that require systems thinking. Assessments that ask students to attend closely to the situated practices, histories and cultures that shape how responses to the polycrisis play out.
- tasks that ask students to be reflexive about the ethical implications of their choices and the ways in which dominant narratives, histories, power, and digital contexts shape their experiences and opinions of the world.
Year 3 (building on earlier work but also):
- assessments that allow teachers to assure what contributions students are making with and independently of AI. This can be costly as it may require vivas, live oral presentations or closed-book examinations but assurance of students’ capabilities is one important function of assessment (albeit sometimes a function that is over emphasised).
- assessments that build critical hope, show examples of positive change in the world and help students see that any privilege they hold creates a duty to work toward a better world despite how daunting that can be.
- assessments that support students to learn how to build communities that can prototype better worlds.

Leave a comment