Bristol Institute for Learning and Teaching

The idea of the human struggle is a perennial representation of how meaning is forged. It’s a truism that distils the complexity of what makes life worth living and how we shape our identities. In this blog, I want to explore the idea of struggle by looking at film representations and then reflecting on our students’ use of AI in relation to struggle and wicked problems.

In the cinematic epic Lawrence of Arabia, the titular character faces unbearable challenges to become a heroic icon. He travels through a harsh desert to face the harsher realities of war, emerging a guerilla hero. In one symbolic scene, Lawrence holds a lit match, watching it burn until he extinguishes the flame by rubbing it between his fingers. A soldier copies the manoeuvre, gets burned and calls out in pain, demanding to know “what’s the trick?”, assuming Lawrence has some method to mitigate the heat. Lawrence replies “the trick …is not minding that it hurts”.

There is an underlying philosophy here – that pain shapes who we are, and we must accept the pain along the way. In our teaching contexts, setting up challenges for students to overcome is a common pedagogic approach. We even have “wicked problems” that we know have no clear solution.

We sometimes give our students challenges knowing that they have neither the information nor the skills to completely resolve them, because the unknown forces them to be creative in their problem-solving and research skills. This can also focus them on the process of learning, the journey through the unknown, rather than on the output. It can support them to look inward instead of only on the final result.

When learners find out for themselves, the experience fires up their minds and helps with knowledge retention and intrinsic motivation. This isn’t always what learners want – so often they want us to hand them the correct answer, but is that always the best thing for their learning?

This is where we need to think about how we integrate generative AI (GenAI) into the classroom. If GenAI is part of the learning experience, how can we ensure that students aren’t skipping that important “struggle” in their learning? The good thing about wicked problems is that they are big, complicated and predicated on real-world problems, in other words, they are authentic. Topics like the climate emergency and decolonising are good examples.

Students need to ensure that they are open to different perspectives and needs, and that different people will have different modes of communication and agency in these contexts. This is a bountiful space for skills development. Teachers benefit from how these topics are easier for students to relate to, they can also model how to approach uncertainties, and create lots of venues for learning experiences in the classroom, in independent learning and through assessments. Many of these issues have a social justice angle, which can be very motivating for our Bristol students.

To make learning from wicked problems successful, it’s best to provide some examples and models to students to learn from and analyse. Start with small, scaffolded tasks before moving onto broader bigger problems that gradually build up the challenge. Along the way, teachers can remind students that there is no one resolution and that disagreements and mistakes are part of the struggle of learning and discovery.

Where does GenAI fit into this experience? That is up to you to determine, but let’s look at some key points for consideration.

• Staff need to ensure that students’ AI data literacy is good enough so that they can be discerning. If students have a lot of AI training already, you might not need to spend much time on this. But if they are new students, you will need to dedicate classroom time and resources to ensure that they know explicitly what they can use and how for this context. Students like to receive guidance that is specific to their subject and their unit.
• Importantly, especially since wicked problems include complex socio-political issues, students will need to have a firm grasp of GenAI ethics. They will need to be aware of how GenAI tools can accidentally and purposefully contain politicised data. Beyond this awareness, they will need to be proactive in their fact-checking. Biases can include things like gender, ethnicity, conflict, and scientific consensus.
• You can get your students to use GenAI for specific times in their learning journey. It might be a good idea to reject its use early on, and allow it later in the process. This ensures that students front-end cognitive application and can then be more selective about GenAI as a supportive tool. This can reduce cognitive offloading. Often, it is helpful to explicitly tell students about such pedagogic choices as it helps them value your teaching choices and understand why they are doing things certain ways.

When our students become increasingly familiar with relying on GenAI for answers and outputs, they may grow distanced from knowing that learning is sometimes a struggle and requires effort and resilience to, like Lawrence, not mind too much when it “hurts”. Active and authentic learning approaches are integral to reducing these potential harms by providing avenues for applied learning skills, learning agency and motivating metacognitive introspection.

You can find out a lot more about authentic approaches to learning and how GenAI fits in, via our staff-facing AI in Education course. Sign up via Develop. Course summary:

• This short online course supports university staff to navigate the opportunities and challenges of AI in higher education. Designed for both teaching staff and professional services colleagues, it provides clear, research-informed guidance you can apply immediately in your own context. 
• This course helps staff plan and design teaching, learning, and assessment activities that use AI responsibly. It builds skills in evaluating opportunities and risks and encourages reflection on appropriate AI use by both staff and students. You will learn to identify practical applications of AI, develop evidence-informed approaches, and consider themes such as inclusion, skills development, student reflection on their learning process, and university policy. A suite of case studies is provided to illustrate practice already happening at the University of Bristol. 
• The core content of the course should take 30 minutes to complete, plus time spent on relevant reflective prompts. All video content is optional and totals c.65 minutes of content.