Slow thinking in the age of AI

Irine Kurtanidze is a PhD Candidate at Queen Mary University of London, researching security narratives, emotional electoral mobilisation and voter behaviour in hybrid and backsliding democracies. Irine combines over a decade of experience in higher education and democracy support with expertise in political communication, campaign strategy, media literacy and quantitative research.

I recently attended the with Lord Clement-Jones, one of SSE’s Inaugural Professors of Practice.

In the talk, Lord Clement-Jones reflected on AI, education and the future of graduate skills and explored how universities can shape, assess and showcase their students’ learning outcomes.

System 1 and System 2 thinking

One point I found very interesting, and keep returning to, was his observation that AI can encourage ‘System 1 thinking’, which is fast, intuitive and relatively effortless. Learning, however, requires ‘System 2 thinking’, which is slower, deliberate thinking that involves questioning assumptions, working through difficult ideas and forming our own judgements.

System 1 and System 2 thinking were originally introduced by psychologists Keith Stanovich and Richard West in 2000, but the framework was popularized globally by Nobel Laureate Daniel Kahneman in his 2011 book, Thinking, Fast and Slow.

Much of our everyday judgement is shaped by System 1, which allows us to respond quickly without deliberating over every decision. System 1 applies to familiar ideas and explanations that feel more convincing, even if they are not supported by evidence.

System 2 corrects for errors in system 1 thinking but demands effort. We tend to engage System 2 thinking only when something forces us to slow down.

This is where I see a potential risk in students’ constant use of AI.

Easy answers = superficial understanding

AI offers fluent answers almost instantly, removing many of the moments of friction that would normally push us to slow down and activate System 2. Because AI produces quick and ready answers, humans tend to accept an explanation without examining how it was reached.

When AI does the work for students, it may produce a polished and clear answer, but the student never develops the understanding needed to explain, defend or apply it to real life situations.

When humans invest effort into recalling information or working out an answer, we strengthen our memory and deepen understanding. If we can follow an argument or an idea, we feel that we understand it. But without putting any energy into constructing, questioning or defending it, we skip the ‘learning’ part of acquiring new knowledge.

AI means that students no longer need to spend energy and effort doing these things, and miss out on deeper, more impactful learning.

Teaching students to question their thinking

In every class I have taught, I have tried to introduce students to System 1 and System 2.

The concept is not tied to any one discipline, because it is about how all of us think and make decisions. It helps students notice that their first reaction to an argument, a news story or an exam question is often produced automatically, before they have had a chance to reason about it.

Once they recognise this, they can learn to take a moment, step back and ask why they made a particular decision or reached a particular conclusion, and whether they would reach the same one if they slowed down.

In my experience, this habit of questioning their own thinking is one of the most valuable things students can take from university, whatever they go on to do.

What we are doing at SSE

Like many academic colleagues across our sector, I worry about what this means for our students. How can we use AI while preserving the intellectual effort that helps them develop independent, critical thinking?

SSE, in wider collaboration with Queen Mary University, is developing a strategy to protect the intrinsic and extrinsic value of the degrees we award our students. In order to keep students questioning and learning in the age of AI, we have divided the ideal student learning journey into three distinct stages.

We are designing innovative classroom activities and assessments that align with each one, so that our students graduate with the confidence and the skills to enter into the AI-driven workplace.

Read more about AI and assessment at SSE.

This article is based on an original Linkedin post.

Follow Irine on Linkedin