Critical thinking is the silver bullet for AI-proofing students

“I think all bullets should cost $5,000,” the American comedian Chris Rock once said, in an iconic routine riffing on the debate in the US about gun control.

“You know why? Cause if a bullet cost $5,000 there would be no more innocent bystanders…People would think before they killed somebody... ‘Man, I would blow your fucking head off – if I could afford it. I’m gonna get me another job, I’m going to start saving some money, and you’re a dead man!’”

In more recent years, another peril besides gun violence has arisen to threaten society – and not just in the US: the threat of mass employment posed by artificial intelligence. But while AI will undoubtedly take over some more mundane jobs, we suggest that Rock’s standup is – to some extent – an antidote to these concerns.

To write a joke like his bullet control gag requires tremendous insight into not only the zeitgeist but also the nuances of constitutional law among other things. And it requires knowing the audience and the context. Could AI similarly make connections that no one has made before, showing us the world in a different light? Could it notice and magnify things we take for granted to reveal their underlying absurdity, as the great comedians do?

As of now, that is hard to conceive. Sure, if you ask a large language model such as ChatGPT to mine the data to write a new stand-up routine or comedy show, it will likely be able to do so. What it produces might even be humorous. But it is unlikely to break new ground. And it is unlikely to lead us to introspect and question the status quo.

That is because novelty and relevance do not stem from the kinds of frequent co-occurrences of phrases in the past that large language models mine. Doing that is a recipe for producing what is technically known as “bullshit” – statements for which their veracity or correspondence to reality are irrelevant. Instead, novelty and relevance stem from understanding what mechanisms are at play. Only profound understanding of why aids in making connections that were previously unthought of.

Art and science operate in a similar manner to the best comedy: coining a unique revelatory metaphor or developing a novel theory implies pioneering a particular symbolic or conceptual combination. And this gives us a clue as to how high school and university students can AI-proof themselves.

They would be best served by seeking to gain in-depth understanding of the mechanisms in their domain of expertise, both learning the fundamentals and staying abreast of the latest developments. The meta-skill that then enables them to extract maximum creative value out of this knowledge is critical thinking – which, of course, should be coupled with prudence. The latter advice can be summarised as follows: be humble enough to admit that you are standing on the shoulders of giants, be wise in choosing your giants, and be brave enough to question the rest.

Developing students’ critical thinking and prudence requires universities to serve as arenas for exchanging ideas and enabling students to see the world in as many different ways as possible. The latter is an indispensable skill to function well in a corporate environment, especially in leadership roles.

Students would also do well to remember that disagreement with each other does not imply disrespect, and a smorgasbord of opinions is better food for the brain than the monotony of a single world view. Faculty might also remember that bold new ideas are worthy of a hearing.

And universities should reconsider the increasing formalisation of how students are taught and assessed. A formulaic approach that seeks to increase “transparency” and “standardise” grading has the unavoidable consequence of restricting an exchange of ideas.

As faculty who have designed their own courses and have studied and taught in North America, Europe and Asia, we can certainly attest to how both teaching and assessment are impoverished when force-fitted into a narrowly defined framework of learning goals, reduced to a predetermined set of questions, and graded using a rigid rubric, among other things. The iron cage of bureaucracy, with administrators dictating how the classroom should be structured, will simply lead to the mass production of students whose skills are replaceable by another algorithmically trained “agent.”

Predicting how AI will evolve is a task best left to fortune tellers. We claim no such powers. But what we can say is that if there is something that makes us uniquely human, it is our curiosity. Curiosity pushes us to seek out new truths in science, create new worlds in arts, and notice absurdities of life that deserve a good laugh. Curiosity fuels critical thinking, thus helping us become better experts.

That is what our education system should tap into – making students curious again. For the possibility of relevant newness and real understanding emerges only when we look at the world with curiosity and enquire – why?

Akhil Bhardwaj and Anastasia Sergeeva are associate professors at the University of Bath School of Management.