‘Magnifica Humanitas’, AI, and suffering

Like many of you, I have been reflecting on Pope Leo XIV’s Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence, his first encyclical, which offers a compelling invitation to reconsider what it means to be human in an era increasingly shaped by algorithmic power. Among its many provocative insights, one line in particular has stayed with me:

“To eliminate suffering entirely would mean, in the end, extinguishing love and desire as well.”

This statement challenges a dominant narrative which is prevalent in technological and economic discourse, and increasingly visible in education, one in which progress is defined by reducing, optimising, or entirely removing friction, discomfort, and difficulty. I find this especially troubling in educational contexts. The idea that learning should be effortless sits uneasily with me, because genuine learning is not simply about acquiring knowledge, but about growing into it. And growth inevitably involves growing pains, in this case, discomfort, uncertainty, and struggle to grasp what we do not yet know or master.

In this reflection, I want to revisit that idea through the lens of the work Mark Murphy and I have written on Gen-AI and education (see here and here)

The technological promise: A nightmare disguised as a dream?

AI represents a turning point in the history of automation. For the first time, knowledge work itself is being automated at scale. This is not merely a technical change; it is an anthropological one, aiming to reshape how we relate to information and knowledge. But it need not mean anthropomorphising the machine. In education, where technological developments have long been framed through the language of optimisation, Gen-AI, and in particular LLMs, intensifies existing tendencies: personalisation, efficiency, and now also the proliferation of “learning assistants” all promising smoother, faster, and less effortful (learning) processes and experiences. We are, after all, in an age where technology is commodified as a service, and this is where the problem begins.

However, as we have argued in our publications on Gen-AI, these affordances do not simply enhance existing practices; they rewire them. When generated content is adopted as if it were one’s own, knowledge becomes detached from experience. What emerges is a kind of epistemic pretence: knowing without having known, or as I often say, experts without expertise; understanding without having struggled with the process of thinking.

At the heart of this transformation lies what we could call the “outsourcing of struggle.” Traditionally, learning has been understood as a process marked by one’s endeavour, but also ambiguity and even frustration. These are not incidental features; they are constitutive of the learning process itself. Gen-AI, however, reframes learning as something that can be streamlined to the point of becoming almost imperceptible, something that can be completed without being “felt” or experienced. It aims to reformulate learning as an emotionless, empty task.

It is here that Magnifica Humanitas offers a powerful intervention. The encyclical suggests that eliminating suffering is neither a neutral nor a purely beneficial goal. Rather, it fundamentally alters our relationship to ourselves, with others and the world around it.

Pope Leo XIV’s assertion that suffering is inherently bound to love and desire echoes a long-standing understanding of desire as emerging from absence, from a sense of incompleteness and vulnerability. In this light, desire within the context of Gen-AI presents a double bind: without encounters with limitation, of not knowing, lacking, or failing, there can be no genuine transformation toward understanding or meaningful connection, whether with knowledge or with others. Yet, desire in a commodified world is also one of haves and wants, which are promised to be achieved without effort, albeit at a price, the price of dormancy, or a quiet submission to algorithmic convenience. We desire precisely because we do not fully possess something; we seek because we do not yet understand, etc. Without the tension caused by our perceived limits, uncertainty, or struggle, there is no substantive movement toward new insights/ideas or relational depth.

What the Pope’s view foregrounds, then, is that epistemic struggle is not a flaw in or of learning; it is its very condition of possibility.

When Gen-AI tools pre-emptively generate answers, summarise complexity, and circumvent interpretive effort, they may produce efficiency, but they also make us lose something: they reshape our relationship with knowledge. Learning can then become a matter of output rather than a dialectics of thinking and understanding; a transaction rather than a transformation.

This then leads me to considerthat there is an ethical dimension as well. Gen-AI mediates relations of care and responsibility by removing the need for struggle. Yet in doing so, it risks removing learning itself. In an Arendtian sense, such support can function as a crutch, a substitute that undermines the development of independent judgment and critical thought. This raises an important question: who, or what, is responsible for learning?

If knowledge production is increasingly delegated to AI, human agency becomes secondary to algorithmic suggestion. In such a context, the broader structures of performativity, i.e., metrics, outputs, efficiency, readily accommodate Gen-AI as yet another tool within technocratic governance.

However, this moment also presents an opportunity. Instead of framing Gen-AI as a tool for eliminating effort, can we reimagine it as a new actor in epistemic practices? The goal of education, then, would shift from optimisation to formation (Building – also see Goethe): not simply producing outcomes, but cultivating individuals who are capable, critical, and engaged knowers, or as Freire would put it, know-how to learn and how to be in a digital society.

The idea that suffering, when understood as effort, limitation, and struggle, has intrinsic value requires resisting the urge to equate technological progress with the removal of all obstacles. What technologies treats as inefficiencies may, in fact, be essential to what makes us human.

The challenge is not to reject Gen-AI outright, but to reconsider how it is integrated into our practices. Rather than erasing difficulty, it would be better if it could coexist with the productive tensions that sustain learning, desire, and human connection.

This is so because ultimately, if we succeed in designing a world without struggle, we may well be designing one without meaning. And that is the antithesis of progress.

References:

Costa, C., & Murphy, M. (2025). Critical education, generative artificial intelligence and the tyranny of freedom: a critique of modern ‘technocracy’. Technology, Pedagogy and Education. Advance online publication. https://doi.org/10.1080/1475939X.2025.2547728

Costa, C., & Murphy, M. (2025). Generative Artificial Intelligence in Education: (What) Are we thinking?. Learning, Media and Technology. Advance online publication. https://doi.org/10.1080/17439884.2025.2518258