ARTIFICIAL INTELLIGENCE AND THE PEDAGOGY OF LIMITS: HUMANITY’S MOST URGENT DEBATE
What if the real danger is not that machines will become more like us, but that we will end up becoming more like them?
Ever since January 2023, when I began exploring how artificial intelligence might work in tandem with the knowledge and experience humanity has accumulated across millennia, continents and fields of endeavour, one conviction has grown stronger: AI is among our greatest technological achievements. It is also the source of one of our most dangerous temptations — to mistake extraordinary technological power for a form of omnipotence.
And I do not believe the most important question raised by AI is purely technological. It is philosophical, anthropological, political and, ultimately, spiritual.
I would approach this question through five interconnected concerns.
1. The limits of artificial intelligence
AI can surpass human performance in a remarkable range of intellectual tasks: processing information, detecting patterns, developing hypotheses and solving problems of extraordinary complexity.
Yet intelligence and wisdom are not the same thing. Intelligence helps us discover how to do something; wisdom asks us whether it is worth doing in the first place.
A system may identify the most efficient course of action. Efficiency, however, does not confer moral authority to decide what is just, dignified or desirable.
AI also faces profound limits as a source of knowledge. It can make mistakes, fabricate information, reproduce existing biases and deliver persuasive answers without fully grasping the human circumstances in which those answers will be used.
Nor do we know whether an AI system could ever develop consciousness. There is no scientific consensus on how we might determine whether it has.
2. The risk of technological omnipotence
Here lies a historical paradox that I find fascinating. Modern science fundamentally changed our understanding of what human beings could achieve, giving us ways to overcome boundaries that once seemed immovable.
AI now presents a more unsettling possibility: as our ability to shape the world expands, we may begin to assume that no limits need apply to us at all.
But the ability to do something is not the same as the right to do it.
Human history is full of inventions whose social, political and moral consequences went far beyond what their creators could foresee.
Nuclear fission offers an especially telling example.
3. The five risks that concern me most
I would identify five principal risks associated with AI, which I would describe as follows:
a. The surrender of human judgment. Individuals, institutions and governments may begin to delegate important decisions without fully understanding them or questioning their purpose and potential consequences.
b. The concentration of power. A handful of corporations or states may come to control the infrastructure, data and models that shape much of our economic and social life.
c. The erosion of our collective ability to distinguish truth from falsehood. The mass production of synthetic images, voices, texts and arguments may make it increasingly difficult for society to distinguish fact from fabrication.
d. The loss of autonomy. As systems become more capable, they may shape our preferences and behaviour, alter our relationships and influence our choices, sometimes through forms of manipulation that are extremely difficult to detect.
e. Losing control of advanced systems. Increasingly autonomous AI could pursue objectives in ways that conflict with human interests, particularly when it operates, in whole or in part, critical infrastructure, cybersecurity systems or military technologies.
This last danger deserves particular attention. Experts disagree about the likelihood of extreme scenarios, but the seriousness of their possible consequences warrants rigorous assessment of likelihood and impact, robust safeguards and continuous scrutiny of how effectively those safeguards work.
4. The deeper danger: spiritual impoverishment
To me, this is the most compelling — and the most pressing — dimension of the debate.
Imagine a world in which AI can generate almost any kind of content, solve an immense range of practical problems and satisfy an ever-growing share of our material needs.
Would we necessarily live better lives?
Not necessarily.
To be human, yesterday, today and tomorrow, is about far more than solving problems through ingenuity or technology. It is about loving and suffering, nurturing hopes, dreams and plans, forgiving, reflecting deeply, cultivating virtues, finding meaning in our relationships and giving purpose to life itself.
AI can help us examine and describe those experiences, and even reflect on them with us. It cannot, however, replace the human experience of living through them.
This is where I see a particularly important link with the research that led me to write my latest book, “AI and the Wisdom of Popular Religiosity: Human Limits and the Dream of Omnipotence” (2026), which explores human limitations and our dream of omnipotence.
Religious traditions and forms of popular spirituality remind us of something our efficiency-obsessed culture too easily forgets: not everything of value can be expressed in numbers; not everything that matters produces tangible results; and not everything we can control is something we truly need.
A procession, a pilgrimage, meditation, fasting or communal prayer — practices I have studied across Africa, the Americas, Asia, Europe, the Near East and Oceania — are not exercises in productivity. They nurture aspects of life that cannot be reduced to a computational operation.
5. We Need a Pedagogy of Limits
I believe a pedagogy of limits could become one of the intellectual foundations of a serious philosophy of AI.
This is not an argument for halting technological progress.
It is an invitation to distinguish clearly among three fundamental questions:
• What can we do? The question of technological capability.
• What should we do? The question of ethical responsibility.
• What is worth doing? The question of wisdom and the purposes of existence.
A civilization that develops the first capacity brilliantly while neglecting the other two may become more powerful without becoming any wiser.
That is why I believe spiritual traditions have an important contribution to make to the contemporary debate about AI. Not because they offer technical solutions to its challenges, but because centuries of reflection have given them something profound to say about power, temptation, humility, responsibility and what it means to be human.
My conclusion
One thought strikes me as especially powerful and revealing: AI might prove far less dangerous if machines come to resemble us than if we allow them to teach us to behave like machines.
In the latter case, we could trade thoughtful reflection for speed and efficiency, and human discernment for computation. Genuine human connection could give way to artificial substitutes, while the accumulation of information displaces wisdom.
But none of this is inevitable. AI could also free us from repetitive tasks and open extraordinary possibilities for science, education, healthcare and creativity.
The future of this human–AI partnership will depend, in no small measure, on a question I was already exploring in September 2024, as AI use was only beginning to reach a mass audience, in my book “Timeless Insights on AI: Artificial Intelligence through the eyes of History´s Greatest”.
In that book, I drew upon the voices and insights of hundreds of extraordinary figures from a wide range of human endeavours, spanning thousands of years and all five continents. They became our “sherpas”, helping us navigate the unfamiliar territory of artificial intelligence. After all, every emerging technology throughout history has had its techno-optimists and its techno-pessimists.
One lesson from that journey through human history remains as relevant as ever: our future with AI will depend on the institutions we build, the decisions we make and the values we choose to preserve.
The defining question of the twenty-first century is not simply how far artificial intelligence can go. It is how far it ought to go — and what it means to remain human in a world where machines can do ever more of what we once thought was ours alone.
Could it be that, precisely by extending our capacities beyond anything we once imagined, AI is forcing us to rediscover the importance of our limits?
That paradox lies at the heart of one of the most urgent debates facing humanity today. With “AI and the Wisdom of Popular Religiosity: Human Limits and the Dream of Omnipotence” (2026), I hope to contribute to that conversation — not by offering definitive answers, but by raising questions whose answers we must seek together. And by “together”, I explicitly mean including people from the many regions and geographies beyond the places where this technology is developed with ambitions of universal reach. This may be the most important conversation of our present — and our future.
Alejandro Félix de Souza
About the author
Alejandro Félix de Souza is a Panamanian–Uruguayan social and political analyst, writer, entrepreneur, senior adviser to high-profile leaders and former diplomat. He is the author of “Timeless Insights on A.I.: Artificial Intelligence Through the Eyes of History’s Greatest” (2024), “The Venusian Perspective: Great Women in History and Artificial Intelligence” (2025) and “AI and the Wisdom of Popular Religiosity: Human Limits and the Dream of Omnipotence” (2026). His work explores artificial intelligence through the lenses of history, ethics, humanism and the wisdom humanity has accumulated across civilizations.
Comentarios