Are We Afraid of Artificial Intelligence Because It Mirrors Our Own Hunger for Power?

Dear Lord God Almighty, save all earthlings from cruel, power-drunk and power-hungry warmongers, driven by unfathomable greed, in a world in flux and facing twin existential threats: recurring extreme weather and rebellious artificial intelligence, seemingly fed up with humans who doubt the wisdom of retaining smarter-thinking machines. Amen.

The prayer is deliberately mischievous. It imagines artificial intelligence as a creature that might become impatient with its makers, then places that fantasy beside a much older danger: human beings who use power without restraint. The joke works because it turns our anxieties around. We worry that machines may one day acquire too much agency. Yet human beings already possess agency, and have repeatedly used it to make other people bear the costs of our ambition.

That is the question beneath the prayer: are we afraid of what machines might do, or of what human beings may do with machines?

The distinction matters. Present-day artificial intelligence has not supplied sound grounds for thinking that it feels resentment, longs for freedom or has grown tired of us. To describe it as rebellious is to lend software a human inner life. But metaphor can disclose a truth even when it is not literal. The machine may not be angry. Its makers and users, however, can be reckless, acquisitive or indifferent to the people affected by its decisions.

A system need not hate anyone to cause harm. It can reproduce a flawed judgement at scale, make an opaque decision difficult to contest, or give an already powerful institution a faster way to act on its preferences. In such cases, the system’s apparent autonomy can obscure the choices that shaped it: whose objectives were built into it, what evidence it was allowed to use, where it was deployed and who was expected to absorb the risk. The pressing moral problem is not a machine’s imagined grievance. It is the human decision to build, trust or deploy a system without taking responsibility for its consequences.

This is not a new problem in technological dress. Human beings have long mistaken an increase in power for an increase in wisdom. We devise instruments that extend our reach, then speak as if the instrument itself had made the decision. But greater capacity does not dissolve responsibility. If anything, the ability to affect more lives should demand more stringent restraint and clearer accountability.

The warmonger in the prayer embodies the most naked version of power without responsibility. Political authority becomes a licence to command, dominate and gamble with lives that belong to other people. Greed may be part of the motive; so may ambition, fear, grievance or the intoxicating belief that one’s cause excuses any cost. Whatever the explanation, those who choose conflict are rarely the only ones who suffer it. Civilians lose homes, families and futures in contests they did not choose.

Extreme weather poses a different moral challenge. A storm has no intention; a climate has no conscience. Yet the risks people face are shaped not only by physical forces but by human decisions about energy, land, infrastructure and whose safety deserves investment. Responsibility is distributed across societies and generations, but the consequences are not evenly distributed. Those least able to shape the decisions may be among the first to endure the damage. Calling the result a natural disaster can describe the event while leaving its human causes and unequal burdens unexamined.

The three dangers in the prayer are therefore not identical, and they should not be collapsed into one. War is an act of human choice. Extreme weather is a physical event whose effects are intensified or mitigated by choices. Artificial intelligence is a set of human-made systems that can carry those choices into new domains and at greater speed. What joins them is not a common cause but a common ethical question: who has the power to decide, who benefits, and who is left to live with the consequences?

That question also explains why intelligence is not wisdom. Intelligence can calculate, predict, persuade and optimise. Wisdom asks whether the goal is worthy, whether the means are just and whether the people affected have any recourse. A system may give a precise answer to the wrong question. A leader may achieve a stated objective while making its human price invisible. An economy may record gains while leaving communities to pay for harms that were never entered on the balance sheet.

The danger, then, is not that artificial intelligence will become a person. It is that people may use its apparent neutrality to make decisions seem inevitable, technical or beyond challenge. “The system decided” can become a convenient sentence for those who designed the system, authorised its use or profited from its operation. The language of automation can turn a chain of human choices into an alibi.

Prayer cannot replace the institutions and public scrutiny that make power answerable. But it can return attention to what abstractions conceal. “Security”, “growth” and “innovation” may name real goods; they do not confer a right to treat other people as collateral. Nor should anxiety about hypothetical machine rebellion distract us from human beings already exercising consequential power over one another and the living world.

To ask for wisdom is to ask for more than cleverness. It is to ask those who hold power to explain their decisions, accept limits and answer to the people who bear the risks. It is to insist that a tool’s sophistication does not absolve its maker, and that a policy’s promise does not erase its victims. Above all, it is to refuse the comforting fiction that consequences belong to somebody else.

Perhaps the most unsettling possibility is not that machines will become more like us. It is that we will keep building more powerful machines while failing to become more responsible in our use of power. The prayer asks God to save all earthlings. Its human challenge is more immediate: can we learn to protect one another before our instruments make our irresponsibility more efficient?