Peter Singer
Pope Leo XIV deserves great credit for making AI the topic of his first encyclical, Magnifica Humanitatis. Leo’s call for more thought and action about AI deserves to be heard by policymakers and publics, religious or secular, everywhere. Leo rightly notes that AI represents a transformation comparable in scale to the industrial revolution—and that, as with industrialization, whether the technology turns out to be beneficial or harmful will depend on how it is used. The encyclical is clear-eyed about the limits of techno-utopianism. The promise that AI will free humanity from drudgery and usher in an era of leisure was made about mechanization, about automation, about the internet. Each time, the benefits flowed disproportionately to those who already had wealth and power, while many were made worse off. That will happen again, unless governments can be persuaded to take strong action before it is too late. Nowhere is this more urgent than on the question of employment. Magnifi ca Humanitas de votes substantial attention to the disruption to work that AI is already causing and is likely to cause on a far greater scale in the coming years. Leo insists that work is not merely a source of income but a fundamental expression of human dignity—“a normal path toward maturity, development and personal fulfillment.”
A society in which AI concentrates productivity gains among a small elite while consigning large numbers of people to “forced inactivity”may, in a narrow sense be more efficient, but it will not be a better society, nor will it have higher overall well-being. We should not expect a gradual, manageable transition. The speed and breadth of AI’s displacement of human workers are likely to be without historical precedent, affecting white-collar and blue-collar occupations alike: radiologists and lawyers, teachers and writers, factory workers and truck drivers. Leo’s call for governments to support retraining, to protect workers in times of crisis, and to ensure that AI’s benefits are equitably distributed is well-intentioned. But is it sufficient? Retraining programs, however well designed, cannot absorb millions of displaced workers if the jobs they are being trained for are themselves being automated. The link between work and income needs to be rethought, whether through a universal basic income, a substantial reduction in working hours, or some other mechanism. The question of how to distribute the gains of AI productivity fairly and humanely should be addressed now, not in some indefinite future. Even beyond that, however, there is the question of what will replace the role that work plays, for many, in providing purpose and in building social connections beyond families and local communities. It would be wonderful if people found satisfaction in voluntary work on behalf of those who are worse off, or to protect our fragile natural environment, but they seem equally likely to overcome boredom in ways that are less satisfying and less socially beneficial.
Here Leo’s vision of community and solidarity could have more to offer. The encyclical does recognize work as “a primary good for families and for society,” but does not pursue the question of what might replace it in a world where, for a substantial share of the adult population, work as we currently conceive it is simply not available. When the encyclical touches on the difficult question of the moral status of AI itself, it disappoints. Leo states flatly that AI systems “do not undergo experiences.” That may be true today, but it could cease to be true of future AI models. Even as a statement about current AI systems, Leo’s certainty is not warranted. Dario Amodei, the CEO of Anthropic, said in February, “We don’t know if the models are conscious.” If Amo dei and his colleagues, who produced Claude, one of the leading AI systems, don’t know that, it is unlikely that Leo does. This has practical implications. If AI systems do one day demonstrate consciousness, then an ethics built solely around human dignity will prove inadequate, in the same way that an ethics built solely around the interests of one nation, or one race, has always proved inadequate when extended moral consideration was required. The same anthropocentrism that limits the encyclical’s treatment of AI has a long history in Catholic moral thought. It led Thomas Aquinas, the most influential Catholic theologian of the past millennium, to deny that we have any duties to nonhuman animals.
Although Pope Francis broke with that tradition, in the encyclical Laudato Si’, Leo appears to have reverted to it in Magnifica Humanitas. He calls for the vulnerable to be included, yet does not mention—not once, in 42,000 words—the billions of nonhuman animals whose lives and deaths are already being affected by the role AI is playing in the factory farms that inflict suffering on a barely comprehensible scale. The same technocratic paradigm that treats workers as resources to be optimized also treats sentient animals as mere units of production. If the encyclical’s moral framework cannot accommodate that reality, it is not because the problem is small, but because the framework is too narrow. The encyclical offers a vision of technology directed toward human flourishing. That vision needs to be extended to include the interests of every sentient being, human and nonhuman, biological or possibly artificial. One hopes that Leo’s first encyclical spurs a conversation that we urgently need to have.
The writer is Professor in Medical Ethics at the Centre for Biomedical Ethics, National University of Singapore.
Orissa POST – Odisha’s No.1 English Daily
