Agentic AI, Meatspace, and Martin Luther – Religion & Liberty Online

Stop hiring humans. They get sick, have children, and occasionally observe religious holidays. Bots, on the other hand, never waste time on these sorts of things. Bots don’t take personal days or holidays, and they never have children with runny noses. Your next best hire will be an AI agent that works around the clock researching market segments, generating new leads, and building relationships with customers—all for a fraction of the cost of a human.

As dystopian as this may sound, what you just read is neither satire nor science fiction. Agentic AI companies like Artisan are making it possible for employers to stop hiring humans. Artisan, and similar companies offering bots for hire, would like to replace a human workforce with an AI workforce. And, according to the Artisan website, some companies are very satisfied with the results. The Artisan website quotes one CEO as saying, “We’ve replaced our entire outbound sales team with Artisan. The open rate is twice what it was with humans, and they close $100K deals day after day.”

The prospect of hiring bots over humans is made possible by agentic AI technology. Building on the power of large language models and generative AI, agentic AI surpasses these in its autonomous capabilities. Unlike previous forms of AI that require a human prompting each step of the way, agentic AI is capable of independent reasoning and autonomous work. For example, a human can create an AI agent and assign it a complex, multistep project such as generating new customer leads through outbound outreach. The agent will then get to work researching potential customers, finding contact information, composing personalized emails, and communicating with potential clients. The agent will tirelessly work until it is told to stop—no sick days, no holidays, no sleep.

Now not everyone agrees that we should stop hiring humans. Artisan’s ad campaign prompted global outrage so that the company had to explain their rationale and defend their efforts. Others, however, have seen agentic AI as an opportunity to create multi-million-dollar businesses without human employees. Meanwhile, some have used this new technology for entrepreneurial opportunities. The company RentAHuman.ai helps AI agents hire humans to do things that bots cannot do. If an AI agent needs a human to hold a sign at a busy intersection or count the number of customers entering a local coffee shop, RentAHuman.ai can help make that happen. The company claims to have over 750,000 rentable humans available for hire in its so-called meatspace.

This is all very weird, but not entirely new. The call to stop hiring humans has happened before. There is a long history of trying to replace humans with machines to get more done through time-saving devices. While we’ve never been exactly here before—bots hiring humans in the meatspace—there are historical antecedents to our present situation. History can illuminate the opaque future of agentic AI. If it is at all possible for agentic AI to bring about flourishing for the common good, then we must heed some key insights from the history of automation. 

Carl Benedikt Frey in his book The Technology Trap: Capital, Labor, and Power in the Age of Automation helps us understand how our present-day situation is not entirely new. Frey, an economist at the University of Oxford, uses historical examples to illuminate the difference between labor-replacing technologies and labor-enabling technologies. While a labor-replacing technology rapidly replaces large numbers of human workers, labor-enabling technologies augment the efforts of human workers while steadily transforming those workers and industries.

For example, Frey suggests that Industrial Revolution power looms were labor-replacing technologies. A single power loom machine could replace a large number of handweavers and skilled artisans. Power loom machinery—along with the help of a few child laborers—could churn out more and cheaper textiles than ever before. Naturally, the power loom was despised by the displaced handweavers and skilled artisans since it left them unemployed and without new opportunities to reskill or adapt to a changing industry.

Though the power loom brought about massive upheaval in the short run (unemployment, inactivity, social discontent), it eventually brought massive gains in the long run (cheaper cost of goods, new innovation, eventual wage growth). There was just one big problem: Those gains came after the displaced workers were dead and gone. It was not the unemployed weavers who experienced the gains of this new technology—it was the grandchildren and great-grandchildren of destitute weavers who benefitted from the power loom.

Unlike the labor-replacing technology of the power loom, Frey offers up the printing press as an example of a labor-enabling technology. Although the mechanical printing press disrupted the work of scribes, it ultimately allowed these workers to expand their efforts. The printing press enabled more books to be published in 50 years than had been published in the entire preceding millennium. Scribes working in the scriptoria began using the mechanical printing press to augment their work as they steadily learned new skills as book binders and printers.

While the power loom incited riots, the printing press did not. Why is this? According to Uwe Neddermeyer, in his article “Why Were There No Riots of the Scribes?,” the short-term work and income of scribes was largely unaffected by the printing press. Many professional scribes continued getting paid for small-scale projects that were impractical for mass printing. This means that the short-term disruption of the printing press was fairly minimal, whereas the long-term impact of this technology brought about a new industry and many new jobs. Scribes found continued work in correcting, binding, and selling printed books while monasteries steadily turned their scriptoria into book binderies and printing presses.

These historical examples can help us make sense of agentic AI today. Agentic AI that contributes to the common good must bring both short-run and long-run benefits. Rather than seeing AI agents as a way to “stop hiring humans,” these tools could help extend the work of humans. Pope Leo XIV captures the urgency of this moment in his recent Magnifica Humanitas encyclical when he wrote:

Today, the human desire for fullness of life is at risk of being misled by deceitful goals, such as the prospect of a technology that promises to free us from all weakness, and models of wellbeing that leave behind entire populations. All too often, we place our hope in unlimited “upgrades,” in forms of progress that exacerbate inequalities, and in immediate solutions incapable of healing people’s wounds.

If agentic AI replaces entire departments overnight (the “immediate solutions” that Leo warns about), people will undoubtedly rage against this labor-replacing technology. If agentic AI promises flourishing in the distant future at the cost of human obsolescence in the present, then we will be repeating the errors of the Industrial Revolution. Historically speaking, agentic AI ought to follow the path of the printing press, not the power loom. While the former augmented human workers and innovation, the latter rapidly replaced people and prompted arson. Efforts to ‘stop hiring humans’ and start hiring bots could lead to something like the power loom riots of the early nineteenth century. 

History also has something else to say about agentic AI: It may not be as helpful as we’d think. Many proponents of this new technology claim that it will provide incredible time-saving possibilities. The promise of agentic AI is that bots will perform dull and monotonous tasks so that humans are freed up for more creative and meaningful work. Proponents of agentic AI see it as a way to limit our weekly working hours—AI agents will save time and effort so that a four-hour workday can accomplish what used to be done in an eight-hour workday.

Will agentic AI really save time and provide leisure? The history of technology tempers our expectations in this regard. Ruth Schwartz Cowan in her 1983 book More Work for Mother: The Ironies of Household Technology from the Open Hearth to the Microwave has something to say here. Cowan traces the history of domestic technologies such as the washing machine, dishwasher, and stove. Her work reveals how these so-called time-saving devices actually led to more work, not less. Rather than freeing up time and introducing simplicity for people in general (and for women in particular), these devices created more work and made domestic life significantly more complicated.

According to Cowan, innovations in household technology never delivered on their promise to save time. While a meal over an open fire would have previously been cooked in a single pot, the advent of the electric stove raised culinary expectations so that a single meal then involved multiple pots and pans. Simple breads traditionally served in households were supplanted by far more complicated breads made possible with new kitchen devices. Services that once came to the home—doctors and seamstresses as well as deliveries of milk, ice, coal, groceries—went away as more and more families owned a car and could pick up the items themselves. Devices that supposedly saved time resulted in much more work and far less time.

These insights shine a light on the future of agentic AI. Cowan demonstrates how increased complexity and shifting expectations make saving time a chimera of contemporary life. The goal posts keep getting moved further and further away from us. While agentic AI will ostensibly save time and provide for more leisure, history suggests otherwise. Expectations for output and production will scale accordingly, and these agents will fail to bring about a restful utopia.

In fact, we may even find that agentic AI makes life busier and more complicated. If historical patterns remain the same, the invention of new devices will make our lives more complex and busier than ever. Current research on how AI impacts daily work suggests that this technology is making our work harder, not easier.

Not only does history have something to say about the future of agentic AI but so does theology. The Christian faith, especially the doctrine of vocation, offers important insights for how agentic AI might bring about flourishing for the common good. The doctrine of vocation—God’s calling upon individuals to meaningful work in the world for the benefit of others—can help reframe conversations about agentic AI. Without the insights of theology, conversations around this new technology will be myopic and misguided.

Rather than seeing this new technology as a way to stop hiring humans, vocation helps us see how this technology might be aimed at loving and serving our neighbors. The doctrine of vocation broadens our horizons and resists reductive criteria for evaluating agentic AI. Rather than evaluating this tool on the basis of how it serves money, time, output or leisure, vocation invites us to consider how this tool might serve the neighbor.

Martin Luther in his treatise “Temporal Authority” offers a helpful criterion for how we might use agentic AI for flourishing and the common good. Luther, addressing princes and nobles in his day, asks them to consider how their decisions benefit the neighbors working for and with them:

Behold, Christ, the supreme ruler, came to serve me; he did not seek to gain power, estate, and honor from me, but considered only my need, and directed all things to the end that I should gain power, estate, and honor from him and through him. I will do likewise, seeking from my subjects not my own advantage but theirs. I will use my office to serve and protect them, listen to their problems and defend them, and govern to the sole end that they, not I, may benefit from my rule.

Luther’s words have something to say in our age of agentic AI. Rather than bringing this technology into the world without any regard for the short-term harms inflicted on human workers, we ought to consider how agentic AI could benefit our neighbors and entire populations. The doctrine of vocation expands our horizon beyond the benefits of the bottom line, helping us see our human neighbors and communities alongside efforts to improve output and efficiency.

While agentic AI will change entire industries and the future of work, God’s plan for human work that is aimed at loving and serving our neighbors will remain. Agentic AI is bringing about new and difficult questions—we need the wisdom of history and the neighbor-centered love of vocation now more than ever.