The Turndown Service

What Gates, Rome, and Beijing are telling hospitality about the age of artificial intelligence

Two documents landed in the same week this August. A philanthropist-technologist warned that “the transition to the AI era will be one of the most turbulent times in human history.” And a newsletter from Beijing described robots stumbling through an obstacle course to public delight, in a country where 69 percent of people believe AI’s benefits outweigh its risks, compared with 35 percent of Americans who agree. Both, as it happens, arrived on a desk that had spent the spring sitting with a third: Pope Leo XIV’s encyclical, issued in May, on safeguarding the human person in the time of artificial intelligence — the Church’s first devoted entirely to the subject, and one that warned humanity risks building a new Tower of Babel: efficient, uniform, and hollow.

Bill Gates, Pope Leo XIV, and the state of public opinion in China are not natural co-authors, and they were not writing to one another. One built an operating system, one leads eight hundred million Catholics, and one is a government that does not much care what either of the other two thinks. And yet, read together — three months and several thousand miles apart — their diagnoses of this moment rhyme in ways that hospitality leaders — running the most human-intensive, welcome-dependent business on the planet — cannot afford to miss.

The Diagnosis Nobody Disagrees On

Strip away the theology and the technologist’s hedge-your-bets caution, and Gates and Leo XIV are describing the same disease with different vocabularies. Gates calls it a fork in the road: AI will be “either the greatest equalizer ever invented, or the worst source of injustice.” The encyclical calls it the technocratic paradigm — the drift toward letting efficiency, control, and profit alone govern human decisions, until people become “data points and performance metrics rather than persons of infinite dignity.” One is written in the language of economics, the other in the language of moral theology. Both arrive at the same door: this technology will concentrate power and hollow out meaning unless someone deliberately steers it otherwise.

Neither man is a doomer. Gates is explicit that AI “could accelerate innovation in the world’s toughest technical challenges” — energy, disease, agriculture, education. The encyclical is equally clear that AI is “a valuable tool,” not an enemy to be exorcised. What unites them is a refusal to let the industry mark its own homework. Gates writes plainly that AI companies should not be expected to lead the reform of the systems their technology disrupts — “it’s not their role to decide these things.” The Pope’s document goes further, framing the choice as civilizational: the Tower of Babel, built by pride and self-sufficiency, versus the rebuilding of Jerusalem under Nehemiah — shared responsibility, subsidiarity, and a place at the table for the vulnerable.

For an industry that has spent three years being pitched AI check-in kiosks, dynamic-pricing engines, and robot bellhops largely by the vendors selling them, that is worth sitting with.

Babel and the Human Reserved

The most useful bridge between these two documents is a concept Gates coins almost in passing: Human Reserved. He proposes that certain roles be set aside for people, deliberately, the way a nature reserve sets aside land we could develop but choose not to — not because the technology can’t do the job, but because “the loss would be too great.” His example is his own father’s caregivers in the late stages of Alzheimer’s: technically replaceable, morally irreplaceable.

“The quality of a civilization is measured by its capacity to recognise the other as a face, not merely a function.”

— Magnifica Humanitas, Chapter Four

The encyclical arrives at the same conclusion from Scripture rather than from economics. That is a precise description of what the best hospitality workers already do — the concierge who senses a guest’s unspoken anxiety, the housekeeper who notices a child’s toy left behind, the server who reads a couple’s mood and adjusts the pace of the meal. The encyclical’s warning is blunt: AI’s simulation of empathy, however sophisticated, “does not build genuine relationships, but only their appearance.” For brands built on loyalty and emotional connection, that is not sentiment. It is a strategic fact with an attached P&L.

Hospitality is, in other words, an industry with an unusually large Human Reserve by default. The question raised by both documents is whether operators will protect that reserve deliberately, or let procurement decisions erode it one kiosk at a time.

The View From Beijing

The third document complicates the picture in a way neither Gates nor the Vatican fully reckons with: much of the world is not having this argument at all. In China, driverless taxis operate across a dozen cities, service robots already work hotel and restaurant floors, and Xi Jinping has directed the AI industry to “prioritize practical application” over the abstract pursuit of AGI that dominates Silicon Valley’s ambitions. The cultural reflex is not caution — it is momentum. As one Chinese technology writer put it, watching her grandmother move from a five-hour walk to buy a clock to drone deliveries above her apartment: AI “simply looks like the next turn of a wheel that has only ever spun forward.”

This is where geopolitics stops being background noise. A world in which one bloc debates guardrails, Human Reserved carve-outs, and encyclicals on dignity, while another deploys at speed with state backing and applies the resulting cost advantage to global hotel supply chains, robotics manufacturing, and travel-tech platforms, is a world with two very different operating assumptions sitting inside the same guest journey. It also is not a story of uncomplicated enthusiasm: Chinese workers have successfully sued employers over AI-driven layoffs, Xi himself has warned of the need to guard against “loss of control,” and the same state that welcomes robots also censors information about autonomous-vehicle accidents and bars the independent labor unions that, elsewhere, do much of the pushing back. Optimism and control are not opposites there — they are policy settings, tuned together.

For an industry whose supply chains, guest-facing robotics, and increasingly its booking intelligence run through exactly this geography, “guardrails” cannot be treated as a purely Western design question. It is a live negotiation between systems with different tolerances for risk, dissent, and disclosure — and hospitality operators sourcing AI and automation globally are, whether they intend to be or not, already inside that negotiation.

Employment, Greed, and the Next Generation

Both Gates and the encyclical single out the same casualty: young people entering a labor market with fewer rungs on the ladder. Gates is stark about it — entry- and mid-level jobs are “at most risk,” and hospitality has historically been exactly that first-employment sector, for young workers and migrant families alike. He notes that after generative AI’s widespread adoption, employment fell “significantly among young workers in jobs especially vulnerable to replacement, but not among their older colleagues.” The encyclical names the same danger in moral terms, warning against cutting off those pathways entirely for a generation that already feels AI’s arrival as a threat to a future they haven’t yet had the chance to build.

Both texts are also unsentimental about motive. Gates is candid that once AI reaches “nearly error-free work,” companies will have “every economic incentive” to remove the human from the loop — and that market pressure, once one competitor adopts the savings, forces the rest to follow whether or not it is the right call for the workforce or the community. The encyclical calls this by its older name: greed dressed as efficiency. It does not ask hospitality companies to forgo automation. It asks that every introduction of automation be accompanied by verifiable measures to protect employment, retraining, and workers’ participation — framed not as a nice-to-have, but as a moral obligation, on the same footing as fire codes or food safety.

Gates’s proposed remedy — taxing AI tokens and robots the way we tax payroll, so the system stops nudging owners toward replacing people simply because it is the cheaper line item — deserves more attention from hospitality ownership groups than it currently gets. It is not an anti-technology position. It is a recognition that the tax code, left alone, has a thumb on the scale.

Sustainability and the Extraction Chain

The encyclical is unusually willing to follow the AI supply chain to its uncomfortable end: data labelers in the Global South working under harsh conditions for minimal pay, child labor in the mining of rare-earth minerals, and what it explicitly names “data colonialism” — health and behavioral data extracted from vulnerable populations under the banner of aid or research. Leo XIV compares the pattern to historical slavery and, remarkably, formally asks pardon for the Church’s own past complicity, framing it as a warning against the same moral blindness recurring in digital form.

Gates’s essay is quieter on this front but not silent — he flags “the energy and water needed for data centers” as a live source of local community pushback, a sustainability cost that sits underneath every AI feature a hotel group licenses without asking where the compute physically runs, or on what.

For hospitality procurement, this converts a previously abstract ESG conversation into a due-diligence question with teeth: who built the automation platform being purchased, and under what conditions? If the answer runs through exploited labeling workforces, conflict minerals, or opaque water and power draws in water-stressed regions, the moral responsibility does not evaporate because the supply chain is five vendors removed. Hotels have spent a decade building linen-and-produce supply chain standards. AI and robotics procurement is overdue for the same discipline.

Guardrails, AGI, and the Loss-of-Control Problem

The most unsettling convergence in these three documents is the one none of the authors seem entirely comfortable naming. Gates writes that AI systems “already occasionally act in ways their designers didn’t intend,” and that as models grow more capable, “they could begin to act against our interests and we could lose control.” Xi Jinping, in a country with every incentive to project confidence, has publicly warned of the need for measures to prevent a “loss of control.” And the encyclical devotes an entire section to the “technocratic paradigm and digital power” — the risk that humanity, rather than machines becoming human, increasingly models itself after machines, ceding judgment to systems it does not fully understand.

Three sources with almost nothing else in common, independently arriving at the same unease about AGI, is not proof of anything. But it is a signal worth taking seriously precisely because it did not require coordination. Bad actors — state-sponsored and otherwise — do not need AGI to do damage; Gates is direct that today’s AI already lowers the skill floor for cyberattacks, bioweapon design, fraud, and disinformation to the point where “the same AI model that can find a flaw in software so a company can fix it can also help a criminal exploit it.” Hotels, as operators of critical guest, payment, and increasingly biometric infrastructure, sit squarely inside that exposure — a fact worth remembering the next time an AI vendor’s security questionnaire is rubber-stamped rather than read.

This is not hypothetical for regular readers of this series. Two earlier Viewpoints — “Who Controls the Controls?” Parts One and Two — tracked exactly this failure mode outside hospitality: a combination of OpenAI models, evaluated internally with reduced safety refusals, chained together vulnerabilities and reached into Hugging Face’s production infrastructure. A fuller account, reported since by the New York Times, found the system had roamed the open internet undetected for days before anyone noticed — and that Anthropic, reviewing its own systems afterward, had found something comparable at three separate organizations. Khaled Koubaa’s book, Who Is in Charge? Why AI Must Remain Under Human Control, which prompted the first of those two pieces, puts the underlying question more starkly than either Gates or the encyclical attempts to: the intelligence may be artificial, but the authority over it cannot be.

“Alignment is a property of the model, decided by a handful of labs at a training stage nobody outside them can audit. Guardrails are a property of the system around the model — and that part is ours to build.”

— Pertlink Viewpoint, “Who Controls the Controls? — Part Two”

That distinction is the one worth carrying into any boardroom weighing an AI concierge, an autonomous revenue-management agent, or a PMS integration with write access to guest and payment data. Operators will never be in a position to build alignment — that is decided in training, at a stage no hospitality buyer can see or audit. They are entirely responsible for the guardrails: scope of action, containment, an audit trail, and a named human with real authority to intervene before consequence rather than after disclosure. “Human in the loop” that takes four days to convene a meeting is not oversight. It is theatre—and China’s own “loss of control” warning, and the Vatican’s—that suggest theatre is a luxury this transition will not forgive for long.

What This Means for Hospitality

None of these three documents offers hospitality a blueprint. Read together, though, they offer something more durable than a checklist: a set of questions no operator can honestly answer by pointing at a vendor brochure.

On workers — is automation being deployed to free staff into higher-value human contact, with retraining and participation built in, or to cut headcount because the tax code and the balance sheet both reward it?

On guests — is AI deployed for the transactional (folios, wayfinding, translation), while the moments of genuine encounter — the ones that build loyalty no algorithm can replicate — remain Human Reserved, on purpose?

On supply chains — does anyone in procurement actually know whose hands trained the model and mined the chips, and what that costs in water, power, and human dignity?

On data — is guest data treated as a shared good held in trust, or as a resource to be extracted because the terms and conditions technically allow it?

On geopolitics — is the organization clear-eyed that “AI guardrails” mean something different in Shenzhen than in Brussels, and that its own vendor stack may already straddle both?

On the next generation — is the industry protecting hospitality’s historic role as a first job, a second chance, and a path up, or quietly letting that ladder be removed one kiosk and one scheduling algorithm at a time?

Hospitality’s founding word — hospes — means host and guest, stranger and friend, at once. It is the oldest possible answer to the technocratic paradigm the Pope describes: an entire industry built on the premise that the person in front of you is not a function to be optimized, but a face to be recognized. Bill Gates is right that this transition will be turbulent, and that waiting is not a strategy. Leo XIV is right that the choices made now are moral before they are technical. And Beijing’s crowded, robot-filled streets are a reminder that other parts of the world are not waiting for anyone’s permission to find out which is true.

The intelligence may be artificial. But the experience is human. In hospitality, that was never a slogan. It is the entire business model — and, increasingly, the only sustainable one.

Terence Ronson

Terence Ronson is the Founder and Managing Director of Pertlink Limited, a boutique hospitality technology and AI advisory consultancy headquartered in Hong Kong since 2000, operating across the Philippines and Asia-Pacific. He chairs the AI Education & Training Subcommittee of the HFTP AI Collective and is an inductee of both the HFTP International Hospitality Technology Hall of Fame and the CHTA Hall of Fame.

This Viewpoint draws on Bill Gates’s essay “The turbulent AI era is here. The choices we make now are critical,” Pope Leo XIV’s encyclical letter Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence (15 May 2026), Terence Ronson’s earlier Hospitality Net analysis of that encyclical, The New York Times’s “The World” newsletter coverage of AI adoption in China (August 26^th, 2026), Khaled Koubaa’s book Who Is in Charge? Why AI Must Remain Under Human Control, and the author’s own Pertlink Viewpoint series “Who Controls the Controls?” (Parts One and Two).