THE MACHINE AND THE CARETAKER

Artificial intelligence, work and responsibility over time of generational transitions

In a city in southern China, in Guangdong, a hotel is preparing to open its doors without any human staff: at the reception there will be humanoid robots, luggage will go up to the rooms on silent wheels, dinner will come from a kitchen managed entirely by algorithms and interactions with other robots. No, it’s not Isaac Asimov and not even Robert A. Heinlein, it’s Reuters[1], and the news comes to us like a postcard sent from the future, but it’s about the present.

Another scene, a small accountant’s office in a small town, maybe Italian, it doesn’t matter: a bookkeeper who has been entering invoices for twenty years looks up from the screen. He is doing the same job as yesterday, but he knows that that job will soon no longer be needed.

The same technology that promises to free us from repetitive fatigue threatens to leave those who made that effort on the ground.

Unexpectedly, on this thin ridge of thought, a very loud voice was raised: Pope Leo XIV in his recent encyclical[2] on artificial intelligence brought the issue back to its center of gravity — man.

Even more surprising was listening, in tune with that call, to the voice of Christopher Olah, co-founder of Anthropic, one of the very builders of these systems: a sign that, when the stakes are high, the pastor and the engineer can look in the same direction.

This newsletter was born from that amazement.

We do not want to offer solutions to problems on a planetary scale, of course. Rather, reflect on the fact that artificial intelligence is indeed an extraordinary opportunity for human progress, but that its power multiplies the responsibility of those who govern it.

• A new social issue

From Rerum Novarum to “Magnifica Humanitas”: history repeats itself at a different pace

When, in 1891, Leo XIII published Rerum Novarum, the steam engine had already redesigned the world: it had filled the cities, emptied the countryside, created a working class and, with it, a “social question” that the Church did not want to leave to the market alone or to the revolution alone. More than a century later, another Leo — XIV — grasps the new social question in Artificial Intelligence.

And it is no coincidence that he chose, for his first major encyclical, a name that evokes that tradition.

The difference lies in the rhythm: then the transformation took two generations to complete, today it risks being completed in one decade, perhaps less.

The time to think has become shorter, but the stakes have never been higher: not only the dignity of those who work, but the very survival of humanity.

“A tool of human genius”: AI has no soul, it has managers

The magisterium of Leo XIV is clear on a point that has precise juridical consequences: Artificial Intelligence is a tool, “the product of human genius,” not a subject endowed with rights or its own will. Its ethical strength , the Pope warns , derives entirely from the intentions of those who design it, those who buy it and those who use it.

With a few words, the Pontiff dismantles the most insidious alibi of our time: that of hiding behind the algorithm. The machine does not decide: someone always decides, and that someone has a name, a face and a responsibility.

Just as the CEO is responsible for the choices of a company, so those who introduce AI into their work are responsible for how they use it.

Disarming technique: power, concentration and the risk of a few hands

“Disarming” artificial intelligence….. the expression used by Pope Leo XIV is striking.

It emphasizes the need to remove AI from the logic of domination, manipulation and control to orient it towards cooperation and communion between people.

Otherwise, the danger is real: a technology whose power is already difficult to conceive today, if controlled by “a handful of companies”, could lead at best to an information imbalance to the detriment of the weakest and… at worst… (dots are intentional)

A concern that, surprisingly and fortunately, is also shared by some of the major players in the technological frontier.

The thread that unites the Pope and the frontier engineers

That a pontiff and a co-founder of one of the large AI companies can be alarmed for the same reasons should make us think seriously.

It should make us think seriously today.

And spur us to act.

If the same subjects who implement these systems speak openly of the risk of systemic “catastrophes” if development is not governed with seriousness and adequate standards, the consequence is both very simple and very demanding: we are not spectators.

We are the supply chain leader.

We are the ones who choose suppliers and technologies, who establish which activities to entrust to the machine and which to guard for the hand and judgment of man, who design the processes in which the algorithm enters the daily work. And we are always the ones who decide where and how AI will touch people’s real lives: that of our employees, whose professional future we guide, and that of customers, whose data, stories and trust we process. In each of these choices — apparently technical, but in reality profoundly human — our responsibility is exercised or abdicated.

• Discerning, accompanying, empowering: the compass for those who guide

Discerning: where AI creates value and where it empties work of meaning

The first virtue, in times of technological abundance, is discernment[3].

Not everything that can be automated deserves to be automated, and not everything that remains human is inefficient for that very reason.

Discernment means distinguishing, case by case, where the machine frees up precious time — that subtracted from repetition — and where instead it risks emptying work of its meaning, reducing the person to a simple overseer of a process that he no longer understands.

“You were not made to live like brute beasts, but to pursure virtue and knowledge”[4] If for Dante discernment was not the talent of animals, even less so can it be of an algorithm.

Accompanying: retraining as a duty, not as a courtesy

Now, AI frees up time and margins, there is no doubt about it.  The ethical question then becomes: how do we reinvest those margins in people?

In the light of Leo XIV’s teaching, accompanying workers in their professional transitions is not a gesture of optional generosity, but a real duty.

It is not enough to “do no damage”: it is necessary to actively take charge of the transition, to build paths, to give a horizon. Keeping a person to carry out a task that is no longer useful is not pity, it is illusion; abandoning it is injustice.

The third way — following virtue and knowledge — is requalification.

Empower: behind every algorithm there must be a face

The third verb, to empower, closes the circle. In all the choices that affect people — hiring, firing, evaluations, access to credit, essential services — it is important that a human, responsible and reachable decision-maker always remains recognizable.

AI can instruct the file, not sign the sentence: maintaining a human decision-making track, especially in the most sensitive matters, is a guarantee that responsibility does not evaporate in the computer code.

• Three stories from the changing world

The accounting firm: from data entry to interpreting numbers

Let’s start from us: Artificial Intelligence, and before that computer science, have greatly reduced the time of pure data processing. The oldest of those who read me will remember the hand-kept accounting. However, the fact that today you can take a package of invoices, scan them and obtain the related accounting records in a few minutes does not mean, unfortunately and absurdly, that managing the accounting of a company has become easier or faster.

The opposite is true.

And entrepreneurs are as aware of this as accountants.

The obligations required by the tax system have multiplied dramatically, and with them the responsibility that weighs on companies and the responsibility for compliance that weighs on the accountants and consultants: deadlines, communications, cross-checks, compliance, privacy, security, constantly changing and expanding obligations.

Never has it been so important, even with all the automations and simplifications, to correctly follow the accounting and administration of a company requires so much work, so much competence and so much attention.

The “job” has not lightened: the center of gravity has shifted, from material typing to the governance of unprecedented regulatory complexity.

From data to meaning: a new value for the customer

And this is where the most interesting opportunity opens up, for accounting firms and for those who rely on them. The energies that technology frees from pure registration do not translate into less work, but into higher quality work, which can finally be dedicated to companies: analyzing numbers instead of just producing them, preventing problems instead of chasing them, offering vision and  governance tools.

In practice, this means that that bookkeeper who has entered invoices for years is becoming, with a gradual path of growth, the interlocutor who prepares monthly dashboards, catches a cash anomaly in advance, sits at the entrepreneur’s table bringing interpretations and proposals.

The law and tax firm: young people are not labour, they are the directors of tomorrow

The second case is more delicate, because it affects those who are entering the profession. In legal research, in the drafting of the first drafts of opinions, contracts and pleadings, generative AI compresses in a few minutes what yesterday occupied young people for days. Thus, the old “ladder” model, for which people grew up doing basic work for years, collapsed. The temptation — convenient and short-sighted — would be to do without practitioners. But burning a generation of young professionals because the machine saves them would mean using AI against the dignity of people and, at the same time, against the very future of professions.

From “jurisprudence seeking” to the governance of the flow

The way is not substitution, but promotion. If until yesterday the trainee was asked to “look for the jurisprudence” and “make the first draft”, tomorrow he can be asked for something more qualified: to design and supervise the workflow with AI. Defining the perimeter of the research, critically validating the sources that the machine proposes, defending the client’s confidentiality, grasping distortions and hallucinations, adapting the text to the human context of the person in front of him. Less time on collection and formatting, more time on strategy and relationship.

In short, the young professionals of the Studio go from being manual workers to assistant directors.

The hidden risk: what happens to “knowing how to think”?

There is, however, a delicate aspect. In the old “ladder” model, the young man learned by doing: in the long hours of research and in the tiring drafting of drafts he learned to connect documents and information, to go deep into concepts, to mature ideas, to build solutions. And, by writing, you learned to write. If the machine takes over the research and the first draft, these skills will no longer be formed, they will remain the heritage of older colleagues for a few decades and then be lost.

It is a high risk and certainly not limited to professional firms.

An attempt at an answer is before your eyes: what you are reading has been written by AI.

However.

However, the human being has reread, corrected, changed a word, created emphasis, eliminated a naivety, fixed a sentence, added a concept.

The gym still exists, only some equipment has changed.

This could be one of the most difficult challenges: avoiding settling for the AI-generated product but using your imagination, experience, sensitivity to adapt and improve it.

It is an open issue. We are sailing through uncharted waters….

The robot hotel: when the relationship becomes the rare commodity

Let’s go back to the postcard from the future. China’s robot-operated hotels — from Alibaba’s pioneering FlyZoo to the new, fully automated project announced in Guangdong — are not an exercise in style: they are a laboratory. Reception, luggage transport, cleaning, catering, all entrusted to robots coordinated by a central system. If an entire hotel can run almost unstaffed, the impact on occupancy can be huge. But it is precisely here that the reasoning becomes broader: because in this scenario the responsibility no longer lies only with the entrepreneur who retrains his own, but also — and above all — with governments, called upon to active policies, training and safety nets up to the time.

What no machine returns

There remains a question that efficiency cannot ask: what does the customer lose when no one welcomes him? What does the territory lose when hospitality ceases to be an encounter between people? The robot hotel questions us today about what we want to continue to consider precious: the look, the care, the word spoken at the right time. These are goods that no machine, no matter how sophisticated, can return — and which, paradoxically, automation makes rarer and therefore more valuable. In short, even in the hotel, the future of human work does not lie in competing with the robot on speed, but in presiding over what the robot remains precluded: the relationship.

The three cases compared: from today’s role to tomorrow’s job

• What we can do

Even before any technicality — which we will address in detail in one of our next newsletters— it is worth setting the direction. There are four construction sites, which we will only mention here.

Mapping, forming, measuring

The starting point is an honest look at one’s own consulting firm or company: which tasks will be absorbed by automation, which skills are transferable, with which paths and in what timeframe. A minimal methodology — mapping, training, measuring — is worth a thousand good intentions.

The tools available

No one is called to do everything alone. There are already interprofessional funds, vouchers for continuous training and professional communities that offer concrete support for retraining. Knowing and using them is part of the responsibility of the driver.

An “AI policy” of the firm or company

Dignity, transparency, proportionality, accountability: four principles can become the backbone of an internal policy on the use of AI — which tells, in black and white, what data is processed, which decisions are never delegated to the machine, who is responsible for them.

The generational transition as an opportunity

For family businesses, the change between generations is the right time to “recalibrate” the use of AI in terms of the common good. Family pacts, statutes, group regulations can already include clear principles on the group’s digital strategy.

We will dedicate some of our next newsletters to these construction sites — with mapping, operational tools, useful references and a draft policy.

• The part that doesn’t depend on us (but we can lend a hand)

The responsibility of governments

It would be naïve to think that the individual entrepreneur can bear the weight of an epochal transformation alone. The reconversion of entire sectors — think of the hotel front-office or repetitive services — calls into question governments: active labour policies, funded training, shock absorbers that accompany transitions without breaking lives. It is a long, difficult, but inevitable process: and if well governed, it can lead not to impoverishment, but to a leap in the quality of work.

The responsibility of those who build AI

At the other end of the supply chain are the manufacturers. The greatest responsibility weighs on them, because an error of direction, on the scale of these systems, can have planetary consequences. The fact that it is the developers themselves who evoke the risk of “catastrophes” is, at the same time, an alarm and a hope: it means that awareness exists, and that the right direction — made up of prudence, transparency and rules — can transform the threat into unprecedented progress.

Our part

And then there is our part. We do not govern the great models, we do not write national laws. But we decide, every day, how AI enters our firm and our company: which tools to choose, which processes to redesign, which people to accompany and to where. It is a small part on the scale of the world, but decisive on the scale of the people around us.

And people, after all, are the only scale that really matters.

§ § § § § §

Do not hesitate to contact us for any further information.

Sincerely,

Aldo Filippini

This newsletter is the result of a collaboration between artificial intelligence and human expertise, with revision and editorial care by Aldo Filippini.

[1] https://www.reuters.com/article/technology/at-alibabas-futuristic-hotel-robots-deliver-towels-and-mix-cocktails-idUSKCN1PG210/

[2]  The complete text of the Encyclical Magnifica Humanitas can be downloaded at this link

[3]“The intellect which cannot clearly distinguish right from wrong, and moral duty from what is forbidden… that intellect is confused and in darkness” Bhagavad Gita (18.30) Already millennia ago the Gita recalled that the true intellect is that which knows how to distinguish ‘the right action from the wrong one, what must be done from what must be avoided’. This profound discernment does not arise from a predictive algorithm, but from human conscience, the only force capable of guiding the use of technology towards the common good.

[4] Dante Alighieri (Inferno, Canto XXVI, verses 119-120)