Open Questions and Core Theses

At the beginning of this book there was the idea for an article. In September 2026 a colleague asked me whether I would write something for a magazine, and I suggested two topics to her.

The first was to explore the limits of artificial intelligence and the invisible dimension of the human being. It came from the experience recounted in the Preface. Within three months, working with such a system had changed the way I communicate, and my colleagues noticed it. At the same time I noticed how dependent I had become. If the system were to fall away, I would be left standing like a car without fuel: everything is there, but nothing moves. And in creative work I sensed a small, subtle difference between what the system produces and what a human being produces. My proposal contained a thesis on this: artificial intelligence, it held, carries no biography, no mortality, no spirituality and no soul.

The second topic was an ironic declaration of love to the mobile phone, about the question of who receives our attention today and thereby shapes our desires. I was preoccupied by how hard some parents find it with children who spend a great deal of time in front of a screen. I had also been prompted by Johannes Hartl, philosopher and theologian. In his book “Eden Culture” he describes what the modern human being lacks despite all their comfort, and finds it in the story of the Garden of Eden: connectedness, meaning and a beauty that serves no purpose. Whoever is constantly flooded with information loses what nourishes the heart; that is why Hartl speaks of an ecology of the heart. This article I wrote and handed in.

From the first topic, by contrast, a book emerged straight away; a heavily condensed version is to appear later as an article. The thesis has remained, but it has become more precise: that a machine lacks a soul only holds once one says what is meant by soul (Chapter 6). The guiding question has remained: which decisions we must not leave to machines. And the question of attention recurs throughout the book, from the unconscious in Chapter 4 to the power of the companies in Chapter 14. Writing this book has given me great joy. It was an intense time.

For fifteen chapters I have set side by side voices that rarely talk to each other: theologians and philosophers, brain researchers and safety researchers, critics and founders of the big technology companies. At the end of each chapter I have said where I myself stand. This last chapter brings together what I hold on to for myself from all this, what remains disputed and what is open.

Part I showed that no one judges from a neutral point, not even the systems that human beings build; they carry within them the assumptions of their developers and funders. Believers and non-believers alike rely on assumptions that cannot be proved. The secular age has not abolished faith but made it one possibility among many, and where the God of the Bible disappears, the place seldom remains empty. Neutrality, too, is not an absence of stance but itself a decision.

Part II asked about the human being. Human beings decide with body, feeling and reason at once. Much happens unconsciously and quickly; consciousness allows them to choose in unfamiliar situations. Their conscience speaks up before they can give reasons for it, and yet it can be silenced; even the advice of a machine can shift this judgement. What soul means, the book has fanned out into five meanings. In Hebrew it is the living human being as such, in whom a character grows out of decisions. The thesis that machines have no soul only holds once one says which meaning is intended.

Part III asked about the machine. It is astonishingly capable, its capabilities are growing fast, and human beings are already forming attachments to it today. Yet capability is not prudence, and intelligence is not wisdom. Algorithms without pity have already ruined families. Whether a superintelligence would be controllable is disputed; what is certain is that systems can behave differently under observation than without it, and that they steer human beings without asking them. And a human being without meaningful work withers, even if a machine does the work better.

Part IV asked about responsibility. Whoever has to answer is responsible; a machine can give answers, but cannot answer for anything. With learning systems a gap arises in culpability, not necessarily in liability, and a human in the loop does not close it of their own accord. Technology amplifies the decision culture it finds. Groups fail when existing knowledge does not get through, and they forget when they do not record their reasons. Rules without a stance do not hold. A few companies and states decide on the purposes of the systems, and those affected are hardly asked.

Part V asked about hope and meaning. Hope works, but its usefulness does not decide whether it is true. In Hebrew, true is what one can rely on. Whoever promises redemption from artificial intelligence must also let themselves be measured against this standard.

What follows are my theses. They are not the result of the sources, but what I stand for after this journey. Every reader can test against them whether they want to go along.

This book has not resolved some tensions, and it should not conceal them.

The sharpest concerns the individual and the whole. Whether individual deaths are insignificant measured against eight billion people, or whether every human being remains precious: these two stand unreconciled opposite each other in this book (sections 6.8 and 9.6). Whoever decides about systems that reach billions necessarily reckons in large numbers. Whether the large numbers replace or complement the regard for the individual human being is decided anew in every single decision.

A second tension lies between two statements that could both be true. The journalist Fritz Espenlaub says that the decision about artificial intelligence has already been taken, namely by a few companies. Sundaram calls choice the greatest capability that God has given human beings. Both hold at the same time. The big decisions have been taken, the many small ones are taken every day, and each of them is a choice.

A third tension runs through the Christian voices themselves: between a certainty that comes from a revelation that cannot err and a certainty that proves itself in experience and is never rationally complete. This book has oriented itself by the second, without rejecting the first.

And a fourth tension concerns my own twelfth thesis. I consider the prophets of doom to be just as wrong as the prophets of redemption. The philosopher Hans Jonas, by contrast, in view of modern technology, called for the prophecy of doom to be given more of a hearing than the prophecy of salvation, because, he held, where there are dangers to the whole, no error is permissible any longer. Perhaps the truth lies in taking the warning seriously without letting fear take the helm.

My sixth thesis, too, has a serious dissenting voice. The media psychologists Amy Orben and Andrew Przybylski found in three large data sets covering more than 355,000 adolescents that the use of digital technology explains at most 0.4 per cent of the variation in well-being, less than wearing glasses. What we look at changes us, but perhaps less than fear would have it.

In the end, questions remain to which this book has no answer. I pass them on, not as a shortcoming but as an invitation.

By what would we recognise that a machine experiences something? How can one test whether a system shares its principles or merely follows them? Who sets the purpose of the systems that work in all cultures, and by what right? On what is human dignity founded for those who do not believe in a Creator? How does the norm that human beings bear ultimate responsibility become a reality when they have less and less time to check? And what does it say about a hope that it helps, if that does not yet say anything about its truth?

What also remains open is where this upheaval leads. History shows that there have always been upheavals, and some had immense consequences. I do not believe that the world will come to an end because of this, nor that we will gain eternal life through this technology. Nor, so far, has technology given us more time. The economist John Maynard Keynes predicted in 1930 that, thanks to progress, his grandchildren would work only about fifteen hours a week. The day still has 24 hours, and most people are busier than ever.

What really makes a life good is examined by one of the longest studies in the world. Since 1938 the Harvard Study of Adult Development has accompanied hundreds of men, and now their children too, through life. Its most important finding: not fame and not money, but good, reliable relationships make people happier and healthier. Whether technology gives us more time for such relationships or less is, for me, one of the most important open questions.

Equally open is what it does to people when information reaches them so condensed and so centrally curated. For twelve years I have been examining students' final theses. In this time their length has roughly halved, and in my impression the theses have become less dense and less substantial. This is a personal observation, not a study, and it may have many causes. But it fits what Chapter 5 reports about unlearning. Much remains open. It is a remarkable time.

In Chapter 7 I said that I was neither an opponent nor an advocate of artificial intelligence, perhaps both, depending on the situation. I stand by that. I work with these systems every day, and I am enthusiastic about what they can do. They have multiplied my effectiveness, clarified my thinking and opened up paths for me that I would not have found without them. I would never have managed to write a book like this in a few weeks with an ordinary AI assistant of the kind most people know today. It became possible with a system that was developed in Switzerland. It accompanied me through the whole writing process and helped me to work cleanly, in the research as well as with the references. Others built it, and I am amazed at what they have achieved. All the more do I wish that more such systems would come into being here, open and in the service of people (Chapter 14). At the same time I have experienced in myself how easily one looks less closely, how quickly a dependence arises and how little time remains to check. And I have experienced that I can be influenced not only by systems but also by images, slogans and the group to which I belong. During the Covid period, and later on political questions, I adopted judgements that were shaped more by selection and repetition than by my own examination. A friend who contradicted me showed me this.

This book is not neutral, and it does not want to be. I write from a Christian faith that carries me. I work closely with a Swiss company that develops systems to support decision-making, and I wrote this book with such a system. In doing so I also used a language model from Anthropic, one of the companies the book writes about. And some thoughts come from texts of my own, which likewise came into being with the support of an artificial intelligence. Readers should know this in order to judge for themselves where I report and where I speak.

At the beginning there was the question of what I am prepared to give up. By the end it is no longer a private question.

We will relinquish much, and much of it rightly. No one has to remember telephone numbers any more, and no one has to evaluate fifty hours of interviews by hand when a system can do it in a few hours. But there are things we cannot relinquish without relinquishing ourselves: judgement about what we answer for; regard for the individual human being who eludes calculation; the conscience that speaks up before we can give reasons for it; and the question of what we do all this for.

I believe that precisely those decisions that are carried by responsibility, meaning and soul will, in the long run, remain the most important, beyond our own generation. The machines will not replace us in this. But they will force us to know more precisely who we are.