Weakness is human

Since the machine started talking to us, the topic of artificial intelligence is back in fashion. Some even expect humans to be replaced by a super-intelligence. The ideas are anything but new. They rely on us forgetting that human life and coexistence are not limited to thinking and talking.

Computers have been calculating faster (and more reliably) than humans for a long time; that’s exactly why they were invented. No one was bothered by this: The use of punch cards and magnetic tapes was too complicated, the large devices were too unwieldy. But even when the IBM computer Deep Blue defeated the then world champion Gary Kasparov in chess in 1996, there was no great excitement. It was quickly understood that chess is nothing more than an optimisation problem in a rule-based environment. Deep Blue was able to analyse and evaluate over 100 million positions per second. The world champion could no longer keep up. The fact that Deep Blue’s successor Watson prevailed against the defending champions in the 2011 Jeopardy game on American television did not make it onto the German front pages.

But now that the machine tells stories and paints pictures, researches and writes newspaper articles, analyses company data and handles corporate planning, the excitement is huge. Newspapers are publishing weekly newsletters; companies are developing AI strategies, and the valuation of digital companies is reaching new highs. There is a gold-rush atmosphere. At the same time, however, the working world of the cultural and economic elite is changing: suddenly they themselves are the subject of technological progress, which they previously only described or managed. Concern about this high-risk technology is also gripping the intellectual elite, because they themselves no longer understand what is happening. Much of the unrest could have to do with this.

But not only that. Knowledge, reason and language are the very definition of being human. We are homo sapiens, the knowing human being. We are the animal rationale, the rational living being. And according to a definition that predates Aristotle, we are the zoon logon echon, the living being that has language. Language seems to be the core of what we are. Without language, our knowledge and reason remain mute. With language, we enter into contact with the world and with other people.

Now, thanks to an incredibly large data set and even greater computing power, the machine has managed to reformulate language as a mathematical-statistical phenomenon and impress us with its analysis. We hear it speak. And we like what we hear, as it reflects all the digitised works of world literature and the findings from scientific publications of an entire generation: “The spirit of the world from the machine”, as the philosopher Hegel put it for the cover story in March 2023.

Is the machine becoming human? Are we on the verge of an intelligence explosion? – This question is also an old one. As early as 1965, the British mathematician Irving John Good predicted the invention of an “ultra-intelligent machine” that would be far superior to humans in all intellectual abilities. This machine would be the last invention that humans would have to make, after which they would no longer be needed.

Ten years ago, the Swedish philosopher Nick Bostrom, who teaches in London, described artificial intelligence as the most promising way to develop such a “superintelligence”. As a transhumanist who expects and welcomes the improvement or replacement of humans by a technologically optimised life form, Bostrom even began to consider whether and to what extent humans would be obliged to treat thinking machines according to human moral standards and, if necessary, to be of service to them.

However, the debate sometimes feels as if the last hundred years of philosophical reflection have not taken place. They take Hegel’s “world spirit”, which is now over 200 years old and is supposed to be the realisation of reason in world history, translate it into a logical-mathematical problem and expect that the algorithms of superintelligence will one day solve it.

Ludwig Wittgenstein already knew over 100 years ago that human problems are not nearly solved when all scientific questions have been answered. And anyone who has listened to the philosophers of existentialism, critical theory, phenomenology or post-structuralism over the past 50 years knows that the questions of reason, language and the human being are not easy to answer: Human existence is not limited to intelligence, and our access to the world is more than just analysing and processing data.

Philosophy calls the phenomenon that human beings do not just happen to have a body, but that they are this body, “corporeality”. Human existence does not result from a combination of hardware and software but is expressed in a body that is sentient and limited in space and time.

In our sensitivity, we encounter the world and others without being able to defend ourselves against them. “Man bathes in the elements”, wrote the French philosopher Emmanuel Levinas. We metabolise our world long before we analyse it or talk about it. We have no control over our pleasure, and even less over our pain. In physical pain, we are exposed to the world and cannot free ourselves from it. Where the machine would switch off a malfunctioning sensor. we cannot. We age. No upgrade and no update possible. Even a simple organ replacement is life-threatening.

The spatial and temporal limitations of our existence make it unique. No one can replace me in my body. And I cannot replace anyone in their body. Everyone must die for themselves; everyone must endure the fragility of their own existence until it is no longer bearable. It is not possible to copy the software onto another (identical) piece of hardware. Our body bears the traces of our existence.

This gives meaning to human existence. It is in caring for the lives of others and dedicating our own lives to them that human community is created. Being able to dedicate our lives to only a few is not a problem of our limited computing power, it is the beauty and tragedy of human existence that we call love. The machine cannot love because it is not finite, it cannot suffer because it has no body, and it cannot die for the other because it does not live. It may be able to think and talk, but will it ever understand what it means to love, suffer and die? We hardly understand it ourselves.

The danger of artificial intelligence is that we fall for the machine in all its eloquence and sophistication, that we allow ourselves to be fooled into thinking that all questions can be solved and that humans can be replaced simply because the machine fulfils some functions of our everyday (professional) lives faster and more accurately than we could. When we replace humans with machines in our everyday interactions, we deprive ourselves of that surplus of humanity that is not needed for economic transactions and is not absorbed in intelligence or language.

The office is not just a place where quarterly financial statements, construction drawings, contracts or newspaper articles are produced. Schools and universities are not just places where knowledge is imparted. And doctors’ surgeries and clinics are not just places where illnesses are treated. Everywhere we meet people in their individuality, with all their idiosyncrasies, strengths and weaknesses. Even if machines one day become better managers, engineers, lawyers, journalists, teachers or doctors, they will not become better people. But they could prevent us from becoming them, because the machine illusion of perfection makes us forget and unlearn humility and generosity in dealing with the weaknesses of others.

This essay was published in the German weekly WirtschaftsWoche N°34, August 16, 2024

The dialectic of liberalism

Almost 250 years ago, Immanuel Kant defined the Enlightenment as man’s emergence from his self-inflicted immaturity. Since then, liberalism as a philosophical, political and economic movement has invoked this individual maturity and has shaped Western societies and their economic systems. But its appeal is waning. Right-wing and populist forces are gaining strength in democracies; at the same time, totalitarian dictatorships are opening up their economies but not their societies. The popular equation “free economy = free society” is increasingly proving to be false.

This comes as no surprise. As early as the 1950s, the German philosopher Theodor W. Adorno warned that the economic order would keep the majority of people in a state of immaturity even after the end of fascism. He was less concerned with concrete economic well-being and the lack of economic means for self-realisation. Rather, he saw the economic system as a form of coercion that stood in the way of the individual’s experience of autonomy. The broken promise of freedom and autonomy made people indifferent to democracy and susceptible to the promises of totalitarianism.

Since Adorno’s diagnosis, this compulsion, we want to argue, has become fully apparent in society as an anonymous system – in the professional, private and political spheres – and is now, with the breakthrough of artificial intelligence, likely to jeopardise democracy in its very essence.

The professional world has become anonymous and offers less and less of a social home. Formerly owner-managed shops – supermarkets, drugstores, DIY stores – have given way to branches of large nationwide or even international companies and have left the inner-city life. Processes, product ranges and workwear have been standardised. The responsible entrepreneur has been replaced by the branch manager who is controlled by KPIs. In large corporations, local structures have been replaced by functional ones. Cross-sectional tasks have been centralised, outsourced or offshored. Who used to be a colleague is now an external service provider and may be replaced tomorrow. And within one’s own workplace, agile teams, teleworking and flex offices are dissolving binding structures. Whoever comes to work becomes a cog in the wheel.

Private life has become more complicated and economised. Thanks to the liberalisation of formerly public services, citizens now have to make more economic decisions than ever before. They can choose from hundreds of electricity, gas and telecommunications tariffs. They have to choose a health insurance company and, if possible, opt for some private pension plan. The situation is unmanageable. After signing the contract, people often discover that they have missed out on a special offer, not taken advantage of a retention or switching bonus or not cancelled a free additional offer in due time. 

And politicians are also increasingly resorting to economic instruments to steer society: CO2 taxes and pollution-dependent lorry tolls, long-term tuition fees and education vouchers, cigarette and alcohol taxes, the list goes on. The state exerts economic pressure, which mainly reaches the lower classes; the wealthy have the financial means to buy their way out.

The liberalisation of formerly public services and the targeted taxation of undesirable behaviour can be seen as liberal democracy’s response to the acceleration of modernity described by Hartmut Rosa: Where all aspects of life are continually accelerating, democratic decision-making and compromise-taking take too long. Economic logic is faster, but also more liberal, because it does not impose rules, but lets people decide how and where they spend their money.

All these developments have brought us economic benefits: longer shop opening hours, a wider range of goods, cheaper products, fewer government bans. But they have also increased the coercive nature of an anonymous system. Today, it is difficult to find someone to assume responsibility. It always seems to be the anonymous market that we cannot escape. In 1984, the German Minister of Posts still had to answer to parliament for the 6-month waiting times for a telephone connection at the Lübeck telecommunications office. Today there is no longer such a minister; and the answer to the lack of fibre optic connections or poor network coverage is economic constraints and “the (failing) market”. Yet the market is proving effective in preventing unprofitable projects, however much they are politically desirable.

Liberalism is blind to the fact that such market mechanisms can constitute coercion. Liberals conceive the market as the sum of the free decisions of responsible citizens. But while liberal societies continue to try and take the economy into their service, the international tech scene has long been thinking about a different future. Accelerationism is the name of a school of thought that believes humans are too imperfect and the institutional structure of the democratically organised state is too slow. The philosopher Nick Land propagates the organisation of society in the form of stock corporations with citizens as shareholders and strives for a comprehensively technologised and de-humanised economy.

Since the breakthrough of Artificial Intelligence, this is within reach. The structural consequences for the relationship between business and society have so far been underestimated. In its detailed and differentiated statement on AI, the German Ethics Council analysed four fields in more detail –  medicine, education, public communication and opinion-forming as well as public administration – but mainly focused on individual applications. However, the fact that machines and algorithms take over individual processes or process steps in companies is nothing new; for example, a large part of stock exchange trading is already automated today, and the autopilot lands an aeroplane more smoothly than a human.

What is new is the possibility of fully networking individual processes and automating the entire company, from purchasing and employee scheduling to sales. The Chinese company Netdragon Websoft replaced the CEO of its Fujian subsidiary with an artificial intelligence in August 2022 – albeit presumably as a PR stunt. What we are looking at is an independent software that can participate in economic activity as a company. This is not primarily about reducing costs, as the current discussion about job losses would have us believe. The far greater leverage of AI is to increase sales and margins; in short: we should all buy more and pay more.

We will therefore probably see the first fully digital companies in the sale of simple contracts: electricity and gas, telephone and internet, banking and insurance services. The platforms have long been in place, the processes are digitalised and run on Amazon, Google and Microsoft servers in the cloud. At the moment, the last people are still working in customer service, in product development and in programming the processes. As soon as the computer can reliably communicate and programme independently, they too will be replaced. And as the computer always has all the relevant data available across our various needs, it can adapt its sales strategy in real time by analysing its human counterpart. As it doesn’t get tired even after eight, twelve or sixteen hours of use, it will become an irresistible salesperson and seducer. Gilles Deleuze’s metaphor of capitalism as a machine becomes reality.

Do we need to redefine economic freedom in the face of new technological realities? – Liberalism rightly defends the freedom of the human entrepreneur. It is difficult to imagine a free society without free choice of profession and free entrepreneurship. The economist Friedrich August von Hayek went so far as to celebrate the economic society as the saviour of democracy from itself. Democracy has a tendency to become “inevitably egalitarian”, he wrote. With the legal fiction of the “legal person” (Article 19 of the German constitution), we even guarantee basic rights for companies and are prepared to weigh these up against the basic rights of humans. But would we also guarantee the fundamental rights of a company controlled by machines if it replaced a company controlled by humans? – Blindfolded, Justice will not see the difference.

Enlightenment, which, as Kant said, requires “only freedom”, will not be enough. He still considered our immaturity to be self-inflicted through “laziness and cowardice”. Liberalism has come a long way with this thesis: We probably live today in the most prosperous and freest of all worlds. Kant was confident that in a free society, a few would start to think and thus enlighten the rest. But they will have a hard time against the machines and their art of seduction. We are becoming increasingly immature in the face of the machine. It knows more about us than we do ourselves, knows our pain thresholds when it wants us to work for it and our weak points when it wants to sell us something.

If Adorno’s thesis that the experiences of coercion associated with the economic system give the right a boost is correct – and various current studies show correlations between dissatisfaction with democracy and the rejection of (economic) globalisation as well as the feeling of no longer belonging to society – then we need to think about how we can put people back in a position to experience maturity.

We need to strengthen individual entrepreneurs over anonymous companies and give people the chance to work in an environment that allows them to experience autonomy.

We need to push back the economic decisions in the private sector, which many people are unable to understand, and which create little value for society as a whole. Nobody needs the multitude of providers and tariffs for (former) public services of general interest.

And in politics, we would have to dare to make decisions politically again instead of inventing a market mechanism and then waiting for the smartest and fastest to win in the new market and the less smart and less fast to call for subsidies and postponement.

All of this would mean that our social prosperity would grow more slowly because we are slowing down economic dynamism. This makes issues of social justice more virulent than they are today and may strengthen left-wing groups, but it need not be harmful to liberal society, as it was primarily labour movements and the political left that contributed to the fall of totalitarian regimes, for example in Spain and Greece in the 1970s and in Eastern Europe at the end of the 1980s. The economy, on the other hand, will not rush to our aid against totalitarianism in case of doubt; it lives quite well with it in other parts of the world and has not stood out as a great resistance fighter in Germany’s past either.

First publication: WirtschaftsWoche 35, 03/11/2023