Most jobs that exist today might disappear within decades. As artificial intelligence outperforms humans in more and more tasks, it will replace humans in more and more jobs. Many new professions are likely to appear: virtual-world designers, for example. But such professions will probably require more creativity and flexibility, and it is unclear whether 40-year-old unemployed taxi drivers or insurance agents will be able to reinvent themselves as virtual-world designers (try to imagine a virtual world created by an insurance agent!). And even if the ex-insurance agent somehow makes the transition into a virtual-world designer, the pace of progress is such that within another decade he might have to reinvent himself yet again.
Since the beginning of human history, the tools and instruments developed by man have been the key to the development of civilization. It enabled humans to free themselves from the constraints of nature, to farm, to secure their food supply and ultimately to form complex societies. This ability to use tools fundamentally distinguishes humans from other living beings like Apes and Gorillas. With the industrial revolution, this process took on a new dimension: technology and increasingly specialized tools made people more and more productive, machines replaced muscle power, computers expanded cognitive abilities and the internet revolutionized access to knowledge. But now we are on the threshold of a new era: artificial intelligence is changing the world of work on an unprecedented scale. The question is no longer whether it will affect jobs, but to what extent and with what long-term consequences.
The ongoing acceleration of technology is the implication and inevitable result of what I call the law of accelerating returns, which describes the acceleration of the pace of and the exponential growth of the products of an evolutionary process. These products include, in particular, information-bearing technologies such as computation, and their acceleration extends substantially beyond the predictions made by what has become known as Moore’s Law.
Technological progress has always been a means of increasing efficiency. Steam power, electricity and the assembly line accelerated industrial production, created prosperity and enabled the rise of modern civilizations - but human labour remained irreplaceable. In the digital revolution, computers took over many office jobs, but their main purpose was to make human labor more effective: Emails replaced letter writing, and more and more people were needed in administrative and management positions instead of the production processes of the past. Modern AI, on the other hand, is no longer just a tool - it is developing into an actor that can independently take over human activities in many areas.
A key characteristic of AI systems is their scalability and endurance. Machines do not need breaks or vacations and do not fall ill. They work around the clock with consistent quality. And they learn. In just a few years, AI models have evolved from simple pattern predictions to complex decision-making instances. In combination with robotics, they are already taking over warehouse work, car journeys, medical diagnoses and legal analyses - with increasing perfection.
This development is not only technologically relevant, but also economically. Demographic change is leading to an increasing shortage of skilled workers in Western countries. At the same time, social costs are rising: Fewer and fewer contributors have to pay for more and more pensioners and people in need of care. AI and robotics are therefore increasingly seen as a necessary response to these challenges. Without them, the existing economic system would be virtually unsustainable.
European demography is characterised by an ageing population, driven not only by increased longevity but also (and mainly) by declining birth rates. This fuels significant imbalances between generations, leading to serious social and economic consequences.
But this creates a paradox: while AI and robotics appear necessary in the short term to compensate for the shrinking workforce, they could fundamentally change the structure of the labour market in the long term. Wage labour as we know it could become superfluous in many areas, as companies are forced to use the most cost-effective and efficient solution - and in many cases this will be AI. After all, this is a law of capitalism: efficiency is driven by the pursuit of profit, and the most cost-effective solution will always be used. The question is whether this change merely means a shift in activities or whether it will actually lead to the complete replacement of human labour by AI and robotics.
Historically, labour was never just a means of earning a living, but also an identity-forming process. People will continue to carry out artistic, creative or manual work - but possibly not out of economic necessity, but out of passion and self-realisation. Those who work as craftspeople today often do so because they enjoy the process. A carpenter, a sculptor, a musician - they could all have their products perfected by machine, but human creativity remains a value in itself.
It’s great we live in such a technologically advanced world as well as a fair and decent one”, someone might object, “but with AI’s doing everything, how will humans have meaning? For that matter, how will they survive economically?
A particularly striking example is music: in theory, AI could create perfect compositions in a matter of seconds (and anyone who has tried Udio and SuniAI will confirm this), but people attend concerts because they appreciate the live experience. Music is more than sound, it is social expression. It could be the same with many forms of labour: as long as labour remains a form of expression, it will not disappear completely.
In light of these developments, the question arises: are there any professions that will survive the AI revolution? The idea that certain professions are “future-proof” is not new. Throughout history, there have been professions that have disappeared due to technological progress: Typesetters, telephone operators, coachmen - they were all replaced by machines. Today, we are experiencing a similar development in both office work and skilled trades. AI agents are already taking over many administrative tasks, while robotics is taking over the physical work. The question remains as to whether there are jobs that can never be replaced by AI and robotics.
The transition to a post-labour world appears to be a real possibility, but the exact extent of this change has yet to be investigated. The coming years will show whether the structure of the world of work is changing or whether we need to prepare for a broader transformation.
In the following, I will first take a look at the so-called “white-collar workers” and the impact that AI is already having in this professional field today, before moving on to the “blue-collar workers”. I will then use empirical data to illustrate the disruptive potential of the AI revolution before drawing a conclusion and providing an outlook for the future. In summary, I ask myself the question: are there professions that are “future-proof” or is the technological revolution leading to an unprecedented change, an unprecedented revolution in the human working world? Let's take a closer look.
The White Collar Worker
The crucial problem isn’t creating new jobs. The crucial problem is creating new jobs that humans perform better than algorithms. Consequently, by 2050 a new class of people might emerge – the useless class. People who are not just unemployed, but unemployable.
Labour is a fundamental part of human existence and involves far more than simply performing activities to earn a living. From an anthropological point of view, work is a purposeful, conscious engagement of people with their environment, through which they not only produce material goods, but also shape their identity and create cultural values. It serves as an expression of human creativity and creative power and is deeply rooted in social structures and cultural practices; it gives people and life meaning and significance.
The development and use of tools played a decisive role in the evolution of humans and played a key role in distinguishing “us” from other animals. Around 2.6 million years ago, our ancestors began to shape stones into sharp-edged tools, which enabled them to procure and process food more efficiently. These early tools not only enhanced the physical abilities of humans, but also influenced their cognitive development and promoted more complex social behaviour.
As an interesting aside, engravings on the side of small rocks did in fact represent an early form of computer storage. One of the earliest forms of written language, cuneiform, which was developed in Mesopotamia circa 3000 B.C., used pictorial markings on stones to store information. Agricultural records were maintained as cuneiform markings on stones placed in trays, and organized in rows and columns. These marked stones were essentially the first spreadsheet.
The Neolithic Revolution, which began around 12,000 years ago, marked another significant turning point. The transition from nomadic hunter-gatherer societies to sedentary communities with agriculture and animal husbandry fundamentally changed the human way of life. People developed specialised tools for agriculture, such as sharpened stone axes and sickles, which increased the efficiency of agricultural work. These new tools and techniques not only enabled a more intensive use of the environment, but also led to social changes such as the division of labour and the emergence of more complex social structures.
https://schoolhistory.co.uk/ks3/local-history-study/neolithic-revolution-in-britain/
Through the continuous development of tools and techniques, humans not only adapted to their environment, but actively reshaped it. This process of ‘domestication’ of labour led to man changing himself, both in his physical constitution and in his social and cultural practices. Labour and the associated production of tools were thus central factors that created modern humans and distinguished them from other living beings. In short: labour and tools are not only things that have helped humans to generate wealth, but have also made humans what they are today. Labour and tools are the anthropological prerequisites of modern man. In this respect, they are also aspects that should not be neglected in the overall analysis of human labour.
The concept of labour has changed over the course of history. In early societies, labour was an integral part of daily life and closely linked to the community. The production of tools made of copper and other metals marked important epochs in which people not only increased their ability to survive through their labour, but also achieved technological and cultural progress. The transition to agriculture and the associated sedentarisation fundamentally changed working methods and laid the foundations for complex social structures.
With industrialisation and the introduction of wage labour, the understanding of work underwent a further transformation. Labour increasingly became a commodity traded on the market and the concept of wage labour established itself as the dominant form of gainful employment. However, the fact that labour is more than just a means to an end of generating income often faded into the background. It also creates identity and offers the individual the opportunity to realise themselves and find social recognition. But in modern society, it has developed into a form of expression, becoming part of the personality and a distinct part of self-worth. Whereas in the previous anthropological phases, work generated necessary utility value and enabled survival, work has thus developed a social status in addition to its productive character; one's own personality is expressed in social life through work: it creates self-worth instead of merely ensuring survival.
In the modern world of work, there are various forms of labour that encompass different requirements and fields of activity. A key distinction is made between so-called ‘white-collar’ and ‘blue-collar’ jobs. The term ‘white-collar worker’ refers to employees who mainly perform intellectual or administrative tasks in office environments. Typical examples are managers, lawyers, financial analysts or IT specialists. These professions generally require a higher level of formal education and are characterised by less physical strain.
The arrival of artificial intelligence and robotics in the world of work is already having a noticeable impact on the activities of white-collar workers. AI-based applications help to analyse large amounts of data, automate routine tasks and enable more efficient work processes. For example, language models such as ChatGPT help with composing emails, scheduling appointments, translating texts or even providing customer support. Specialised AI systems such as Claude 3.5 provide support with programming and can generate or check code. Models such as AlphaFold are already revolutionising science and OpenAI's DeepResearch is replacing consulting activities. These technologies increase productivity and allow specialists to focus on more complex and creative tasks. In some cases, however, specialists are already being replaced entirely by AI models, as the financial services provider Klarna recently illustrated.
In the near future, AI and automated systems are expected to further transform the world of work. AI agents, such as OpenAI's ‘Operator’, are designed to independently take on tasks and make decisions that were previously reserved for human workers. As early as 2025, such systems could be able to take over the work of mid-level software developers, as discussed at Meta. In other areas too, such as finance, law and HR management, AI systems could increasingly take over tasks that were previously carried out by white-collar workers.
Anthropic, the major AI opponent of OpenAI, recently evaluated the user data of its AI Claude and analysed the areas of work for which Claude is already being used. It is used for computer and mathematical work in an overwhelming 37.2% of applications. No wonder, because Claude is ideal for programming.
One day later, Anthropic's CEO Dario Amodei an urgent speech in France, in which he also refers to the disruptive effects of AI on the labour market.
Time is short, and we must accelerate our actions to match accelerating AI progress. Possibly by 2026 or 2027 (and almost certainly no later than 2030), the capabilities of AI systems will be best thought of as akin to an entirely new state populated by highly intelligent people appearing on the global stage—a “country of geniuses in a datacenter”—with the profound economic, societal, and security implications that would bring. There are potentially greater economic, scientific, and humanitarian opportunities than for any previous technology in human history—but also serious risks to be managed.
What Dario says should make you sit up and take notice. In a few years, perhaps as early as next year, we will have AI entities capable of performing tasks on a scale equivalent to an entire city of highly intelligent humans.
It's no secret that Anthropic is developing its own agents. If his thesis proves true, it will pose a major challenge to white collar workers. The OpenAI operator gave a small foretaste of what AI agents can already do today and will be able to do even better in a few months' time. Today's reasoning models, on the other hand, are a testament to outstanding mathematical and programming skills (Sam Altman said a few days ago that their latest internal reasoning model is currently ranked among the top 50 best coders in the world and will probably be number 1 by the end of the year). Combined with Anthropics Claude communicating very creatively, the question is what activities can be left for humans.
And there is actually one area that has so far remained relatively untouched by AI: emotional labour. Although AI has excellent communication skills, it can neither ‘understand’ nor authentically ‘reproduce’ emotions. However, these skills are needed in psychotherapy, for example (what is currently available is so-called ‘affective computing’, where an AI interprets emotions based on images. But that is not understanding). Of course, not all psychotherapeutic skills require emotional quality. Behavioral psychotherapy in particular is primarily about the skills that patients learn in order to deal with their challenges. This requires little emotionality. In long-term therapies, in psychoanalysis and in trauma therapy, however, empathy is required. And AI does not yet have this quality of emotion - which does not mean that it cannot be developed (and development is the key word here). Sam Altman recently said that new emergent qualities that do not yet exist appear in the models every 10 orders of magnitude or so (every 100x; Perhaps we see precisely these qualities emerging as part of a philosophical process in which quantity is transformed into quality). On the other hand, there is an urgent need in this field of work. More and more young people are seeking help from psychotherapists.
This phenomenon extends beyond psychotherapy and affects numerous professions that rely on emotional intelligence and authentic human connection. AI can analyze speech patterns, recognize facial expressions, and even simulate responses that seem empathetic, but it fundamentally lacks the core of true human emotional intelligence: subjective experience, self-awareness, and lived empathy.
Take nursing, for example. While robotic assistants and AI-driven monitoring systems are already improving efficiency in healthcare, the emotional labour performed by nurses—providing reassurance to anxious patients, understanding the unspoken distress of an elderly person, or offering comfort to a grieving family—is something AI cannot genuinely replicate; AI can't empathise. Patients don’t just need medical care; they need to feel seen, heard, and understood. A machine may calculate a pain score based on biometric data, but it will not hold a patient’s hand with genuine compassion.
A similar argument applies to teachers and educators. While AI-powered tutoring systems can deliver knowledge efficiently, they lack the ability to inspire, to read the subtle cues of confusion or disengagement in a student’s posture, or to provide the nuanced encouragement that fosters real intellectual and emotional growth. A teacher who believes in a struggling student’s potential can change the course of their life—not through algorithmic optimization, but through human connection.
Then there’s the domain of leadership. Great leaders—whether in politics, business, or activism—do more than just process data and make rational decisions. They inspire, rally people around a vision, and create trust through their emotional presence. Employees do not follow a leader just because of strategic decisions; they follow because they feel heard, understood, and valued. AI might generate optimized business strategies, but it cannot give a rousing speech that makes people believe in a mission.
Moreover, consider creative fields that involve deep emotional labor, such as acting, writing, and music. While AI can generate scripts, compose music, or mimic artistic styles, it does not feel. It does not experience heartbreak, longing, joy, or nostalgia. When an actor delivers a performance that moves an audience to tears, it is not just the words in the script—it is the subtle, deeply human expression of pain or triumph that resonates. This is why AI-generated music, while technically impressive, often lacks the raw soulfulness of human composition.
At the core of all these examples is the fundamental distinction between simulation and genuine emotional experience. AI can predict, optimize, and simulate, but it does not experience emotions. And while it is possible that future AI systems could develop some form of emergent emotional cognition, the time horizon for such advancements is unclear. Sam Altman’s observation about emergent properties appearing every few orders of magnitude in scale raises the question: at what point does quantity turn into quality? But for now, jobs that require authentic emotional intelligence remain future-proof—not because AI lacks computational power, but because it lacks the lived human experience that gives emotions their depth and meaning.
To summarise, it can be said that AI is excellent at taking over the hand tools, the static or even mechanical tasks and performs outstanding one-dimensional tasks. However, it has fundamental problems with recognising and really caring for people. This does not mean that this field cannot also be developed by AI and that there will be possibilities to transfer feelings into the dimension of artificial intelligence, but so far it is more of an adaptation, an attempt to convert emotions into numbers using images and words. However, emotions are transported, they are more than just text modules. Although studies show that people increasingly prefer to talk to AI doctors rather than human doctors, this is mainly because AI takes its time and is not in a hurry. For this reason, the domain of emotional labour has so far been left to humans.
In any case, these developments raise serious questions about the future role of people in the world of work. While, on the one hand, efficiency gains and new opportunities for self-realisation are created by relieving people of routine tasks, there is also the challenge of redefining the value of human labour and ensuring that technological advances benefit everyone. A balance needs to be struck between utilising AI to increase productivity and preserving the human aspects of work that are essential for identity, creativity and social interaction.
Overall, it is clear that work is much more than just generating income. It is a central component of human existence and culture that is constantly changing and adapting to new social and technological realities. At the same time, we can see that white-collar workers are currently more endangered than ever. There are only a few areas in which AI is not yet on the rise. And even these areas only seem to have a certain grace period until AI also develops these emergent qualities.
Our first approximation therefore shows that white-collar workers are particularly at risk. Due to the increasing and evolving capabilities of AI, we are already seeing a significant increase in their use. So far as tools that make us humans more efficient. However, the first agents such as Operator have caused a small outcry in the community. This is because processes and tasks can now be handled one after the other in context by an AI. Agents are the first AI entities that come very close to humans thanks to their complex, multi-stage problem solving. Agents are no longer just focussed on one task and act like a chatbot, but can perform real actions; you can not only take over one task on the PC, but also control the entire PC including the web interface! It has been leaked that the first outstanding software engineering (SWE) agents will be released towards the end of the year. This should be a wake-up call. But overall, white-collar workers are extremely vulnerable. Let's now take a look at the blue-collar worker. We then take a closer look at empirical data.
—
Ready for part 2? Subscribe to FF Daily for free!
