But it may be difficult to quell those fears when 41% of employers plan to reduce staff whose skills are becoming less relevant by 2030, according to a new survey from the Work Economic Forum. And younger workers—digital natives who are more aware of tech’s abilities, and may be more vulnerable to job cuts—are especially afraid. About 62% of Gen Z staffers think AI will replace their roles within the next decade, according to a 2024 survey from General Assembly.
Broadening digital access is expected to be the most transformative trend – both across technology-related trends and overall – with 60% of employers expecting it to transform their business by 2030. Advancements in technologies, particularly AI and information processing (86%); robotics and automation (58%); and energy generation, storage and distribution (41%), are also expected to be transformative.
While many studies (e.g. by the World Economic Forum) assume that overall job losses will be compensated by the new jobs created, I am much more skeptical and essentially share the view of Dario Amodei and Sam Altman. Most of the studies I know of so far ignore the fact of exponential development - they merely assume the status quo, but do not extrapolate the data. However, what we saw at the beginning of this analysis in Part 1 and Part 2 is the rapid development, progress and emergence of new artificial intelligence capabilities. If we take this development in conjunction with the billions of dollars of investment (e.g. Stargate project $500 billion, EU AI investment $200 billion) as a premise, we can assume with a probability bordering on certainty that AI will certainly create new jobs in the future, but that future AI will also be able to do this work. Consequently, I view all the studies with a certain degree of skepticism. Not because I don't trust the studies, but because it is hardly possible to map the development statistically; the development is too erratic with too many undetermined variables.
In addition, most studies are based on a time frame of 5 years, i.e. 2025-2030. The development of AI is far too fast for long-term predictions to be possible. In short, all empirical data should be read with caution and always against a certain theoretical background. Nevertheless, let's take a look at what we know today.
Since the introduction of generative AI tools such as ChatGPT at the end of 2022, concrete effects on digital freelance markets have been demonstrable. According to a DIW study, demand for automation-prone jobs fell by an average of 20% within eight months, with peaks of 30% for typing and 20% for software development. Graphic designers and 3D modelers recorded a decline of 17% as AI algorithms increasingly take over routine tasks. This development exceeds the impact of traditional industrial robots and is also reflected in the IT sector: in the USA, the unemployment rate among IT specialists rose from 3.9% in December 2024 to 5.7% in January 2025 - an indication of AI-related rationalization.
ChatGPT and image creation AI tools have quickly shaken up the freelance labor market in certain areas. Generative AI is still in its infancy, so the world of work is likely to change even further and much more than before,” says Jonas Hannane from the Companies and Markets department at DIW Berlin. Together with Ozge Demirci from Harvard Business School and Xinrong Zhu from Imperial College London Business School, Hannane collected and analyzed over one million jobs advertised on a large online platform for freelance work from July 2021 to July 2023 for the study
A study by the McKinsey Global Institute predicts that up to 30% of current total working hours could be automated by 2030, leading to potential annual productivity growth of up to 3%. At the same time, however, hundreds of thousands of jobs could be lost, particularly in administrative office activities, customer service and production. By far the greatest upheaval will affect administrative office activities: Up to 54% of expected job changes in Germany, for example, fall into this area.
The advance of (generative) artificial intelligence will fundamentally change the labour market in Europe and the USA. Rapid use of the new technologies can lead to productivity growth of up to three per cent per year. This assumes that up to 27% of individual activities within all job profiles will be automated and would result in up to 12 million job changes by 2030.
Nevertheless, the OECD again emphasizes that AI does not so much destroy jobs as reorganize job profiles. In experiments, AI support led to productivity increases of 14% for programmers and 35% for customer service employees, with more complex tasks resulting in higher budgets. This trend illustrates a dualism: on the one hand, entry-level jobs for repetitive tasks are decreasing, while on the other hand, more demanding positions requiring AI-supported problem solving are emerging. Although it must be emphasized once again that the OECD study is already more than a year old and therefore assumes a different basis (as I wrote above), which is one of the biggest problems. This disparity is particularly evident here. While the OECD's study assumes that AI can be expected to increase productivity in programming, Mark Zuckerberg from Meta assumes that coding AI agents will completely replace mid-level SWEs (Software Engineers) this year! A stark contrast. And it is almost certain that Zuckerberg has excellent knowledge of the internal research and development of AI models.
Overall, AI systems are increasingly taking over repetitive and predictable tasks, which is leading to a transformation of many job profiles. In accounting, for example, AI-supported programs can automate routine tasks such as data entry or invoice verification. This allows professionals to focus on more analytical and strategic tasks, such as financial analysis or consulting.
In customer service, chatbots are used to answer frequently asked questions. This allows employees to focus on more complex customer issues that require human empathy and problem-solving skills.
The impact of AI automation varies greatly depending on the industry. High-risk areas such as writing and content creation are particularly affected. AI-supported tools such as ChatGPT are increasingly taking over text creation, translation and editing tasks. According to studies by OpenAI, up to 50% of the activities of accountants, journalists and interpreters could be automated. IT and software development is also undergoing a profound change: AI-supported code generators such as GitHub Copilot significantly accelerate development processes while at the same time reducing the need for junior programmers, as many standard tasks are automated.
Similar changes are emerging in medicine and the legal sector. Diagnostic algorithms support doctors in the evaluation of medical data, while contract analysis systems relieve the burden on legal professionals. However, these developments not only lead to the automation of existing tasks, but also require specialists to work more closely together across disciplines and acquire new skills at the interface between man and machine. At the same time, new job profiles are emerging, for example in the areas of AI ethics, data protection and human-machine collaboration.
The World Economic Forum predicts a net increase of around 58 million jobs worldwide by 2025, while up to 700,000 new jobs could be created in Germany alone. However, this change often requires advanced digital skills, further exacerbating existing skills gaps. Those who want to survive in the new world of work must continuously adapt and develop technological skills.
According to the studies mentioned above, AI is not a singular shock, but a continuous transformation process. In the short term, repetitive knowledge workers in particular will be replaced, while hybridized activities that combine human creativity with AI efficiency will emerge in the medium term. The greatest challenge lies in the fair distribution of the burden of adaptation. Without comprehensive educational reforms and social dialogue, the division between AI beneficiaries and losers threatens to jeopardize social cohesion.
In Summary, the Following Can Be Stated
The debate on the influence of artificial intelligence on the labor market reveals a fundamental discrepancy between theoretical forecasts and current empirically measurable developments. While empirical studies are mostly based on retrospective data and analyze the status quo, theoretical assessments operate on a more abstract level and anticipate possible qualitative leaps in AI development.
This difference is particularly evident in the work of thought leaders such as Ray Kurzweil, Dario Amodei and Sam Altman. Kurzweil, for example, uses logarithmic scales to model technological developments and predicts the singularity - the point at which machines surpass human intelligence and continue to develop independently - for the year 2049. Amodei, on the other hand, predicts a superintelligence that surpasses humans in all areas as early as 2027. It is striking that neither Amodei nor Altman provide concrete data or measurable parameters for these predictions - but this is less due to a lack of foundation than to the nature of the object of investigation. This is because the progress of AI does not follow a linear path, but rather an exponential development that is difficult to predict.
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."
This is the core problem with empirical analyses: they are inevitably backward-looking. Studies by institutions such as McKinsey, the OECD or the World Economic Forum (WEF) are based on quantitative methods that primarily measure past developments. So when these studies describe the influence of AI on the world of work, they are providing a snapshot of the period from around 2021 to 2023 - a period in which AI was still barely noticeable for many people in their everyday lives. But the picture is changing rapidly.
The technological breakthroughs of recent years, particularly in generative AI, have meant that AI-supported systems are no longer just support tools, but are increasingly making autonomous decisions and performing complex tasks independently. While earlier studies examined the impact of chatbots such as ChatGPT, OpenAI's “Operator” at the beginning of 2025 already provided a glimpse of a future in which autonomous AI agents play a central role. These technological leaps are hardly taken into account in existing empirical analyses, as they simply did not exist at the time of data collection.
This is precisely why theorists such as Kurzweil rely on higher-level growth models to estimate the further course of AI development. They consider not only the current performance of AI systems, but also the underlying acceleration of technological innovations. This approach takes into account the fact that new capabilities are emerging at ever shorter intervals - an observation that is already being confirmed today.
By definition, the singularity itself defies empirical analysis. It describes a kind of technological novelty, a tipping point after which AI systems develop independently and accelerate their own progress exponentially. Such a development cannot be depicted in classic economic models or empirical data sets, as it moves outside of historical reference points.
Empirical studies undoubtedly have their value when it comes to quantifying the impact of AI on the world of work to date.However, they are like looking in the rear-view mirror as we head at breakneck speed into an unpredictable future. The dynamic nature of AI development therefore requires a combined approach: on the one hand, precise empirical analyses to capture current trends and, on the other, visionary models that take into account the exponential nature of progress. Those who rely solely on historical data run the risk of underestimating the speed and scope of the changes to come. The real challenge is to close the gap between theory and empiricism - before reality overtakes us.
Concluding Thoughts
In any case I think meaning comes mostly from human relationships and connection, not from economic labor. People do want a sense of accomplishment, even a sense of competition, and in a post-AI world it will be perfectly possible to spend years attempting some very difficult task with a complex strategy, similar to what people do today when they embark on research projects, try to become Hollywood actors, or found companies. The facts that (a) an AI somewhere could in principle do this task better, and (b) this task is no longer an economically rewarded element of a global economy, don’t seem to me to matter very much.
I will never tire of emphasizing the need for a broad social discussion about life and the distribution of resources in a post-labor economy. On the one hand, we need to discuss what work means as a meaningful activity and, on the other hand, what wage labor is; and in particular, if wage labor is eliminated due to its replacement by AI, how to allocate the distribution of all produced goods. And even more profoundly: is capitalism a sustainable system, or has capitalism done its job for people and the world through its productive efficiency? But what is needed in the transition phase, in the change from a laboratory to a post-labor society, in which traditional professions are gradually falling victim to AI rationalization? Is universal basic income (UBI) needed to bring about socially acceptable change?
I suspect that some new and stranger thing will be needed, and that it’s something no one today has done a good job of envisioning. It could be as simple as a large universal basic income for everyone, although I suspect that will only be a small part of a solution. It could be a capitalist economy of AI systems, which then give out resources (huge amounts of them, since the overall economic pie will be gigantic) to humans based on some secondary economy of what the AI systems think makes sense to reward in humans (based on some judgment ultimately derived from human values).
At this point in time, it makes sense to at least try to benefit from the technological revolution. I would like to emphasize the short-term nature of this, as AI can certainly call everything into question. In the short term, you can participate in the steady rise on the stock market via shares or, more conveniently, via ETFs. Be it via an S&P 500 ETF, which to a certain extent tracks all the beneficiaries of AI, or more concentrated in an AI/semiconductor themed ETF or via the direct purchase of share positions in well-known AI companies or hyperscalers (provided they have even had an IPO and gone public. OpenAI, for example, has so far deliberately decided against an IPO so as not to pave the way for possible takeovers). But when I say short-term, I have to point out in this context that the AI revolution is leaving no area of society untouched. When I write these lines, I have to keep reminding myself that AI will really change all areas of life in a valid, empirically verifiable and theoretically plausible way. The AI revolution is not a pipe dream or esotericism, but a scientific fact. And therefore also empirically valid with an impact on the entire financial market (it is no coincidence that the founder of DeepSeek was the first to launch a fund that outperforms using machine learning).
But what if AI becomes so good in its predictions that the future can no longer be traded on the stock market because AI can predict everything? What if the superintelligence uses its power to turn a stock market speculation into a stock market fact? The purpose of the stock market would no longer apply, price determination through supply and demand would become meaningless because an AI would determine the future development of a share 100 percent. In short, when I say that you can participate in short-term success by investing in AI on the stock market, then this statement must always be understood against the background of the serious social influence of AI, namely that no area of life remains unchanged and “safe”.
Another crucial point in the debate about the influence of artificial intelligence on the world of work is the fundamental question of the future of work itself. Previous analyses have often discussed which professions will be replaced or transformed by automation. In the long term, however, it is not just about the redistribution or adaptation of existing activities, but about the possible end of wage labor as we know it. If machines and algorithms take over not just individual tasks, but entire occupational fields, work in its current form will simply no longer exist.
This development requires an in-depth social discussion about the role of work in our lives. Is work merely a means of earning a living, or does it fulfill a meaningful function that goes beyond this? Historically, work has always been a central element of social organization - whether in feudalism, industrialization or the service society. But with the advance of automation and the replacement of human labor by AI and robots, we may be facing a paradigm shift: a post-labor society in which wage labor becomes obsolete as a cornerstone of the economic system. To put it very clearly and unequivocally: in my opinion, there is no profession that is absolutely future proof. The technology revolution is taking place at such a rapid pace that all predictions are being disgraced by empirical reality. All safety guarantees can only be given for a few years, everything after that is uncertain and vacant. Future proof in times of AI is an illusion. All renowned scientists from OpenAI to Anthropic agree: the speed at which Ai is currently developing is unlimited. There is no end in sight!
This change will have far-reaching consequences. On the one hand, it raises the question of the material security of the population: if people no longer have to work to earn a living, alternative mechanisms of resource distribution will be needed. The debate about a universal basic income (UBI) is becoming increasingly important in this context, but whether such a model is sufficient or whether completely new economic structures are required remains to be seen. On the other hand, we need to consider what people will do with their newly acquired time. Will they take on new creative, scientific or social activities? Or is there a threat of a crisis of meaning if the traditional concept of work loses its significance?
At first glance, a world without work might seem utopian - a state in which everyone exists without economic constraints and can devote themselves to self-realization. However, such a development also harbors risks. In a society that has regarded work as an identity-forming factor for centuries, the elimination of wage labor could lead to disorientation and social tensions. Without a profound cultural reassessment of what constitutes a fulfilling life, there is a risk of a crisis of meaning in which people will have to redefine their role in society.
One thing is certain: The transformation towards a post-labor society is no longer just a theoretical possibility, but a foreseeable reality. The challenge now is to actively shape this change instead of just reactively managing it. Education systems need to be restructured, social safety nets rethought and philosophical questions about the meaning of work in an AI-dominated age answered. Without a forward-looking societal debate, we risk being overwhelmed by technological development instead of steering it in a direction that benefits humanity as a whole.
But the time to discuss this is right now.
—
Ready for more content from Kim Isenberg? Subscribe to FF Daily for free!
