Robots, artificial intelligence. And the crucial question about jobs. There will be fewer and fewer jobs that a robot cannot do better. Brave new world of work? AI will revolutionize our working world, to a greater and faster extent than we can imagine. How well prepared are we for this? And how good is the politics? Digital Minister Karsten Wildberger is already talking about an unconditional basic income because of AI. While the Chancellor is still demanding of the people: That's why we have to work more. But what if AI takes over more and more jobs? That's what this film is about. What other work do we need? And who will pay for our pensions, our kindergartens, or the repair of the roads if there are fewer and fewer taxpayers? Because of AI. This is Ella, a robot dog. Ella can do all sorts of tricks. (Man) Give me paw. Fine. Come with me, little man. Sometimes you forget that Ella is not a dog, but a machine. Ella lives in the AI ​​Village in Hürth near Cologne. TV shows were previously produced on the site. Today, the aim is to give artificial intelligence a boost in the Rhineland from this location. And robot dogs like Ella are supposed to take over jobs. Jobs that are still done by humans today. Robot dogs that inspect trains... ...or monitor company premises. In the production sector, we are already seeing the first use of four-legged animals that walk around reading sensors, measuring temperatures, etc. We are seeing automated analysis of machine data in the production area. So we are already seeing a great deal in use. Faster, more efficient, more precise. Through AI. The robot dog is just one example. Since artificial intelligence arrived for most of us with ChatGPT just 4 years ago , we have been able to write texts, generate images and videos in seconds. AI has become an integral part of everyday life. The possibilities seem limitless. But what does this mean for our jobs? Let's ask the AI ​​directly. How drastically will AI transform the world of work? The Industrial Revolution. Production was massively increased. More and more people were working in the factories. As wage laborers, often under inhumane conditions. But the fear that these technological revolutions would cost jobs has not materialized, says Christian Kellermann from the German Research Center for Artificial Intelligence. If we think in terms of two machine ages: The first machine age was still about work and power machines. The second machine age is the data-processing machine. As in the first industrial development or revolution with other types of machines, it did not lead to mass unemployment. On the contrary. We ultimately experienced an increase in employment. Will it be the same now? Who is afraid of AI? And who should be afraid? Researchers asked themselves this question in an international study. Result: AI primarily affects professions that are highly cognitive, not physical. For example, data analysts, software developers, or translators. This is therefore different from previous historical upheavals. Back then, machines primarily replaced muscles and hands. AI is replacing brains, computer work in the office. But what does that mean for the people at the desks? We are on our way to becoming a company where AI has been used for a longer period of time. In Hanover. Edgar Ennen is part of the management team at the consulting firm EGGERS & PARTNER. The company employs around 40 consultants. AI helps them to familiarize themselves with the businesses of their clients. They call this training "preliminary projects". Depending on the size of such a project, we sometimes had to invest up to 20 working days in this preliminary phase to familiarize ourselves thoroughly with the topic. This can now be reduced to one week through good data collection and structuring . His people only need about 1/4 of the time to do the same work. Ennen says he will not lay off any people because of this. Rather, they could provide more intensive support to their customers and take on more projects. But let's assume all management consultancies worked so much more efficiently: Would their clients also have so much more work to employ the consultants who were freed up? Where will this end? Let's ask the AI ​​again. Does AI primarily increase the efficiency of companies, or does it also lead to layoffs? The answer: So far, so imprecise. And what do economists say? Will AI make us more efficient or unemployed? Marcel Fratzscher is the head of the German Institute for Economic Research. Our research shows that, first and foremost, over the coming years, the potential to support people at work is greater than the potential to displace them. But that will change eventually. Eventually, AI will be so efficient and intelligent that it will be able to replace humans, and in the long term, most of the paid jobs we know today will probably be done by AI. And no longer by humans. Perhaps these are already the harbingers, the latest reports from tech companies. HP is expected to cut between 4,000 and 6,000 jobs. Meta, Facebook's parent company, announced in the spring that it would be cutting 8,000 jobs. Amazon plans to cut a total of around 30,000 jobs in its office and administration departments. And AI is always cited as a justification. Job losses due to AI. How is politics reacting to this development? During his visit to China at the end of February, Chancellor Friedrich Merz seemed quite impressed by what his hosts' humanoids were capable of. The message: Robots and AI are capable of taking over almost any human job soon. And at home? At home, Merz demands that people should work more to boost the economy. Overall, the productivity of our economy is not high enough. To put it even more clearly: Work-life balance and a 4-day week will not allow us to maintain the prosperity our country enjoys today. That's why we need to work harder. Are people too lazy? Is Friedrich Merz perhaps completely unaware of the transformation of the working world brought about by artificial intelligence? We meet Christoph Bornschein. He has been studying the consequences of AI for the economy and society for years. What does he say to Merz's announcement? Overall, it must be said that the political landscape is emerging as one that is very much driven by such an idea of ​​"back to". So, the blissful era when we were world market leaders in everything and everyone. I believe the one crucial point where a change of thinking is necessary is that there is no "going back to". The basic assumptions and premises that defined the economic system of the 1950s, 60s, 70s and 80s in Germany – cheap energy and all that, combined with industrial production – are simply no longer there today. But AI is also creating new jobs. And there are still professions where people cannot simply be replaced. Craft professions, for example, or social professions, where human skills are essential. It is the teacher, the female teacher, the doctor, the female doctor, nurses, nursing professions. The question then becomes: How can we make them more attractive and better paid? And how do you create processes that make people want to do them? The volume of work isn't decreasing, that's true. But there is the problem of timing. An insurance claims adjuster will not become a nurse next week. People who lose their old jobs to take on new ones? But how can this transition succeed? What's coming our way? We asked Finance Minister Lars Klingbeil about this. His ministry is downplaying the situation. So everything's fine then? Should we continue for now? Klingbeil's cabinet colleague, Digital Minister Karsten Wildberger of the CDU, apparently sees things differently. He even came up with an unusual proposal, especially for a Union politician. A universal basic income as part of the solution. So, a basic financial provision from the state that covers people's living expenses without any obligation in return. Ironically, Elon Musk demanded this almost 10 years ago. There will be fewer and fewer jobs that a robot cannot do better. What can be done about mass unemployment? This will be a huge societal challenge. I think that ultimately we will need some kind of unconditional basic income. It seems like Musk cares about people. In reality, many tech entrepreneurs and experts in the US expect AI to create a new underclass. Poorly paid and underemployed. According to Marcel Fratzscher's logic, a basic income is more of a form of social assistance. Mark Zuckerberg and Elon Musk have a different idea behind it. They say that AI is destroying so much work, making so many people unemployed, that we need to appease them with a universal basic income. But that is not the logic behind the original idea, behind the unconditional basic income. Namely, that everyone can choose what kind of work they do. This job no longer exists at Musk and Zuckerberg. It could therefore be that AI replaces old jobs faster than humans are even able to take on new jobs. That millions of people will need to be retrained, and that the value of human labor is decreasing, thus putting pressure on wages. One scenario. But what does that mean for our society? Who then pays taxes and social security contributions? We'll ask the AI ​​about that first, too. Do we need a different tax system because human labor is being replaced by AI in many areas? Less tax on income, more on capital. The demand is not new, but it is becoming even more relevant now. In 2025, total tax revenue amounted to approximately 990 billion euros. Approximately 263 billion came from income tax, which is directly deducted from wages or salaries. About a quarter. Only around 25 billion euros were generated from share profits through the "withholding tax" . Just under 2.5%. (Kellermann) If the wage share decreases constantly due to higher levels of automation, the classic value-added share of labor becomes smaller. One can then ask the question: Is the current taxation arithmetic between capital and labor still properly balanced, or does it need to be rebalanced? More taxes on capital gains, then. But for that, one would primarily have to consider the companies that will make the biggest profits with AI in the future. Not least the large US corporations, which also earn billions in Germany. Bergheim, March 12. Groundbreaking ceremony for a massive project. The US software giant Microsoft is building 3 large data centers in the middle of the former lignite mining region. Everyone is beaming, happy about a bit of structural change. However, experts say there is a catch. When Microsoft builds a data center, Microsoft operates a Microsoft cloud in a Microsoft data center. This means that the value creation takes place at an American corporation. Such data centers are the basis for future billion-dollar profits for the tech giants. Experts have been criticizing for years the fact that they pay hardly any taxes on these profits in Germany because the income ends up in tax havens. Only a small fraction of Microsoft's profit can actually be taxed. Because what Microsoft does with these data centers, namely providing results for AI models to a company in Italy, South Africa or Germany, cannot be taxed. Because this service is booked in Ireland or elsewhere where Microsoft pays minimal taxes. But what sums of money are we actually talking about here? The Tax Justice Network has calculated that the five US corporations Microsoft, Apple, Google's parent company Alphabet, Amazon and Meta alone earned an estimated 20 billion euros in Germany in 2025, but only paid an estimated 560 million euros in taxes. Effectively just 2.8%. That means, despite all the euphoria, the profits from new data centers will probably mostly end up here in the USA. Microsoft stated in response to an inquiry that it was not yet certain whether and to what extent the data centers would be used for AI. Furthermore, we fully comply with all applicable laws. AI is the next stage of capitalism. And one of the crucial questions could be: How do we create the necessary acceptance when it intervenes so massively in people's lives? The implementation of artificial intelligence in our daily lives will fail, and our democracy will fail, if we do not succeed in creating social acceptance for this change. And today, for most people, AI is more of a threat than an opportunity. That's something we need to work on. We must design this AI in such a way that it acts in the interests of all people. Ultimately, this acceptance also depends on the question of who benefits. Society? Or just a select few who profit directly or indirectly from the AI ​​boom?