For years, it has been repeated that artificial intelligence (AI) primarily threatens office jobs, while manual trades seemed safe. However, the arrival of robotics presents a new scenario. Axelle Arquié, an economist at the Center for Prospective Studies and International Information and director of the Observatory of Threatened and Emerging Jobs, warns that while manual jobs currently resist, they will not be exempt from technological impact in the long term, as she stated in an interview with Le Monde. Geoffrey Hinton's recommendation, Nobel Prize winner in Physics 2024 and a reference in AI, to "become a plumber" to avoid automation was taken as a joke, but it contains a temporary truth: first comes AI, then robotics, or both combined. The trend to create humanoid robots is not just for empathy but for their ability to adapt to human environments and perform manual tasks, which expands the scope of automation. Arquié warns of the risk of a "social catastrophe" if the extent of these changes is not properly measured. She recalls that while the internet revolution destroyed jobs, it also created many others, but AI represents a different leap: it is not just a tool but an agent capable of replacing entire tasks. The debate intensified after statements from Mustafa Suleyman, head of Microsoft AI, who claimed that most office tasks could be replaced by AI in the next eighteen months. Arquié calls for caution: developers tend to exaggerate the disruptive potential, while politicians and economists often downplay the risk. The transition, she warns, could be as profound as the Industrial Revolution, which took decades to improve the general standard of living and caused significant social costs. Economist Philippe Aghion argues that new jobs will emerge, but Arquié doubts they will be sufficient to compensate for those lost. Recent examples, such as European deindustrialization, show that labor retraining is neither automatic nor simple. The threat is not just the disappearance of jobs but their transformation. Drawing inspiration from sociologist Juan Sebastian Carbonell, Arquié speaks of "cognitive assembly lines," where tasks are fragmented: part is done by AI, part by the worker, relegating humans to more routine and less creative tasks, weakening their bargaining power and reducing wages. There is an optimistic view: AI as a tool to increase productivity. However, Arquié points out that many companies will prefer to replace rather than complement, especially if AI proves more profitable and efficient. Examples like the legal sector, where AI can draft documents and search for case law, illustrate this trend. Unlike industrial robotization, generative AI can take on non-routine tasks typical of skilled professionals, affecting high-salary jobs and tax contributions. Moreover, agentic AI is already capable of executing complete processes autonomously, approaching the final frontier of job replacement. The main current obstacle is the lack of explainability: models function as black boxes and still make errors, although less frequently. Arquié recalls that fifteen years ago, autonomous cars seemed impossible, and today they are already circulating in cities like San Francisco. Technical evolution and social acceptance are advancing rapidly. Many see manual trades and vocational training as a safe exit, as they require improvisation in new situations, something that AI and robotics have yet to master. But Arquié warns that, in the long term, the combination of AI and robotics, especially with advancements in China, could automate even these jobs. In this context, the economist identifies two major challenges: Europe's dependence on foreign technology and the need to redistribute the wealth generated by automation, which could lead to debates about universal basic income. She proposes not to halt innovation but to capture part of that wealth to avoid a social and democratic crisis. "Training is necessary but not sufficient," concludes Arquié. The real challenge is to reorganize the economy so that the technological revolution does not leave millions behind. The advice to become a plumber may be useful today, but the real debate is how to adapt society to a future where technology redefines work. (Source: huffingtonpost.es)