Public debate keeps asking how many jobs artificial intelligence will destroy. Yet a deeper transformation is already underway: the boundaries between professions are dissolving.
Pietro Paganini’s latest analysis on artificial intelligence, jobs and the future of work was also published by HuffPost Italia, where it was featured on the homepage.
In a nutshell:
- AI is not just automating work. It is recombining skills that once belonged to separate roles.
- Jobs are being unbundled into tasks that can be automated, AI-enhanced, or left to human judgement.
- The real advantage will be knowing how to frame problems, verify answers, and take responsibility for decisions.
The AI and Jobs Debate Is Asking the Wrong Question
The debate about AI and the future of work tends to swing between two predictions: artificial intelligence will eliminate millions of jobs or, like previous technological revolutions, create new ones.
Both may be partly right. But both risk missing the transformation already underway.
A salesperson builds a website. An engineer prepares a financial analysis. A researcher develops communications materials. Tasks once associated with different professions are becoming accessible to the same person.
AI is not simply automating work. It is recombining skills.
The decisive question is no longer only how many jobs AI will destroy or create. It is how much time workers, businesses and institutions have to redesign work before technology redesigns it for them.
AI Does Not See Professions. It Sees Problems.
OpenAI’s Work at the Frontier report analysed more than 800,000 work-related ChatGPT messages. It found that 43.5% of occupation-specific, non-generic messages concerned tasks historically associated with another occupation, a pattern the report calls task crossover.
For more than two centuries, productivity has been organised around specialisation. Companies built departments, universities divided knowledge into faculties, and workers shaped their identities around professional titles.
AI operates differently. It does not follow organisational charts or recognise boundaries between marketing, finance, research and communications. Given a problem, it combines the capabilities needed to address it.
For AI, the question is not: “What is your profession?” It is: “What problem needs to be solved?”
Work Is Not Disappearing. It Is Being Unbundled.
Specialisation will not disappear. Experience, technical expertise and deep knowledge will remain essential.
But a job will increasingly look less like a single block and more like a bundle of tasks: some automated, some enhanced by technology, and others made more valuable precisely because they require human accountability.
AI does not make everyone an expert in everything. But it lowers the cost of complexity, giving professionals, entrepreneurs and small businesses access to capabilities that once required specialists and dedicated departments.
AI distributes tools. It does not distribute talent.
The same technology can amplify knowledge and creativity, but also errors, bias and superficiality. That is why an almost Renaissance-like quality may become valuable again: not knowing everything, but knowing what to ask, what to verify, what to connect and what to decide.
The Human Advantage Is Judgement
The better AI becomes at execution, the more responsibility shifts to human judgement: understanding context, asking better questions, connecting disciplines, challenging persuasive answers and taking responsibility for decisions.
The defining skill will not be using AI. It will be judging it.
AI can produce an answer. A human being must decide whether that answer deserves to become a decision.
Preparing people for the future does not mean adding an AI course to programmes designed for the past. It means teaching them to define problems, interrogate tools, verify information and move across disciplines.
The deepest transformation brought by artificial intelligence may be the end of the idea that a person’s economic value can fit inside a single job title.
The greatest risk is not that a machine takes a job. It is that education and training continue preparing people for rigid roles while work is already reorganising around problems to solve.