Something important is developing, but it is quite curious that such a decisive political issue has not yet entered the political agenda of those who govern, with the exception of Spain. A few days ago, while a guest of the most influential American podcaster, Joe Rogan, the Democratic senator from Vermont, Bernie Sanders, explained it simply, perhaps too much so: “You are a worker, and your productivity is increasing because we give you artificial intelligence, right? Instead of throwing you out on the street, we’ll reduce your work week to 32 hours”. I would add an “if” to his reasoning. If AI improves the productivity of companies and services in many sectors, what do we do with that increase in productivity? Should technology improve the lives of people or only those who produce it? Could the time freed up by AI be used, for example, to spend more time with our children or elderly parents, to do sports or volunteer work, or not?
In Spain (the European country that had the greatest growth in 2024, 3.2 percent compared to the EU average of 0.8 percent), the government reached an agreement with the unions to cut weekly working hours with the same salary by the end of 2025, going from the current 40 hours to 37.5, working four full-time days or five part-time. It’s a first step, and it was motivated not by ideological motivations but by the increase in productivity resulting from the use of technology. The bet is that by making that free time available at the same salary levels, consumption will not be reduced, but rather that citizens will be able to save on expenses such as babysitters or carers (and therefore on the cost of living) and improve the nation’s quality of life by reducing public costs for healthcare and welfare. Finally, those who work less by leaving the most repetitive tasks to technology should work better. In organizations, time savings can lead to reduced operating costs (for example energy costs) and greater efficiency, allowing resources to be redirected towards more strategic initiatives, thereby promoting competitiveness.
Simple, one might say: why isn’t everyone doing this? In fact, many companies around the world are already doing it, from Microsoft and Panasonic in Japan to Unilever in New Zealand. According to the “Future of Professionals Report 2024” by Thomson Reuters, the use of AI tools could free up an average of four hours a week, reaching eight hours in three years and twelve hours in five years: an “AI dividend” that translates into roughly two hundred hours a year. In Great Britain, in a test developed by the government, twenty thousand civil servants from twelve agencies used AI to process documents, emails and reports, saving an average of 26 minutes a day, equivalent to almost two weeks of work a year. If this was extended to the entire system in the UK (around 514,000 employees), AI could “free up” nearly 29,000 thousand people a year. Is it better to make fewer people work more or more people work less?

According to a recent study by the Alan Turing Institute, about 41 percent of public sector work time could be supported by the use of GenAI, which translates to 3 to 4 hours per day out of a total of 8.3 hours worked. Obviously, the effectiveness of AI varies across sectors, reflecting the nature of different activities and the degree of reliance on human interaction and tolerance for error, ranging from 49 percent in education (primary and secondary) to 33 percent in healthcare. On average, public-sector workers spend thirty minutes a day on email, time that could be reduced by over 70 percent with AI.
Of course, increased productivity can also be simply “exchanged” for linear and brutal cuts to employment. This is what is happening in Silicon Valley. The series of testimonies collected by Brian Merchant (a former tech columnist for the Los Angeles Times) with the project “AI Killed My Job” is shocking. “AI first” strategies are becoming the pretext under which companies such as Duolingo, Klarna and Dropbox are firing their employees. AI is used to justify job cuts, accelerate existing work and make the remaining employees take on the workload of laid-off colleagues. CrowdStrike has laid off five hundred employees, presenting the operation as a “strategic realignment” to double its investments in AI and accelerate efficiency.

Big Tech company CEOs prefer to talk about something else. An exception to this, and because of this noise, was the alarm raised at the end of May by Dario Amodei, founder and CEO of Anthropic (the company behind the chatbot Claude), about the very high risk of unemployment among white-collar workers: in the US, artificial intelligence could eliminate half of entry-level office jobs in the next five years, bringing American unemployment to 10 or even 20 percent. Activities such as data analysis, report writing, contract reviews and planning could soon be largely automated. Amodei urgently recommends the creation of tools such as large-scale professional retraining programs, new welfare networks (even if he is not convinced about a universal basic income) and “productivity sharing” to redistribute the gains of the best productivity, what we might call the “artificial dividend”.
Here we return to the initial “if”. If AI does not improve work productivity and the quality of life for all, what is its purpose? Moreover, is it not also in the interest of Big Tech for artificial intelligence to be perceived as an opportunity rather than a threat? Work will probably no longer be so central in the society of the future, but there is the issue of wealth redistribution which should be addressed. The problem is that, with few exceptions, those who govern in the United States and Europe prefer to remain silent.

Cover image from “ Future of Professionals Report 2024 ”


Aggiungi la tua firma e il codice fiscale 94097630274 nel riquadro SOSTEGNO DEGLI ENTI DEL TERZO SETTORE della tua dichiarazione dei redditi.
Grazie!

