Predicting the future has always been an imperfect art, especially when it comes from those with direct interests in the industry. Bill Gates recently stated that by 2035 we could be working only two days a week, thanks to artificial intelligence.
An idea that, let's face it, is tempting for anyone who spends 40 hours a week (or more) in the office. And yet, listening online his words on the US talk show “The Tonight Show” with Jimmy Fallon, I can't help but notice how distant this vision is in my opinion from the current reality of AI.
On the one hand, Gates compares the evolution of artificial intelligence to that of computers: before they were expensive, now they are accessible to everyone. And this is an assumption, of course. On the other hand, I wonder: is the scheme also applicable to AI? Is this prediction based on concrete data or is it just another case of technological optimism disconnected from reality?
“Rare” intelligence will become common, and we will work two days a week
During the interview, Gates expressed his vision: we are entering an era in which intelligence, which is rare and expensive today, will soon become free and widespread. “The era we are just entering is one in which intelligence is rare. A great doctor, a great teacher… And with AI in the next decade, that will become free. Common, you know? Great medical advice, great mentorship, and so on,” he explained to Fallon.
The prospect is certainly fascinating: a world where medical expertise is available instantly, where every student has access to personalized tutoring, and where the shortage of professionals is solved thanks to artificial intelligence. Above all, according to the founder of Microsoft, this transformation could be so profound that it will drastically reduce our need to work.
AI brings so many changes. What will jobs be like? Should we only work two days a week, or maybe three? – Bill Gates
When Fallon asked if we'll still need humans, Gates responded with disarming frankness: "Not for most things." A statement that drew mixed reactions online, and in me.
Why it might work…
The idea of a drastically reduced workweek is not completely unfounded. We already have concrete evidence that working less can bring significant benefits. In the UK, a pilot program testing a four-day week produced such positive results that 86% of participating companies decided to keep it. Productivity has not decreased, while workers' well-being has improved significantly. And there are also other examples.
If AI could truly automate much of the repetitive tasks that occupy our workdays, we could theoretically focus on what humans do best: creative thinking, empathy, complex problem-solving. In this scenario, two days of human labor might actually be enough to complete the tasks that AI cannot handle.
…but he won't
The current reality of AI tells a very different story than Gates’s optimistic predictions of a two-day workweek. There are at least two flaws in his vision, in my opinion.
First: the shortage of professionals in crucial sectors such as healthcare and education is not a technological problem, but structural: reflects the lack of investment and support in these professions. To be mean, an unfortunate choice.
Second: Current AI is still deeply problematic in the very industries Gates envisions revolutionizing. Google's chatbot, Gemini, continues to provide incorrect information in search results. Medical machine learning algorithms, despite advances, continue to discriminate against women and people of color, failing to correctly diagnose certain diseases.
In a context where health inequalities are already widespread, the premature adoption of these systems risks exacerbating them further.
More than two days
To the two flaws I add two more things. First of all, a bitter irony: AI is primarily trained to perform creative tasks (writing, creating art) that humans could do in their spare time if they worked less, not yet to replace repetitive or dangerous tasks. Moreover, these models are often trained using “stolen” works of art or pirated material, as demonstrated by recent controversies that involved ChatGPT and Studio Ghibli.
Artificial intelligence is not designed to replace boring or dangerous jobs, but primarily to perform creative tasks.
Finally, an elephant in the room: the environmental impact of AI. Artificial intelligence technologies (except perhaps DeepSeek) have an alarming carbon footprint and exorbitant water consumption, so much so that companies like Microsoft itself are failing to meet their own climate goals precisely because of the race to AI.
Gates’ prediction, in short, seems to ignore all these issues, projecting us into an idealized future that has little to do with the current reality of technology and the world of work. I don’t want to rule it out, of course, but let’s put it this way: to see something similar it will take… Much more than two days.
