Artificial intelligence is reshaping the way we think about technology — and about work itself. Between fears of mass job loss and the very real limits of current technology, the question of AI’s true impact on employment remains open. So what do we actually know right now?
Over the past three years, since the debut of ChatGPT, AI has evolved at lightning speed, changing how we navigate the digital world. Chatbots can now handle countless online tasks, from writing and research to coding and customer service. As a result, more routine and repetitive work is being automated, fueling fears that AI might eventually replace humans entirely — pushing millions out of the workforce.
The real impact: mostly on young graduates
Recent headlines seem to back up those fears. Major companies are announcing mass layoffs and blaming AI for the cuts. Yet, more often than not, that’s a convenient excuse. A new study from Yale University found that AI hasn’t yet had a measurable effect on employment in the United States. The data points instead to a slowing job market linked to broader economic conditions.
Meanwhile, research from King’s College London found that job losses have mainly affected recent graduates — particularly in companies offering the highest salaries. In the long term, those same companies could struggle to fill more senior roles that demand greater experience and expertise.
Are we overestimating AI’s abilities?
Despite the hype, today’s AI systems are far from perfect. Large language models — or LLMs — are known for their “hallucinations,” confidently producing incorrect or misleading information. A study published earlier this year found that chatbots made factual errors in nearly 60% of web-based research tasks.
Many experts now argue that the architecture behind current models simply can’t match the depth of human reasoning or lead to so-called “general artificial intelligence.” Even AI pioneer Yann LeCun, one of the field’s founding figures, has left Meta to pursue a new kind of model called Jepa. According to him, we’re still decades away from machines that can truly think — or work — like humans.
Yann LeCun details his vision of artificial intelligence, which he considers will not replace human intelligence. © University of Geneva
Evolution, not extinction, for human work
Even if AI one day matches human capability, history suggests that every major technological revolution creates new kinds of jobs. Two decades ago, few would have guessed that being a YouTuber or influencer could be a full-time career. And now, new roles like prompt engineer are emerging to guide and fine-tune AI systems — roles that might evolve or even vanish as technology matures, but won’t disappear entirely.
And if the worst happens — if automation wipes out most human jobs — the world already has an idea on standby: universal income. Countries like Finland and Germany have tested this system, in which the government pays citizens a fixed sum regardless of their resources. Early results suggest it can improve both employment and overall well-being.
In the end, the fate of work in the age of AI won’t be decided by the machines themselves — but by us. The balance between innovation, opportunity, and policy will determine whether AI becomes a threat or simply the next step in human progress.
Edward Back
Journalist
My passion for programming began with my very first computer, an Amstrad CPC 6128. I started coding in Basic, then moved on to Turbo Pascal on a 286, eventually exploring more modern languages including web development. I’m also deeply interested in science, which led me to attend a math-focused preparatory program. Later, I studied psychology with a focus on the cognitive aspects of artificial intelligence.