{"id":198548,"date":"2026-06-29T11:53:14","date_gmt":"2026-06-29T09:53:14","guid":{"rendered":"https:\/\/www.erasmusmagazine.nl\/en\/?p=198548"},"modified":"2026-06-29T11:53:22","modified_gmt":"2026-06-29T09:53:22","slug":"de-invloed-van-ai-op-startersbanen-geen-reden-tot-paniek-wel-tot-nadenken","status":"publish","type":"post","link":"https:\/\/www.erasmusmagazine.nl\/en\/2026\/06\/29\/de-invloed-van-ai-op-startersbanen-geen-reden-tot-paniek-wel-tot-nadenken\/","title":{"rendered":"The impact of AI on entry-level jobs: no reason to panic, but much to think about"},"content":{"rendered":"\n

Is AI destroying students\u2019 job prospects? In reality, the picture so far is more nuanced than newspaper headlines suggest. Yes, the market for entry-level jobs has changed. But that is not only because of artificial intelligence.<\/p>\n\n\n\n

\"\"

\n\t\n\n<\/svg>\n<\/span>\n\tImage by:<\/span>\n\n\t\t\t\n\t\t\t\t\t\t\tEva Gomb\u00e1r-Krishnan<\/a>\t\t\t\t\t<\/span>\n\t<\/p>\n<\/figcaption><\/figure>\n\n\n\n

\u201cI have read alarming reports in various newspapers that jobs will disappear because of the rise of artificial intelligence\u201d, says history student Sam, sitting on a bench in the sun during his break. \u201cI increasingly wonder what this means for me.\u201d IBEB and philosophy student Roman recognises this feeling. \u201cAI can take over many entry-level tasks. If companies stop hiring starters and mainly use the technology to boost their profits, that will be bad for us.\u201d Further into the campus grounds, business administration student Phedra says she is mainly worried about \u201cthe speed at which everything is changing\u201d. Because of this, she is unsure about what she wants to do after her bachelor’s.<\/p>\n\n\n\n

At the same time, these students, each in their own way, use AI in their daily lives. Whether it is for making summaries, using the technology as a sparring partner or improving their cover letter: they are all using the very technology they fear.<\/p>\n\n\n\n

Figures from CBS (Statistics Netherlands) show that half of young people between 18 and 25 think their work can be taken over by AI. But according to various experts, that view is too simplistic. In reality, the impact of artificial intelligence on entry-level positions in the Netherlands is more nuanced.<\/p>\n\n\n\n

Too simplistic <\/strong><\/h2>\n\n\n\n

In recent months, alarming reports have appeared in various media about the loss of entry-level jobs because of AI. That picture is partly correct. The share of entry-level positions is lower than a few years ago. While more than 20 percent of all vacancies were for starters three years ago, this has now halved to 10 percent. Moreover, the UWV (Employee Insurance Agency) has seen a decline in jobs that are sensitive to AI, such as translators, marketers, account managers and software developers.<\/p>\n\n\n\n

Even so, the idea that artificial intelligence is the only culprit is too simplistic, says professor of Work and Institutions Ferry Koster. He does a great deal of research on the impact of AI on work. According to him, other developments also play a role in this shrinkage, such as geopolitical unrest and restructuring. \u201cIt has always been difficult for young people to enter the labour market, especially at times of uncertainty such as now.\u201d <\/p>\n\n\n\n

Mervyn Nankoe, career adviser at the university, agrees. \u201cWe are talking about two things here: AI and the decline in jobs. It is important to separate the two.\u201d Nankoe stresses that the first is not automatically the main cause of the second. \u201cThere are also fewer jobs because of, for example, corporate cutbacks. AI is only one part of the bigger picture.\u201d<\/p>\n\n\n\n

Tasks are changing<\/strong><\/h2>\n\n\n\n

At the same time, both acknowledge that the rise of AI is indeed causing changes, just not always in the way people expect. \u201cThe debate often focuses on tasks that are taken over, or tasks that emerge\u201d, says Koster. What he misses in the debate is attention for how employees will work differently within organisations. <\/p>\n\n\n\n

\u201cInstead of employees carrying out their tasks individually, this technology can actually lead to more collaboration between departments.\u201d He compares this with earlier technological tools, such as Dropbox and Teams, which changed the way of working at the time. New AI tools can make it easier, for example, to brainstorm together.<\/p>\n\n\n\n

Jo\u00e3o Gon\u00e7alves, assistant professor at the Erasmus School of History, Culture and Communication, observes similar shifts. \u201cAlthough the output of work remains broadly the same, the way it is produced is changing\u201d, he says. \u201cWhere employees used to focus on working out a single solution, they now work on several solutions at the same time.\u201d<\/p>\n\n\n\n

He takes journalism, a profession in which he himself also worked for two years, as an example. \u201cIn the past, a journalist would bounce ideas off a colleague about the headline for a story. Now they can present it to AI, which immediately comes up with several headlines. The journalist then chooses the best one.\u201d In this way, employees\u2019 roles shift. Less creating, more evaluating.<\/p>\n\n\n\n

Learning to work smartly with AI<\/strong><\/h2>\n\n\n\n

Educational institutions play a crucial role in preparing students for a changing labour market. Maurice de Rochemont, lecturer on the master’s in Strategic Entrepreneurship, has been experimenting for some time with the role of AI in his classes.<\/p>\n\n\n\n

For his course Validation and Pivoting<\/em>, students have four weeks to identify and solve a problem. These may be social issues, or something they encounter themselves. De Rochemont sees opportunities for AI at every stage of that process. <\/p>\n\n\n\n

For instance, in the problem exploration phase, students use the technology to explore different perspectives on a problem. \u201cFirst, students formulate their own assumptions about the problem and the target group. They then ask an AI system to assess the same issue. Afterward, they critically examine which ideas and perspectives are useful and which are not.\u201d<\/p>\n\n\n\n

These tools also prove useful when searching for solutions. Some students enlisted the help of AI while working on solutions for spoiled food in the refrigerator. \u201cThe students themselves came up with the idea of an expiry-date app for consumers. AI, on the other hand, suggested developing a product aimed at businesses. That gave these students new ideas.\u201d<\/p>\n\n\n\n

Previously impossible <\/strong><\/h2>\n\n\n\n

AI is also lowering the barrier for building prototypes such as apps and websites. Where students once needed technical skills, they can now bring their ideas to life themselves. \u201cThis allows them to use their creativity in a way that previously was not possible\u201d, says De Rochemont. He mentions a student who built an app that now has about one hundred users. \u201cThat is not a great number of users, but a few years ago something like that would simply have been impossible.\u201d<\/p>\n\n\n\n

More worries<\/strong><\/h2>\n\n\n\n

Career advisers are also noticing the change. Nankoe and his colleague Nikola Janou\u0161kov\u00e1 have recently been speaking to more students who see their future profession changing rapidly and therefore doubt their job prospects. \u201cThey are mainly Marketing, Communication and Arts & Culture students, but they can come from any programme\u201d, says Janou\u0161kov\u00e1.<\/p>\n\n\n\n

Experienced workers have practical knowledge that is harder to automate. Young people, on the other hand, are more often engaged in tasks that can easily be structured, such as summarising or drafting first versions of reports. AI can more easily take over these tasks, and that worries students. <\/p>\n\n\n\n

Yet students are not adjusting their study choices because of this, the career advisers note. \u201cThere are indeed students who more often choose a technical master’s\u201d, says Nankoe, \u201cbut they were already interested in that master’s anyway.\u201d Janou\u0161kov\u00e1 agrees: \u201cStudents\u2019 passion remains the guiding factor.\u201d<\/p>\n\n\n\n

\"\"

\n\t\n\n<\/svg>\n<\/span>\n\tImage by:<\/span>\n\n\t\t\t\n\t\t\t\t\t\t\tEva Gomb\u00e1r-Krishnan<\/a>\t\t\t\t\t<\/span>\n\t<\/p>\n<\/figcaption><\/figure>\n\n\n\n

Reassuring students<\/strong><\/h2>\n\n\n\n

Janou\u0161kov\u00e1 and Nankoe have a clear plan of approach when students come to them with concerns. \u201cFirst of all, it is important to pause to consider those worries\u201d, says Janou\u0161kov\u00e1. \u201cWhat are they, what do they mean, and are they accurate?\u201d Nankoe adds: \u201cStudents have read something or heard something from fellow students, but the reality is often unknown to them. By testing assumptions and setting everything out clearly, more clarity emerges.\u201d <\/p>\n\n\n\n

One concrete tip Nankoe gives is to talk to alumni about how they see their work changing. Various faculties provide opportunities for this. For example, the Erasmus School of Law has a mentoring programme with alumni, while career days are regularly organised across the university. Students who want extra support can make an appointment with Career services and draw up an action plan together with a career adviser.<\/p>\n\n\n\n

Skills that will become more important<\/strong><\/h2>\n\n\n\n

Janou\u0161kov\u00e1 expects human skills to become even more important now that AI is taking over certain tasks. She is thinking, among other things, of intercultural skills and the way students present themselves: their so-called \u2018personal touch\u2019. <\/p>\n\n\n\n

The need for fact-checking will also increase. \u201cStudents must be critical about what language models churn out. They must be able to properly assess AI\u2019s work before applying it\u201d, says Nankoe. \u201cThey can gain this knowledge and experience during their studies.” <\/p>\n\n\n\n

De Rochemont explicitly trains his students in this. During his lectures, they analyse different prompts and investigate which assumptions and biases are hidden in the models. \u201cStudents must know what they are working with. And sometimes their own ideas simply fit the problem of the target group better than what AI proposes.\u201d <\/p>\n\n\n\n

Finally, adjustability is crucial. \u201cThe work landscape is changing rapidly. The ability to cope with these changes is becoming increasingly important. Our advice to students is to develop this skill\u201d, says Janou\u0161kov\u00e1.<\/p>\n\n\n\n

Mismatch<\/strong><\/h2>\n\n\n\n

Expectations surrounding new technologies matter. They shape the behaviour of students, universities and companies. For example, students may make different study choices, while organisations adjust their recruitment policies. In doing so, they act not only on the basis of what AI can actually do, but also on the basis of what they think it can do. <\/p>\n\n\n\n

Koster believes that what is written in the media does not always match what happens in the workplace. \u201cOn the one hand, you hear the grand narrative that AI is going to change everything. At the same time, various reports show that productivity gains vary greatly from situation to situation.\u201d As a result, this can lead to a mismatch between what employers expect from AI and what people who work with it think is possible. <\/p>\n\n\n\n

\u201cSometimes it is questionable whether AI actually adds value\u201d, says the professor. He illustrates this with a metaphor. \u201cI also think it is very clever if someone can make a bear ride a bicycle, but that is of no use to me and is even harmful. In short, AI simply does not always work equally well yet.\u201d <\/p>\n\n\n\n

\u201cWe are now in a situation where organisations are mainly experimenting\u201d, says Koster. \u201cBut we are still a long way from a world in which we can outsource all our knowledge. AI is not a miracle device. There will always need to be checks and balances<\/em>.\u201d<\/p>\n\n\n\n

Setting the direction <\/strong><\/h2>\n\n\n\n

Koster stresses that the direction of artificial intelligence is not fixed. \u201cThere are theories that assume technological determinism\u201d, he says. In tech jargon, this is the idea that new technologies change us, rather than the other way round. \u201cIn a perfectly rational world, this might be true, but that is not how our world works. People always have views on technologies, and they work with technologies they do or do not like.\u201d <\/p>\n\n\n\n

\u201cIn that respect, there are indeed choices we can make about how we use AI.\u201d Fortunately, more and more people, especially young people, seem to be realising this. That is particularly important, Koster believes, because they will be working with this technology for the longest time.<\/p>\n\n\n\n

People at the centre <\/strong><\/h2>\n\n\n\n

The success of AI will depend on the choices organisations make. If the transformation is gradual, education programmes can prepare students well for the work that does increase. If the change is rapid, the consequences for the labour market will be more painful. A recent report by TNO also notes this: the challenge does not lie in a shortage of work, but rather in the match between supply and demand. <\/p>\n\n\n\n

Koster is hopeful in this respect. \u201cIn the short term, it may seem attractive that fewer costs are incurred, but organisations know very well that people remain indispensable. In the end, organisations will realise that there will always be circumstances in which they get more out of people than out of AI systems.\u201d<\/p>\n\n\n\n\t

\n\t\t\t\t\t

Een lijst met artikelen<\/h2>\n\t\t\n\t\t
\n\n\t\t\t\n\t\t\t