Half of Dutch jobs might be significantly changed by generative AI

The researchers looked at the impact of generative AI on the tasks and skills needed to perform 113 occupations. What is missing from other analyses of the potential impact of generative AI is the application of specific labour market data, including labour shortages. ‘According to  our knowledge, this is the first of this kind of analysis for the Netherlands’, say AI specialist Mona de Boer and Bastiaan Starink, leader of PwC's People & Organisation practice. ‘We then combined this analysis with estimations on labour shortages in occupations and sectors in the Netherlands and with an adoption factor, the likelihood of sectors actually adopting the new technology.’

Sectors that are AI-exposed, where labour shortages are expected to be high and that have a high technology adoption factor include information and communication, financial institutions, education, specialised business services and public administration and services. In these sectors, more than 75 per cent of jobs are at least highly exposed to generative AI. Hence, AI offers opportunities for increasing productivity and thus contributing to partially solving the labour scarcity problem in those sectors.

Occupations much less affected (25 per cent or less) are in sectors such as retail, agriculture, forestry and fishing, construction and in manufacturing industries. This is a marked difference from previous automation waves that actually had a greater impact on the manufacturing industries, or so-called ‘blue collar’ jobs.

‘In the past, we have often seen that jobs don’t disappear because of technology, but that it changes jobs’, says De Boer. ‘With previous technological advancements, we have seen that many jobs did not disappear as expected, but the nature of those occupations changed. AI takes over administrative and repetitive work, giving workers time for things that require human contact, knowledge, creativity and empathy. AI is really not going to replace the human factor in work. But it offers very good opportunities to make work more efficient and perhaps more interesting. Compare ChatGPT and similar tools to a new colleague joining you to do the work together.'

De Boer continues: ‘The possibilities AI offers for reducing workloads are now being explored by organisations that have acute staffing problems. Those are already turning to this technology in the short term to increase the workforce, so to speak. However, it is equally important to also think about the medium term. If you provide services or products or have a profession that is changing under the influence of generative AI, you have to ask yourself how relevant your service or your skills will still be in a few years' time.'

Bastiaan Starink points out the urgency for organisations to analyse the impact of AI on their professions and their services. 'Earlier, you had to be a data scientist to work with AI, but with tools like ChatGPT, everyone can get by, even if you know nothing else about AI. Just look at how many people are experimenting with it, even in the workplace. Generative AI has started to take its place in organisations, just not yet in a structural way.'

Like Mona de Boer, Starink emphasises the opportunities: it can help organisations that are currently stuck in their operations. But, he says, ‘if you want to make good use of the potential of AI, there is a challenge. Organisations that start talking about change and upskilling often face resistance. It makes employees insecure; they don't like it because it reminds them of the distress they felt in high school. Our research Hopes & Fears shows that young employees are enthusiastic about AI, but older ones are a bit more indifferent. The trick is to get an entire organisation on board.’ 

Mona de Boer thinks the success of technology depends on how people ultimately interact with it and are willing to change. 'You have to put serious effort into making the promise of technology come true. It's not a magic wand.'