AI and jobs: what the research says about automation and how to prepare

TUREFF• •0

Few questions about the future worry people more than whether artificial intelligence will take their job. Predictions range from mass unemployment to a productivity boom that creates more work than it destroys. The honest answer is that nobody knows precisely, but research offers useful clues about who is affected and what you can do.

Exposure is not the same as replacement

Many studies measure how exposed jobs are to AI, meaning how many of their tasks AI could plausibly perform or assist with. Exposure does not mean the job will disappear. In January 2024 the International Monetary Fund estimated that almost 40 per cent of global employment is exposed to AI, rising to around 60 per cent in advanced economies. Crucially, the IMF suggested that roughly half of those exposed jobs might benefit from AI raising productivity, while the other half could see AI take over key tasks, lowering demand for labour. It also warned that, in most scenarios, AI was likely to worsen inequality without policy responses.

Tasks, not whole jobs

Economists generally find that technology automates tasks rather than entire occupations. Most jobs are bundles of different tasks. When some are automated, the job often changes rather than vanishes: people spend more time on the parts machines handle poorly, such as judgement, relationships and dealing with the unexpected.

How this wave is different

Earlier automation mainly affected routine manual and clerical work. Generative AI is different because it touches tasks involving language, writing, coding, analysis and images, which means many office and professional roles are more exposed than before. Some research also suggests that less experienced workers may gain the most from AI assistance in certain tasks, which could narrow skill gaps within some roles while putting pressure on entry-level positions in others. Evidence is still emerging, and studies conducted in one workplace may not generalise.

Lessons from history

Past technologies, from electricity to computers, eliminated many jobs while creating new occupations that were hard to imagine beforehand.
Transitions have often been painful for the individuals and places affected, even when the overall picture improved over decades.
Adoption usually takes longer than early predictions suggest, because businesses must reorganise around new tools.

What is hard to predict

How quickly AI capabilities will keep improving.
How fast companies will adopt them, given costs, reliability, regulation and trust.
Whether productivity gains will translate into more hiring, lower prices, higher wages or higher profits.
What new kinds of work will appear.

Be wary of anyone offering precise figures for jobs lost or gained by a given year.

How to prepare

You cannot control the economy, but you can improve your position:

Audit your own tasks. List what you do in a typical week and consider which parts AI tools could already help with or take over.
Learn the tools. People who use AI well in their field are likely to be in a stronger position than those who avoid it. Start with low-risk tasks and check outputs carefully.
Strengthen human skills. Judgement, communication, negotiation, leadership, care and hands-on work remain hard to automate.
Build domain expertise. Deep knowledge of a field lets you spot AI's mistakes and apply it well.
Keep learning. Short courses, professional training and employer schemes can help you adapt. Check what is available through your government, union or professional body.
Watch the trends in your sector. Industry bodies and professional associations often publish guidance on how AI is changing work.
Build a financial cushion where possible, so a change of role is less frightening.

For employers and policymakers

Researchers commonly point to retraining, social safety nets, support for workers moving between jobs and involving staff in decisions about new technology as ways to make transitions fairer. How is AI changing your work so far? Share your experience in the comments.