Analysis: How does AI affect teachers' work?
Danmarks Statistik (Statistics Denmark) has, together with researchers from the Copenhagen Center for Social Data Science (SODAS) at Københavns Universitet (University of Copenhagen), published an analysis examining how large language models (LLMs), such as ChatGPT, affect the Danish labour market. In this article, I will very briefly introduce the key findings, focusing on the parts that show how the education sector may potentially be affected. It is well worth delving into the analysis yourself; it is genuinely fascinating reading and can serve as a marker for the transformation our society is undergoing as a result of developments in artificial intelligence.

By applying the AIOE score, which measures employment in relation to AI exposure, to Danish labour data from Danmarks Statistik, the article highlights which occupational functions and sectors may potentially experience changes as a result of artificial intelligence. It is important to point out that the analysis does not distinguish between whether language models augment or replace job functions, and according to the analysis there are currently few job functions where artificial intelligence takes over all work.

The main conclusions are that people whose work involves many cognitive routine tasks will face significant change; and here the education sector is very likely to be affected by the new opportunities. This means both that teachers will be able to exploit the advantages of language models, but also that they will experience changes in their daily work. Examples include personalised learning, automated feedback on assignments, or the construction of interactive learning environments.
In addition, the analysis indicates that women generally have better opportunities to make use of language models in their work compared with men, and high-income earners appear to have more advantages than those with low incomes.
The analysis provides insight into how the labour market is developing for educators, and also an indication of which sectors will need to engage with language models and work on further training.
Below, I have included the overview of occupational functions and sectors they expect will be affected. Both tables appear in the analysis.


