Saving Teacher Time with AI
This page is from a teacher’s perspective, specifically addressing an aspect relevant to practically everyone working in higher education: to what extent can generative AI help free up valuable teacher time?
Below are some potential areas. The quality of the prompts you write and how you follow up the responses with new prompts are crucial, but different subjects and knowledge areas may also be better or worse served by different language models.
As always, when it comes to using generative AI, teachers need to consider the short- and long-term pros and cons in each individual case. Additionally, the material generated always needs to be reviewed and often adjusted.
Formulating Texts
Generative AI can formulate:
- welcome letters for a course
- suggestions for content for web pages
- email responses (never upload sensitive or other information that can identify the recipient!)
- suggestions for assessment criteria for assignments
- course plans
- instructions
- exercises and discussion questions
- exam questions in various subjects, with suggested answers
- MCQ questions with different answer options
- meeting minutes based on notes
- text for slides in a presentation
Summarising Texts
Generative AI can summarise:
- content from web pages (e.g., a current overview of a specific research area)
- the content of documents
Note: Documents that are copyright-protected should not be uploaded without permission, and sensitive material should not be uploaded at all.
Producing Material
Generative AI can produce:
- images for slides in a presentation
- audio files
- translations
- subtitles for videos
- suggestions for the structure of course modules
- drafts for the outline of a lecture
- program code
Long-term Possibilities
In the USA, generative AI is already used to assess answers in written exams and to provide feedback on student texts. In cases where systematic comparisons have been made with assessments and feedback from teachers, AI has often been as good or better, and it is likely that the systems will develop even more.
In a Swedish context, however, there are still ethical and legal questions that need to be answered before teachers can upload student work, including exams, for analysis by generative AI.
If it becomes formally allowed, and if teachers deem it otherwise appropriate, interesting possibilities for teaching open up. For example, it may suddenly become possible to provide qualified feedback on student work even in large courses with hundreds of students, freeing up teacher time for other efforts in the courses.