What’s of interest? What knowledge workers use LLMs for – Jill Walker Rettberg
Tell me more!
I gave a talk about AI today for a group of knowledge workers in the public sector – people who are asked to write reports summarising research on a specific topic, and who coordinate funding schemes and advise on policy – and I asked them how they use LLMs in their work. Based on their free-text responses on Mentimeter it looks like there are four types of things they have found LLMs actually useful for:
1.Translation and copy-editing – checking the comma rules was one example, another was revising a text to be easier to read
Summarising long texts
3.Structuring information, comparing data (but this was also in the “didn’t work” category)
I also asked for examples of things they had tried to use LLMs for that hadn’t worked:
Literature searches and finding factual information
Analysing a topic or texts, reflective analysis, complex topics, public consultations (høring, an institutional response to a policy proposal),
Summarising large or complex documents
Excel
Making a seating chart
Maintaining ambiguity or multiple perspectives, e.g. in an article with two argumments
Anything related to specialised expertise that is not heavily discussed online
This is one among many items I will regularly tag in Pinboard as oegconnect, and automatically post tagged as #OEGConnect to Mastodon. Do you know of something else we should share like this? Just reply below and we will check it out.
It’s very interesting to see how knowledge workers are actually using LLMs in practice especially the distinction between what works (like translation and summarising) and what doesn’t (such as complex analysis or factual searches).
It clearly shows the importance of using AI as a support tool rather than a replacement for human expertise.
I agree Wisal it is helpful to find out directly how people are actually using LLMs rather than making judgements from news stories.
This reminds me of a bit of work I want to do. In the last two years of the Open Education Awards for Excellence our form has had a place for nominators to be transparent about their use of AI in creating their submissions - we do not place any restrictions.
The use has definitely increased this year, and in many cases it is for checking English language syntax, grammar, some on organizing outlines or brainstorming ideas. I hope to do this soon!
I completely agree , it’s really valuable to see how people are actually using AI tools instead of relying on assumptions from news stories.
I really like the idea of including a space for transparency in the Awards submissions — it’s a great practice that reflects the real use of AI.
From what I’ve seen among knowledge workers, AI is often helpful for summarising, editing, and organizing ideas, while deeper analysis and critical thinking still rely on humans.
AI should be used as a supportive tool, not a replacement for human intelligence.
“AI is a powerful tool, but human intelligence remains superior in reasoning, creativity, and critical thinking.”
