Two Days of Learning How to Use Generative AI in My Research
(Text & Photos: Muqiu Du)
In August 2026, I attended the doctoral course "Using Generative AI in Research". The course was organised by KATAJA and held at the Jyväskylä University School of Business and Economics on 26 and 27 August. I am a second-year PhD student and I work with a very large amount of text data, so the course came at a good moment for me. In this blog post, I would like to share what I learned during these two days and how it has already changed my research plans.
My PhD research focuses on Work Exchange Sharing Accommodation, a niche sharing economy model where travellers exchange their labour for accommodation. My material is a large collection of reviews written by hosts and travellers on platforms such as Workaway and HelpX. In my current article, I use large language models to run aspect-based sentiment analysis on these reviews. Before the course, I had already finished a pretest where I compared different prompting strategies across four LLM APIs on a small sample. That pretest showed me how much potential this method has, and it left me wanting to learn how to apply it with a more rigorous method.
The course was very practical. Instead of talking about artificial intelligence in general, we spent most of the two days working with the tools ourselves. We practised prompt engineering for academic tasks, used AI for literature work and academic writing, coded qualitative material, and compared how different models answered exactly the same instruction. We also got to know some AI tools and learned how to set up open-source models on our own computers using LM Studio, which is a good option for researchers who work with sensitive data. Ethical and legal questions came up on both days and led to good discussions.
Beyond these practical skills, the course also made me think about the place of AI in research more generally. Using it in everyday research now looks like an unstoppable trend. I watched how the instructor works. He sets the instructions, supervises the process, leaves the rest to the model, and then reports openly how he did it when he submits a paper. That is a bolder way of working than anything I had considered. At the same time, journals and reviewers have become sensitive to these questions, and yet AI-use policies also seem to be loosening over time. That combination makes me anxious about how to learn to use these tools well. Where is the real boundary in everyday research? There does not seem to be a clear one. Instead, we simply keep pushing it. And is pushing in this direction actually a good thing, especially once we count the energy costs and the waste that come with it?
The course was not only about tools and methods. It was also nice to meet doctoral researchers from other fields and other universities. Our topics were completely different, but our worries turned out to be very similar. The informal discussions between the sessions were just as useful as the exercises.
Overall, these two days in Jyväskylä were a very valuable experience for my doctoral work. I am sincerely grateful to Matkailukoulutuksen ja -tutkimuksen kannatusyhdistys for their generous support, which made my participation possible.


Kommentit
Lähetä kommentti