9. Conclusions and next steps#

Warning

This page needs expanding.

Here a page for the conclusions. Some important points should be mentioned:

9.1. Summary#

AI might automate some tasks, but it adds the important task of always verifying its work
AI is good to automate the trivial or tedious part of the process where the expert can quickly verify the reliability of the output
AI is good for brainstorming and planning things together, it is a further resource along with articles, books, and other materials you find useful for planning your research
AI might be needed when the amount of data to process is big. Although this is not common in qualitative research, AI can help the researcher with big amount of data. In these cases a sample validation (verifying only a subset of AI results) might be accetable considering the amount of the processed records (e.g. if the researchers verify the AI output for 50% of the data, and the output was always valid, then it is safe to assume that the task was reasonable for the AI and also for the remaining 50% the output might be valid). This of course requires to be disclosed in the manuscript in the dedicated section of ‘Declaration of generative AI use’.

Start using these tools and be transparent about their use. Share the good, the bad, and the ugly about the tools with your local (or global) community.
Follow the recommended practices in your field.
Explore other resources, and feel free to expand this handbook with things you have learned elsewhere
If you have found this handbook helpful, please let us know.