In the previous post I sketched three possible directions for the prototyping phase of the research, with the promise of developing one of them further. This one will start from openings left by the literature review draft of my ongoing thesis and it will explain its possible applicative idea.
Summarising gaps
Reviewing the literature made one absence particularly clear. A considerable amount has been written about how artificial intelligence could, or should, be integrated into the creative process and into data-driven storytelling. Yet, much less has been written about how practitioners actually behave when they sit down and use these tools in their everyday work. The theory on good integration of AI in our work is a lot, but the empirical picture of real habits in the storytelling and data-driven design practices field is almost inexistent.
Sierra Shell makes a related observation in The Human Touch(point), where she notes that understanding the current state of how people use AI features and how they give feedback on them is still an open area for research, rather than settled knowledge. Additionally, the recent surveys on data-driven storytelling and visualization point in a similar direction, treating a human-centred account of the process as an unresolved question rather than as an established one. In other words, before proposing how designers should work with AI, it would be much more useful to document how they are already doing.
Building the idea
To effectively address this empirical gap, there is the need to collect firsthand data. The thesis, therefore, opens up to the need of a survey aimed at students, workers, and experts across communication design, storytelling, and data visualization, with the aim of collecting data and make them later available for further studies and fellow researchers as open source.
The instinctive move would be to reach for a common survey platform. Yet, I would rather build a different tool for ensuring having open data together with the possibility of taking the survey, with also the opportunity of having always updated data in real time. A generic form provider tends to lock results away in a private account, whereas the aim here is the opposite, making the collected evidence a shared and accessible resource.
For the survey section, the priority is currently to define how the tool works and which steps it moves through, not to finalize every question. What remains fixed is the intent: to capture habits of AI usage together with a basic professional profile of who is answering. As for the open data hub, once results begin to accumulate, they will be presented in an aggregated and anonymized form that anyone can consult, without having to complete the survey first and keeping sensitive data out of this public layer.
Encouraging personal reflection
Both parts in this idea, in the end, serve the same purpose, which is to encourage reflection on our behaviours. For the respondent, answering the survey is already a small prompt to consider one’s own reliance on these tools. For a visitor to the open hub, seeing collective patterns laid out invites a comparison with their own practice. The intention is not to lecture nor to attach a score of guilt to anyone’s choices, but to make current habits visible enough to become the source for discussion.
The full bibliography of the thesis will be attached to this documentation, for anyone who wishes to follow the sources behind the literature review.
